Release v6.1.0

Add new integral int1e_grids_ipip
sunqm · Nov 11, 2023 · b9a340f · b9a340f
1 parent 47e9b55
commit b9a340f
Show file tree

Hide file tree

Showing 5 changed files with 113 additions and 8 deletions.
diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -1,7 +1,7 @@
 cmake_minimum_required (VERSION 3.5)
 project (qcint C)
 set(qcint_VERSION_MAJOR "6")
-set(qcint_VERSION_MINOR "0")
+set(qcint_VERSION_MINOR "1")
 set(qcint_VERSION_PATCH "0")
 set(qcint_VERSION_TWEAK "0")
 set(qcint_VERSION "${qcint_VERSION_MAJOR}.${qcint_VERSION_MINOR}.${qcint_VERSION_PATCH}")

diff --git a/ChangeLog b/ChangeLog
@@ -1,3 +1,6 @@
+Version 6.1.0 (2023-11-10):
+	* New second order derivative integrals int1e_grids_ipip
+
 Version 6.0.0 (2023-09-17):
 	* Add polynomial root finder using the eigenvalue algorithm
 	* Fitting Rys roots of SR integrals with Chebyshev polynomials

diff --git a/README.md b/README.md
@@ -3,8 +3,8 @@ qcint (quick libcint)
 
 An optimized libcint branch for X86 platform
 
-version 6.0.0
-2023-09-17
+version 6.1.0
+2023-11-10
 
 
 What is qcint

diff --git a/scripts/auto_intor.cl b/scripts/auto_intor.cl
@@ -225,15 +225,15 @@
   '("int1e_ipiprinvip"          ( nabla nabla \| rinv \| nabla ))
 )
 
-(gen-cint "deriv3.c"
-  '("int1e_ipipipnuc"           ( nabla nabla nabla \| nuc \| ))
-  '("int1e_ipipiprinv"          ( nabla nabla nabla \| rinv \| ))
-  '("int1e_ipipnucip"           ( nabla nabla \| nuc \| nabla ))
-  '("int1e_ipiprinvip"          ( nabla nabla \| rinv \| nabla ))
+(gen-cint "deriv4.c"
+  '("int1e_ipiprinvipip"        ( nabla nabla \| rinv \| nabla nabla ))
+  '("int1e_ipipiprinvip"        ( nabla nabla nabla \| rinv \| nabla ))
+  '("int1e_ipipipiprinv"        ( nabla nabla nabla nabla \| rinv \| ))
 )
 
 (gen-cint "int1e_grids1.c"
   '("int1e_grids_ip"            ( nabla \| grids \| ))
   '("int1e_grids_ipvip"         ( nabla \| grids \| nabla ))
   '("int1e_grids_spvsp"         ( sigma dot p \| grids \| sigma dot p ))
+  '("int1e_grids_ipip"          ( nabla nabla \| grids \| ))
 )
diff --git a/src/autocode/int1e_grids1.c b/src/autocode/int1e_grids1.c
@@ -302,3 +302,105 @@ return CINT1e_grids_spinor_drv(out, dims, &envs, cache, &c2s_si_1e_grids);
 } // int1e_grids_spvsp_spinor
 ALL_CINT1E(int1e_grids_spvsp)
 ALL_CINT1E_FORTRAN_(int1e_grids_spvsp)
+/* <NABLA NABLA i| 1/r_{grids} |j> */
+void CINTgout1e_int1e_grids_ipip(double *RESTRICT gout, double *RESTRICT g,
+int *idx, CINTEnvVars *envs, int gout_empty) {
+int ngrids = envs->ngrids;
+int bgrids = MIN(ngrids - envs->grids_offset, GRID_BLKSIZE);
+bgrids = ALIGN_UP(bgrids, SIMDD);
+int nf = envs->nf;
+int nrys_roots = envs->nrys_roots;
+int ix, iy, iz, n, i, ig;
+double *RESTRICT g0 = g;
+double *RESTRICT g1 = g0  + envs->g_size * 3;
+double *RESTRICT g2 = g1  + envs->g_size * 3;
+double *RESTRICT g3 = g2  + envs->g_size * 3;
+__MD rs[9];
+G1E_GRIDS_D_I(g1, g0, envs->i_l+1, envs->j_l);
+G1E_GRIDS_D_I(g2, g0, envs->i_l+0, envs->j_l);
+G1E_GRIDS_D_I(g3, g1, envs->i_l+0, envs->j_l);
+if (gout_empty) {
+for (n = 0; n < nf; n++) {
+ix = idx[0+n*3];
+iy = idx[1+n*3];
+iz = idx[2+n*3];
+for (ig = 0; ig < bgrids; ig += SIMDD) {
+for (i = 0; i < 9; i++) { rs[i] = MM_SET1(0.); }
+for (i = 0; i < nrys_roots; i++) {
+rs[0] += MM_LOAD(g3+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[1] += MM_LOAD(g2+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[2] += MM_LOAD(g2+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iz+ig+i*GRID_BLKSIZE);
+rs[3] += MM_LOAD(g1+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[4] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g3+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[5] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iz+ig+i*GRID_BLKSIZE);
+rs[6] += MM_LOAD(g1+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iz+ig+i*GRID_BLKSIZE);
+rs[7] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iz+ig+i*GRID_BLKSIZE);
+rs[8] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g3+iz+ig+i*GRID_BLKSIZE);
+}
+MM_STORE(gout+ig+bgrids*(n*9+0),  + rs[0]);
+MM_STORE(gout+ig+bgrids*(n*9+1),  + rs[3]);
+MM_STORE(gout+ig+bgrids*(n*9+2),  + rs[6]);
+MM_STORE(gout+ig+bgrids*(n*9+3),  + rs[1]);
+MM_STORE(gout+ig+bgrids*(n*9+4),  + rs[4]);
+MM_STORE(gout+ig+bgrids*(n*9+5),  + rs[7]);
+MM_STORE(gout+ig+bgrids*(n*9+6),  + rs[2]);
+MM_STORE(gout+ig+bgrids*(n*9+7),  + rs[5]);
+MM_STORE(gout+ig+bgrids*(n*9+8),  + rs[8]);
+}}} else {
+for (n = 0; n < nf; n++) {
+ix = idx[0+n*3];
+iy = idx[1+n*3];
+iz = idx[2+n*3];
+for (ig = 0; ig < bgrids; ig += SIMDD) {
+for (i = 0; i < 9; i++) { rs[i] = MM_SET1(0.); }
+for (i = 0; i < nrys_roots; i++) {
+rs[0] += MM_LOAD(g3+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[1] += MM_LOAD(g2+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[2] += MM_LOAD(g2+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iz+ig+i*GRID_BLKSIZE);
+rs[3] += MM_LOAD(g1+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[4] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g3+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iz+ig+i*GRID_BLKSIZE);
+rs[5] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iz+ig+i*GRID_BLKSIZE);
+rs[6] += MM_LOAD(g1+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iz+ig+i*GRID_BLKSIZE);
+rs[7] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g1+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g2+iz+ig+i*GRID_BLKSIZE);
+rs[8] += MM_LOAD(g0+ix+ig+i*GRID_BLKSIZE) * MM_LOAD(g0+iy+ig+i*GRID_BLKSIZE) * MM_LOAD(g3+iz+ig+i*GRID_BLKSIZE);
+}
+MM_STORE(gout+ig+bgrids*(n*9+0), MM_LOAD(gout+ig+bgrids*(n*9+0)) +  + rs[0]);
+MM_STORE(gout+ig+bgrids*(n*9+1), MM_LOAD(gout+ig+bgrids*(n*9+1)) +  + rs[3]);
+MM_STORE(gout+ig+bgrids*(n*9+2), MM_LOAD(gout+ig+bgrids*(n*9+2)) +  + rs[6]);
+MM_STORE(gout+ig+bgrids*(n*9+3), MM_LOAD(gout+ig+bgrids*(n*9+3)) +  + rs[1]);
+MM_STORE(gout+ig+bgrids*(n*9+4), MM_LOAD(gout+ig+bgrids*(n*9+4)) +  + rs[4]);
+MM_STORE(gout+ig+bgrids*(n*9+5), MM_LOAD(gout+ig+bgrids*(n*9+5)) +  + rs[7]);
+MM_STORE(gout+ig+bgrids*(n*9+6), MM_LOAD(gout+ig+bgrids*(n*9+6)) +  + rs[2]);
+MM_STORE(gout+ig+bgrids*(n*9+7), MM_LOAD(gout+ig+bgrids*(n*9+7)) +  + rs[5]);
+MM_STORE(gout+ig+bgrids*(n*9+8), MM_LOAD(gout+ig+bgrids*(n*9+8)) +  + rs[8]);
+}}}}
+void int1e_grids_ipip_optimizer(CINTOpt **opt, int *atm, int natm, int *bas, int nbas, double *env) {
+int ng[] = {2, 0, 0, 0, 2, 1, 0, 9};
+CINTall_1e_grids_optimizer(opt, ng, atm, natm, bas, nbas, env);
+}
+CACHE_SIZE_T int1e_grids_ipip_cart(double *out, int *dims, int *shls,
+int *atm, int natm, int *bas, int nbas, double *env, CINTOpt *opt, double *cache) {
+int ng[] = {2, 0, 0, 0, 2, 1, 0, 9};
+CINTEnvVars envs;
+CINTinit_int1e_grids_EnvVars(&envs, ng, shls, atm, natm, bas, nbas, env);
+envs.f_gout = &CINTgout1e_int1e_grids_ipip;
+return CINT1e_grids_drv(out, dims, &envs, cache, &c2s_cart_1e_grids);
+} // int1e_grids_ipip_cart
+CACHE_SIZE_T int1e_grids_ipip_sph(double *out, int *dims, int *shls,
+int *atm, int natm, int *bas, int nbas, double *env, CINTOpt *opt, double *cache) {
+int ng[] = {2, 0, 0, 0, 2, 1, 0, 9};
+CINTEnvVars envs;
+CINTinit_int1e_grids_EnvVars(&envs, ng, shls, atm, natm, bas, nbas, env);
+envs.f_gout = &CINTgout1e_int1e_grids_ipip;
+return CINT1e_grids_drv(out, dims, &envs, cache, &c2s_sph_1e_grids);
+} // int1e_grids_ipip_sph
+CACHE_SIZE_T int1e_grids_ipip_spinor(double complex *out, int *dims, int *shls,
+int *atm, int natm, int *bas, int nbas, double *env, CINTOpt *opt, double *cache) {
+int ng[] = {2, 0, 0, 0, 2, 1, 0, 9};
+CINTEnvVars envs;
+CINTinit_int1e_grids_EnvVars(&envs, ng, shls, atm, natm, bas, nbas, env);
+envs.f_gout = &CINTgout1e_int1e_grids_ipip;
+return CINT1e_grids_spinor_drv(out, dims, &envs, cache, &c2s_sf_1e_grids);
+} // int1e_grids_ipip_spinor
+ALL_CINT1E(int1e_grids_ipip)
+ALL_CINT1E_FORTRAN_(int1e_grids_ipip)