Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[AIE2P] Legalization and instruction selection support for G_FADD/G_FSUB #274

Open
wants to merge 2 commits into
base: aie-public
Choose a base branch
from

Conversation

katerynamuts
Copy link
Collaborator

No description provided.



---
name: test_fadd_bfloat16
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this requires name change.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

so do rest of the tests.

; AIE2P-NEXT: [[UV2:%[0-9]+]]:_(<16 x s32>), [[UV3:%[0-9]+]]:_(<16 x s32>) = G_UNMERGE_VALUES [[UV]](<32 x s32>)
; AIE2P-NEXT: [[INT:%[0-9]+]]:_(<16 x s16>) = G_INTRINSIC_W_SIDE_EFFECTS intrinsic(@llvm.aie2p.v16accfloat.to.v16bf16), [[UV2]](<16 x s32>)
; COMMON-NEXT: [[AIE_SEXT_EXTRACT_VECTOR_ELT32:%[0-9]+]]:_(s32) = G_AIE_SEXT_EXTRACT_VECTOR_ELT [[INT]](<16 x s16>), [[C1]](s32)
; COMMON-NEXT: [[ASSERT_SEXT:%[0-9]+]]:_(s32) = G_ASSERT_SEXT [[AIE_SEXT_EXTRACT_VECTOR_ELT32]], 16
; COMMON-NEXT: $r0 = COPY [[ASSERT_SEXT]](s32)
; COMMON-NEXT: PseudoRET implicit $lr, implicit $r0
%0:_(s32) = COPY $r1
Copy link
Collaborator

@SagarMaheshwari99 SagarMaheshwari99 Jan 20, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you add a couple of tests for vectors, for ex. <16 x s32> , same for fadd..

@mludevid
Copy link
Collaborator

mludevid commented Feb 4, 2025

I've updated the legalization commit. It now legalizes the non-vector case properly. I've also split up the test for vectors on aie2 and aie2p, because the two architectures perform the G_FADD and G_FSUB operations on different vector sizes.

@mludevid mludevid marked this pull request as ready for review February 4, 2025 18:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants