llvm-project
[AMDGPU] Ensure v_mfma_scale_f32_{16x16x128|32x32x64}_f8f6f4 instructions are convergent
#178627

Merged

[AMDGPU] Ensure v_mfma_scale_f32_{16x16x128|32x32x64}_f8f6f4 instructions are convergent #178627

frederik-h merged 3 commits into llvm:main from frederik-h:mfma_scale-convergent

[AMDGPU] Ensure v_mfma_scale_f32_{16x16x128|32x32x64}_f8f6f4 instruct…

9d05c22c

frederik-h force pushed to 9d05c22c 196 days ago

frederik-h marked this pull request as ready for review 196 days ago

llvmbot added backend:AMDGPU

frederik-h commented on 2026-01-29

frederik-h requested a review from

arsenm 196 days ago

frederik-h requested a review from

LU-JOHN 196 days ago

arsenm commented on 2026-01-29

Use autogenerated test checks

40574978

arsenm commented on 2026-01-29

Remove IR defines from test

177cb996

arsenm approved these changes on 2026-01-29

frederik-h merged ac00a114 into main 195 days ago

frederik-h deleted the mfma_scale-convergent branch 195 days ago

Reviewers

arsenm

LU-JOHN

Assignees

No one assigned

Labels

backend:AMDGPU

Milestone

No milestone