[all-commits] [llvm/llvm-project] 975c6c: AMDGPU: Shrink used number of registers for mfma s...
Matt Arsenault via All-commits
all-commits at lists.llvm.org
Thu Nov 21 09:00:44 PST 2024
Branch: refs/heads/users/arsenm/gfx950/shrink-mfma-scale-inputs-based-on-format
Home: https://github.com/llvm/llvm-project
Commit: 975c6c066c577171bf51b4dc5b09384ec2d98540
https://github.com/llvm/llvm-project/commit/975c6c066c577171bf51b4dc5b09384ec2d98540
Author: Matt Arsenault <Matthew.Arsenault at amd.com>
Date: 2024-11-21 (Thu, 21 Nov 2024)
Changed paths:
M clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl
M llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp
A llvm/test/Transforms/InstCombine/AMDGPU/mfma-scale.ll
Log Message:
-----------
AMDGPU: Shrink used number of registers for mfma scale based on format
Currently the builtins assume you are using an 8-bit format that requires
an 8 element vector. We can shrink the number of registers if the format
requires 4 or 6.
To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications
More information about the All-commits
mailing list