[mlir][nvgpu] Fix transposeB in nvgpu.warpgroup.mma
#79271
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The llvm#76150 fixed meaning of `transposeB` in NVVM dialect which was initially implemented with opposite meaning. This PR fixes the lowering of `nvgpu.warpgroup.mma` to NVVM dialect. This will fix two integration tests: gemm_f32_f16_f16_128x128x128.mlir gemm_pred_f32_f16_f16_128x128x128.mlir
|
@llvm/pr-subscribers-mlir-gpu @llvm/pr-subscribers-mlir Author: Guray Ozen (grypp) ChangesThe #76150 fixed meaning of This PR fixes the lowering of This will fix two integration tests: Full diff: https://github.com/llvm/llvm-project/pull/79271.diff 1 Files Affected:
diff --git a/mlir/lib/Conversion/NVGPUToNVVM/NVGPUToNVVM.cpp b/mlir/lib/Conversion/NVGPUToNVVM/NVGPUToNVVM.cpp
index 43d05b872a4fbc8..5080956a4589828 100644
--- a/mlir/lib/Conversion/NVGPUToNVVM/NVGPUToNVVM.cpp
+++ b/mlir/lib/Conversion/NVGPUToNVVM/NVGPUToNVVM.cpp
@@ -1407,7 +1407,7 @@ struct NVGPUWarpgroupMmaOpLowering
NVVM::WGMMAScaleOutAttr scaleOut = generateScaleOut();
NVVM::WGMMAScaleInAttr scaleIn = generateScaleIn();
NVVM::MMALayoutAttr layoutA = generateWgmmaLayout(op.getTransposeA());
- NVVM::MMALayoutAttr layoutB = generateWgmmaLayout(op.getTransposeB());
+ NVVM::MMALayoutAttr layoutB = generateWgmmaLayout(!op.getTransposeB());
auto overflow = NVVM::MMAIntOverflowAttr::get(
op->getContext(), NVVM::MMAIntOverflow::wrapped);
|
|
I am submitting this PR without reviews as it is fixes the sm_90 integration tests. The nightly tests did not catch them, because we don't test sm_90 targets yet. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The #76150 fixed meaning of
transposeBin NVVM dialect which was initially implemented with opposite meaning.This PR fixes the lowering of
nvgpu.warpgroup.mmato NVVM dialect.This will fix two integration tests:
gemm_f32_f16_f16_128x128x128.mlir
gemm_pred_f32_f16_f16_128x128x128.mlir