Update Triton to c75c6b034756629b891e7b2df406f634552331d5 #223
Conversation
This file contains 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
Add `sanitize_overflow: bool = True` to class CPUOptions in compiler.py and `get_benchmarker(self)` to class CPUDriver in driver.py to run the tests. XFAILing TritonToLinalg tests since this pass will be retire soon: test/Conversion/TritonToLinalg/wraparound_side_by_side.mlir test/Conversion/TritonToLinalg/wraparound_stacked.mlir XFAILing StructuredToMemref tests due to LLVM commit 889b67c9d30e3024a1317431d66c22599f6c2011 asserts that dynamic shapes like <2x?> and <?x?> are mismatch: test/Conversion/StructuredToMemref/wraparound_side_by_side.mlir test/Conversion/StructuredToMemref/wraparound_stacked.mlir
Add rewriteSplatOp() in PtrAnalysis pass. This function creates a
tts.makeptr for the case below:
%6 = tt.splat %arg0 : !tt.ptr<i32> -> tensor<1x!tt.ptr<i32>>
Previously we rely on rewriteAddPtrOp to create the tts.makeptr:
%3 = arith.constant 0 : index
%6 = tt.splat %arg0 : !tt.ptr<i32> -> tensor<1x!tt.ptr<i32>>
%7 = tt.addptr %6, %3 : tensor<1x!tt.ptr<i32>>, tensor<1xi32>
Creating a constant 0 and adding it to a pointer is optimized away by
Triton.
This reverts commit 2153c53. The commit being reverted will be sent in a separate PR.
Thanks Nhat Nguyen for the fix. UnXFAILed and updated wraparound_side_by_side.mlir and wraparound_stacked.mlir in test/Conversion/StructuredToMemref.
We use git submodule to checkout Triton at specific commit.
Fix various build issues and lit test failures: LLVM API changes: x.get<M> --> cast<M>(x) y.is<N> --> isa<N>(Y) applyPatternsAndFoldGreedily() --> applyPatternsGreedily() Float32Type::get(somType.getContext()).getWidth() --> cast<FloatType>(Float32Type::get(elemType.getContext())).getWidth() Fix link error related to TritonGPUTC05MMAPipeline.
This reverts commit 1cbcc57.
Tests added are either hard-coded to use 'cuda' device or with 'ttg' dialect.
zhaoshiz
changed the title
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.
Trying to fix #178
This PR includes cosmetic changes due to LLVM API change, fix for link error, lit test update and adding unsupported tests in conftest.py