[MLIR][XeGPU] Add lowering from transfer_read/transfer_write to load_gather/store_scatter #152429
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
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
akroviakov
reviewed
Aug 7, 2025
chencha3
reviewed
Aug 8, 2025
adam-smnk
reviewed
Aug 11, 2025
adam-smnk
reviewed
Aug 11, 2025
silee2
reviewed
Aug 11, 2025
chencha3
reviewed
Aug 11, 2025
adam-smnk
reviewed
Aug 13, 2025
charithaintc
approved these changes
Aug 13, 2025
Comment on lines
291
to
295
| for (int64_t dim : vectorShape) { | ||
| auto stepType = VectorType::get({dim}, rewriter.getIndexType()); | ||
| auto stepOp = vector::StepOp::create(rewriter, loc, stepType); | ||
| stepVectors.push_back(stepOp); | ||
| } |
There was a problem hiding this comment.
consider using llvm::map_to_vector
There was a problem hiding this comment.
changed. Not sure this is better? The lamba function has lots of side-effects like IR creation and writing to setpVectors, not a pure function.
chencha3
approved these changes
Aug 13, 2025
| Value localOffsets = broadcasted[0]; | ||
| for (size_t i = 1; i < broadcasted.size(); ++i) { | ||
| localOffsets = | ||
| arith::AddIOp::create(rewriter, loc, localOffsets, broadcasted[i]); |
There was a problem hiding this comment.
nit: braces are not necessary.
adam-smnk
approved these changes
Aug 14, 2025
7 tasks
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Outdated suggestions cannot be applied.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.
Lowering transfer_read/transfer_write to load_gather/store_scatter in case the target uArch doesn't support load_nd/store_nd. The high level steps: