[Layout] Enhance layout inference pass #427
Merged
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Outdated suggestions cannot be applied.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.
This pull request introduces enhancements to layout inference and optimization for parallel operations, along with a minor change in a Python wrapper file. The most significant updates involve adding support for multiple inference levels and improving layout computation logic.
Enhancements to layout inference and optimization:
Added multiple inference levels for layout optimization: Introduced
InferLevelwith three levels (kStrict,kCommon,kFree) to control the aggressiveness of layout inference and optimization. Each level has distinct behavior, ranging from conservative to highly permissive. This change is documented inParallelOp::InferLayout.Updated layout inference logic in
ParallelOp::InferLayout: Enhanced the method to handle cases whereread_source_bufferis defined, improving layout computation flexibility and accuracy.Iterative layout inference in
Stmt Copy::Lower: Modified theLowermethod to iterate over all inference levels (kCommon,kStrict,kFree) for layout inference, providing more robust layout optimization.Miscellaneous:
tilelang/jit/adapter/wrapper.py: Added areturn 0;statement to theinitfunction, ensuring proper function termination.