[TIR] Enhance LowerThreadAllreduce
pass to automatically infer shared memory scope
#17442
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Outdated suggestions cannot be applied.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 pass
LowerThreadAllreduce
enables efficient block reduction. However, block reduction often requires a large amount of shared memory space. The current implementation ofLowerThreadAllreduce
only enable static shared memory reduce buffer allocation and prevents the shared memory merging when another shared memory scope is defined asshared.dyn
.This pull request addresses this issue by first collecting buffer allocations, and then determining the memory scope of the reduction buffer, allowing for memory space fusion in the following
MergeSharedMemoryAllocations
pass.