Fix TensorMemoryAllocation to correctly implement getAlloc #6614
+49
−2
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
ThomasRaoux
reviewed
Apr 28, 2025
Comment on lines
183
to
187
| auto arg = cast<BlockArgument>(subOp.getSrc()); | ||
| auto wsOp = op->getParentOfType<triton::gpu::WarpSpecializeOp>(); | ||
| auto capture = wsOp.getExplicitCaptures()[arg.getArgNumber()]; |
There was a problem hiding this comment.
you need to check or assert that the arg belongs to wsOp. Best would be to get wsOp from arg.getOwner()
Mogball
reviewed
Apr 28, 2025
| while (isa<triton::gpu::MemDescSubviewOp>(op)) { | ||
| op = op->getResult(0).getDefiningOp(); | ||
| while (auto subOp = dyn_cast<triton::gpu::MemDescSubviewOp>(op)) { | ||
| if (subOp.getSrc().getDefiningOp()) { |
There was a problem hiding this comment.
Suggested change
| if (subOp.getSrc().getDefiningOp()) { |
I don't think this is ever null.
There was a problem hiding this comment.
It is null for the lit test case where it comes from arguments of warp_specialize. Is the lit test case valid?
There was a problem hiding this comment.
Oh sorry you're right. I think I had a brain fart
Summary: Test Plan: Reviewers: Subscribers: Tasks: Tags:
FindHao
pushed a commit
to FindHao/triton
that referenced
this pull request
Apr 30, 2025
…ng#6614) Prior to the fix, getAlloc will hit an infinite loop. This patch also handles the case where tmemDesc comes from argument of warp_specialize.
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Outdated suggestions cannot be applied.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.
Prior to the fix, getAlloc will hit an infinite loop. This patch also handles the case where tmemDesc comes from argument of warp_specialize.