[backend] Fix improper mma->dot shortcut when warpsPerCTA[1] > 1
#4803
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
triton-lang#4492 started causing an issue where chained MMAs on hopper would segfault with 8 warps. It seems that previously this was checked, but the check got removed in this PR and it's still unsupported. Adding back this check means these MMAs will have to go back to shared memory, but it's better than segfaulting until it's actually supported. Resolves openxla/xla#17356
|
Can you copy & paste me the mma->mma instruction and the layouts? |
Does openxla/xla#17356 (comment) contain what you need? |
|
Oh, I think it's mma->dot instead of mma->mma. It's not supposed to be handled by my previous PR. Why did I take it out...I need some further investigation before getting back to you. |
|
I agree with the fix. I actually didn't take out the Fix improper mma->dot shortcut when |
chsigg
changed the title
warpsPerCTA[1] > 1
Jokeren
approved these changes
Sep 25, 2024
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Outdated suggestions cannot be applied.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.
#4492 started causing an issue where chained MMAs on hopper would segfault with 8 warps. It seems that previously this was checked, but the check got removed in this PR and it's still unsupported.
Adding back this check means these MMAs will have to go back to shared memory, but it's better than segfaulting until it's actually supported.
Resolves openxla/xla#17356