[WaveTransform] Skip lane-mask accumulator reset inside a uniform origin branch - #3854
Merged
cdevadas merged 2 commits intoAug 13, 2026
Conversation
|
PSDB Build Link: http://mlse-bdc-20dd129:8065/#/builders/10/builds/757 |
cdevadas
reviewed
Aug 11, 2026
cdevadas
reviewed
Aug 11, 2026
Author
Sorry, I missed it in a hurry. Updated the comments as suggested now. |
|
PSDB Build Link: http://mlse-bdc-20dd129:8065/#/builders/10/builds/759 |
TejaX-Alaghari
force-pushed
the
amd/dev/tmahatej/wave-transform/fix-reset-in-uniform-originblock
branch
from
August 12, 2026 06:43
1a20290 to
3fd4ba3
Compare
|
PSDB Build Link: http://mlse-bdc-20dd129:8065/#/builders/10/builds/761 |
cdevadas
approved these changes
Aug 13, 2026
cdevadas
deleted the
amd/dev/tmahatej/wave-transform/fix-reset-in-uniform-originblock
branch
August 13, 2026 14:20
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 unconditional accumulator reset inside an origin branch (in step 1.1) is only valid when the accumulator has just been consumed in full. That is, when EXEC has been set to it - which happens only at divergent origin branches (in step 2.2).
At a uniform origin branch nothing consumes the accumulator as EXEC is unrelated to it.
So, zeroing there would discard the contributions of lanes that are inactive in this block but still reach the lane target through another origin.