[BUG] [CONV] Fix incorrect stride calculation when w=1/h=1 in MISA solvers #3786
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
SreecharanGundaboluAMD
requested review from
BradPepersAMD,
BrianHarrisonAMD,
JonathanLichtnerAMD and
adickin-amd
as code owners
BradPepersAMD
approved these changes
Jun 23, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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Outdated suggestions cannot be applied.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.
Overview
This pull request addresses a critical bug in the FWD, WRW and BWD solvers where the stride was calculated incorrectly due to buggy manual flipping of inputs and outputs.
Can be reproduced using:
MIOpenDriver conv -n 3 -c 1 -H 1 -W 1 -k 3 -y 3 -x 2 -p 2 -q 1 -u 2 -v 1 -l 1 -j 1 -m conv -g 1 -F 4 -t 0Bugfix Details
The manual direction flips previously implemented in the FWD, WRW and BWD solvers have been replaced with methods from the
ProblemInterpreterclass which perform flips based on the convolution direction, ensuring consistency and accuracy. TheProblemInterpreterclass is already utilized in most CK solvers.Enhancements
1.Regression Tests
Added regression tests to relevant unit tests to ensure the correctness of the stride calculation and the overall functionality of the WRW and BWD solvers.
2. Code Cleanup:
Removed WORKAROUND_ISSUE_2867, as this issue is effectively addressed by the bugfix implemented in this pull request.