Add reads_merged flag to track scale read merge status in E2E tests#1227
Open
Hardcode84 wants to merge 3 commits intoiree-org:mainfrom
Open
Add reads_merged flag to track scale read merge status in E2E tests#1227Hardcode84 wants to merge 3 commits intoiree-org:mainfrom
Hardcode84 wants to merge 3 commits intoiree-org:mainfrom
Conversation
Adds a reads_merged parameter to _dbuf_mxfp4_helper and the 4-wave MXFP4 preshuffle E2E tests. The flag records whether merge_contiguous_reads successfully merges all vector<1xi8> scale loads into wider vectors. Tests assert that the actual merge status matches the declared flag, catching both regressions (merged->unmerged) and improvements (unmerged->merged) that need flag updates. Currently only 128x256x256 with wave_shape (1,4) merges; all other configs have unmerged scale reads. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Ivan Butygin <ivan.butygin@gmail.com>
Auto-generates test IDs from block dimensions instead of duplicating them manually in each pytest.param call. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Ivan Butygin <ivan.butygin@gmail.com>
Add 128x256x256 with wave_shape (4,1) and 256x192x256 with wave_shape (2,2) to MXFP4 preshuffle E2E tests. Both configs merge scale reads successfully. Include wave_shape in auto-generated test IDs to distinguish configs sharing the same block dimensions. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> Signed-off-by: Ivan Butygin <ivan.butygin@gmail.com>
Hardcode84
commented
Apr 1, 2026
Hardcode84
commented
Apr 1, 2026
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
1 participant
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.
Adds a reads_merged parameter to _dbuf_mxfp4_helper and the 4-wave MXFP4 preshuffle E2E tests. The flag records whether merge_contiguous_reads successfully merges all vector<1xi8> scale loads into wider vectors. Tests assert that the actual merge status matches the declared flag, catching both regressions (merged->unmerged) and improvements (unmerged->merged) that need flag updates.