Fix SM resource alignment discovery test#2389
Open
rwgk wants to merge 1 commit into
Open
Conversation
rwgk
added
P0
Contributor
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
Contributor
Author
|
/ok to test 3e7f0fb |
rwgk
added
the
PR review get-together
|
rwgk
marked this pull request as ready for review
July 19, 2026 18:38
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Fix the SM resource discovery alignment test for CUDA 13.4 (currently still in developer preview) reported under nvbug 6464867.
Analysis (gpt-5.6-sol)
SMResourceOptions(count=None)produces an unconstrained structured split request. CUDA 13.4 may satisfy that request with every available SM; on B100, that is 148. The device independently reports its default co-scheduling alignment as 8, but that alignment is not imposed unless the request supplies a co-scheduling constraint. The previous148 % 8 == 0assertion therefore tested a guarantee the API no longer makes.Both values come directly from driver-populated resource structures, and
cuda.bindingsforwards the CUDA 13.4 structured split ABI without calculating either value. This is acuda.coretest expectation issue.Fix
Explicitly request
coscheduled_sm_count=sm_resource.coscheduled_alignmentbefore asserting that the discovered group count is divisible by that alignment. Skip the test when the CUDA driver or bindings predate structured split support (CUDA 13.1), or when the device reports no co-scheduling alignment.