Add test for bind() and CUDA initialization. - #1147
Draft
TomAugspurger wants to merge 3 commits into
Draft
Conversation
NVIDIA/cuCascade@8a3b42f changed cucascade's topology discovery in a way that ultimately broke multi-GPU cudf-polars. Some part of rapidsmpf's ``bind`` is sensitive to whether or not CUDA is initialized during topology discovery. This PR adds a test to ensure we catch it sooner if this happens again. #1145 pinned release/26.08 to a known-good commit. #1146 forward-merged that to `main`. But we don't want that pin on `main`: we'd prefer to catch regressions and benefit from improvements immediately. So the second commit reverts that pin.
|
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 69075ec |
TomAugspurger
added
improvement
Contributor
Author
|
I expect the new test to fail, since we should be building rapidsmpf against a cucascade commit with the problematic rmm usage in topology discovery. |
Contributor
Author
|
/ok to test d4bf9ee |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
NVIDIA/cuCascade@8a3b42f
changed cucascade's topology discovery in a way that ultimately broke multi-GPU
cudf-polars.
Some part of rapidsmpf's
bindis sensitive to whether or not CUDA is initializedduring topology discovery. This PR adds a test to ensure we catch it sooner
if this happens again.
#1145 pinned release/26.08 to a known-good commit.
#1146 forward-merged that to
main. But wedon't want that pin on
main: we'd prefer to catch regressions and benefit from improvementsimmediately. So the second commit reverts that pin.