[UR] Fix some tests that are broken when run with multiple cuda devices available. #17216
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
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
force-pushed
the
aaron/fixTestsForMultiDevRunner
branch
from
45dac1b to
b5c390b
Compare
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
force-pushed
the
aaron/fixTestsForMultiDevRunner
branch
from
b5c390b to
2b62065
Compare
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
aarongreig
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
|
e2e fails are unrelated: this only touches UR test code |
kbenzie
approved these changes
Mar 3, 2025
|
just want to confirm, after this is merged do we want to change the cuda runner to the multi-gpu one? or perhaps we want to add a separate workflow for the new runner...? |
I think the idea was to use it to replace the current one because the multi-gpu one is faster, maybe @pbalcer can confirm |
|
ping @intel/llvm-reviewers-cuda |
Seanst98
approved these changes
Mar 4, 2025
|
@intel/llvm-gatekeepers please merge, the e2e fails aren't related as this only touches UR tests |
|
FYI, enabled new runner, turned off the old one. If you see any issues please let me know |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Also removes a test and adds known failures where appropriate (typically where the test only runs when multiple devices are available so the skip doesn't affect behaviour of single-device runs).
The test removed is
cudaUrContextCreateTest.ActiveContext. This test seems to be testing the assumption that aurQueueCreatefollowed byurMemBufferCreatewill set the active cuda context to the one associated with the context passed to those calls. Neither of these calls set the active context, this may have changed at some point as the test dates back to a PI unit test. The test currently passes as long as only one device is available because a previousurDeviceGetInfosets the active context to the one associated with that device, which is inevitably the same as the one associated with the UR context used in the test. Since the test is based on a faulty assumption about the adapter I think we can just delete it.