nixlbench: fix cudaFree problem #965
Merged
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
e-ago
requested review from
a team,
aranadive,
brminich and
ovidiusm
as code owners
|
👋 Hi e-ago! Thank you for contributing to ai-dynamo/nixl. Your PR reviewers will review your contribution then trigger the CI to test your changes. 🚀 |
Contributor
Author
|
/build |
Contributor
Author
|
/build |
e-ago
force-pushed
the
fix_gpunetio_bench
branch
from
1caba49 to
25503bd
Compare
Signed-off-by: eagostini <[email protected]>
e-ago
force-pushed
the
fix_gpunetio_bench
branch
from
25503bd to
f9ca79d
Compare
e-ago
changed the title
Signed-off-by: eagostini <[email protected]>
ovidiusm
approved these changes
Oct 30, 2025
aranadive
reviewed
Oct 30, 2025
rakhmets
reviewed
Oct 31, 2025
rakhmets
approved these changes
Nov 3, 2025
Contributor
Author
|
/build |
Contributor
|
/build |
Contributor
|
/ok to test 268bf6a |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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Outdated suggestions cannot be applied.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.
GPUNetIO plugin has a persistent kernel running on a dedicated stream.
NIXL Bench uses the sync memory op cudaFree to free VRAM buffers but the cudaFree requires to synchronize the whole GPU device before proceeding.
Due to the gpunetIo persistent kernel, nixlbench executed with gpunetio plugin hangs on the cudaFree.
This PR resolves this issue replacing cudaFree with cudaFreeAsync