Cleanup usage of NCCL error codes #357
Merged
Conversation
This file contains 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
topo_file_lock is static initialized, so there is no need to call pthread_mutex_init() during the rdma_init() call. Signed-off-by: Brian Barrett <[email protected]>
rajachan
requested changes
Feb 27, 2024
AmedeoSapio
reviewed
Feb 28, 2024
Last year, we chose to simplify error code management and only use Libfabric error codes or -errno codes (which should be equivalent) in the core of the code, translating at the API layer. This patch finishes the previously incomplete patches to make that transition. Signed-off-by: Brian Barrett <[email protected]>
bwbarrett
force-pushed
the
cleanup/errorcodes
branch
from
4b95c2b to
a77cdbc
Compare
AmedeoSapio
approved these changes
Feb 28, 2024
rajachan
approved these changes
Feb 28, 2024
bwbarrett
added
BuildTriggerRequest
AvivBenchorin
added
the
BuildTriggerRequest
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Outdated suggestions cannot be applied.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.
The majority of this PR is removing the use of NCCL error codes in the core of the library, since we only use Libfabric error codes / -errno error codes to simplify reasoning about what error code is what. This is a follow-on to last year's commit that made the change, since it looks like we missed a couple hundred places.
While cleaning that up, I noticed a place where we double initialized a mutex. While harmless, it's also wrong, so clean that up as well.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.