Query CUDA forward compatibility elf note if available#1598
Merged
elezar merged 3 commits intoNVIDIA:mainfrom Jan 23, 2026
Merged
Query CUDA forward compatibility elf note if available#1598elezar merged 3 commits intoNVIDIA:mainfrom
elezar merged 3 commits intoNVIDIA:mainfrom
Conversation
Signed-off-by: Evan Lezar <elezar@nvidia.com>
This change is a minor refactor to the enable-cuda-compat hook to allow the mechanism for determining the compat libraries in the container to be extended more easily. Signed-off-by: Evan Lezar <elezar@nvidia.com>
e23b06e to
7902588
Compare
Pull Request Test Coverage Report for Build 21216900667Details
💛 - Coveralls |
This change queries an ELF note section in libcuda in the container to determine whether the forward compat libraries in the container should be used over the host drivers. If the elf note section is not available, we fall back to the heuristics using the major numbers of the host and compat driver libraries. Signed-off-by: Evan Lezar <elezar@nvidia.com>
7902588 to
e1428d9
Compare
ArangoGutierrez
approved these changes
Jan 23, 2026
Collaborator
ArangoGutierrez
left a comment
There was a problem hiding this comment.
LGTM - One non blocking nit
| m.logger.Debugf("Compat major version is not greater than the host driver major version (%v >= %v)", hostDriverVersion, compatDriverVersion) | ||
| return "", nil | ||
| // First check the elf header. | ||
| cudaCompatHeader, _ := GetCUDACompatElfHeader(libcudaCompatPath) |
Collaborator
There was a problem hiding this comment.
Nit: Could we at least log the error?
Member
Author
There was a problem hiding this comment.
Firstly, this is in a hook and the logs are not captured and as such logs are not that helpful. Secondly, I want to explicitly ignore errors at this stage since I want to fall back to the heuristics that we already have in place.
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
3 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.
This change queries an ELF note section in libcuda in the container to determine whether the forward compat libraries in the container should be used over the host drivers.
If the elf note section is not available, we fall back to the heuristics using the major numbers of the host and compat driver libraries.