Add Oracle Linux (ol) support

[bolt1777]

Description

Adds support for Oracle Linux nodes (os-release ID=ol), mirroring the existing Rocky Linux support

Why

Nodes running Oracle Linux are detected via the NFD label feature.node.kubernetes.io/system-os_release.ID=ol. So this PR brings support Oracle Linux OS.

Testing

• Tests: new cases in TestGetGPUNodeOSInfo and TestGetOSTag covering Oracle Linux 9 and 10.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[rajatchopra]

We need to add OL to end-to-end tests. Now or eventually.

[tariq1890]

@bolt1777 We do not officially support driver containers for oracle linux. Are you using your custom-built driver images for Oracle Linux?

[bolt1777]

Hi, we are using images made for rocky linux, but would be also great to have a native support of OL

[tariq1890]

Thanks for the context @bolt1777. Are you able to share driver container scripts for Oracle Linux that you have used for testing? Until we validate these changes on an Oracle Linux environment with the appropriate driver container, we can't merge this PR.

[bolt1777]

The existing script works well on Oracle Linux. I tested it with the following driver container image:

nvcr.io/nvidia/driver:595.71.05-rocky9

It handles both installation and uninstallation correctly without requiring any additional custom steps. The only actions needed are loading the necessary kernel modules and disabling SELinux filesystem.
The main missing piece is support for building drivers directly on Oracle Linux base images, along with adding native support for Oracle Linux in the driver container repository.

Repository reference:
https://github.com/NVIDIA/gpu-driver-container

I’m happy to prepare a PR that adds this support so it can be reviewed and validated. Would that approach work for you?