Retry device enumeration on startup to prevent empty ResourceSlices

[kasia-kujawa]

Fixes #1008

Added a retry loop in NewDeviceState().
If the first enumeration returns 0 devices, the plugin retries every 5 seconds for up to 5 minutes before proceeding.
Errors still propagate immediately without retry.

Before the fix, no log was emitted after Traverse GPU devices and the empty ResourceSlice was published silently.

With the fix (nvidiaDriverRoot: /home/kubernetes/bin/nvidia/, GKE COS, Tesla T4):

I0410 08:11:16.614833  1 nvlib.go:197] Traverse GPU devices
I0410 08:11:16.779628  1 device_state.go:96] No GPU devices found yet (driver may still be initializing), retrying in 5s...
I0410 08:11:21.780288  1 nvlib.go:197] Traverse GPU devices
I0410 08:11:23.111832  1 nvlib.go:278] Adding device gpu-0 to allocatable devices
I0410 08:11:23.111862  1 allocatable.go:254] Adding allocatables for PCI bus ID: 0000:00:04.0

Full logs with the fix from nvidia-dra-driver-gpu-kubelet-plugin when GPU initialization needed slightly more time:
https://gist.github.com/kasia-kujawa/1082b48357a0ae80d663f12ee665e34c

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: kasia-kujawa
Once this PR has been reviewed and has the lgtm label, please assign shivamerla for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment