fix node permissions for computedomain plugin

[rahulait]

Description

These permissions are coming from: https://github.com/kubernetes-sigs/dra-driver-nvidia-gpu/blob/9c52a7d50994adbf2fbb5f1ce2f6466fa3f9936f/deployments/helm/dra-driver-nvidia-gpu/templates/rbac-kubeletplugin.yaml#L17-L19

Checklist

• No secrets, sensitive information, or unrelated changes
• Lint checks passing (make lint)
• Generated assets in-sync (make validate-generated-assets)
• Go mod artifacts in-sync (make validate-modules)
• Test cases are added for new code paths

Testing

Verified using:

kubectl auth can-i patch nodes   --as=system:serviceaccount:gpu-operator:nvidia-dra-driver-kubeletplugin
Warning: resource 'nodes' is not namespace scoped

yes

[tariq1890]

Don't we need to add the other permissions mentioned in the ClusterRole from the upstream DRA driver helm chart?

[rahulait]

Looks like ubuntu/canonical apt servers are having issues, driver installs are taking ages to finish and our e2e tests won't finish on time.

[rahulait]

Don't we need to add the other permissions mentioned in the ClusterRole from the upstream DRA driver helm chart?

Other required permissions remain in the existing general clusterrole https://github.com/NVIDIA/gpu-operator/blob/main/manifests/state-dra-driver/0200_kubeletplugin-clusterrole.yaml. Are you looking for any specific RBAC which is missing?