nvidia arm64 & GPU operator test

[jepio]

• Add SkipFunc implementation for skipping test on unsupported instance types
• Add GPU operator test (includes nvidia-runtime sysext test)
• Add Arm64 support to both tests
• Add AWS support

PR Overview

This pull request adds support for NVIDIA GPU testing by introducing a SkipFunc for unsupported instance types, adding a GPU operator test (including an NVIDIA runtime sysext test), and extending support to the ARM64 architecture and AWS platform.

• Introduces skipOnNonGpu to conditionally skip tests on unsupported instances.
• Adds a new test (cl.misc.nvidia.operator) with a complete GPU operator installation and validation workflow.
• Updates existing NVIDIA installation test to incorporate ARM64 support via template configuration.

Reviewed Changes

File	Description
kola/tests/misc/nvidia.go	Added new constants, skip logic, GPU operator test implementation, and expanded platform/architecture support

Copilot reviewed 1 out of 1 changed files in this pull request and generated no comments.

Comments suppressed due to low confidence (2)

kola/tests/misc/nvidia.go:162

• The multi-line helm installation command uses backticks, which preserve literal newlines. Verify that the shell execution handles these newlines as intended, or consider converting it to a single-line command.

_ = c.MustSSH(m, `curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3 \

kola/tests/misc/nvidia.go:101

• The SSH check in waitForNvidiaDriver only verifies for the substring 'active (exited)', which may be too specific if the nvidia service enters other valid states. Consider broadening the check or adding comments to clarify the expected state.

out, err := c.SSH(*m, "systemctl status nvidia.service")