[truongbo17]

Summary

Log warnings when nvsandboxutils or NVML dGPU discoverer initialization fails, for both full GPU and MIG devices. The warning names the failing backend and includes the original error, while the existing fallback and joined-error behavior stays intact.

Fixes #2114.

Tests

• Added regression tests for all four warning paths (confirmed failing before the fix).
• make build, make fmt, and make test passed in a Linux Go container.

[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.

[myeolenv]

Thanks for the contribution. We will review this PR soon.

[henry118]

@truongbo17 Thanks for picking up the TODO. A couple of thoughts:

• All-failed case: the error is already returned and logged at the top level.
• One discoverer fails: this is the case nothing logs today.

imo we just log when at least one discoverer was created and another failed, at debug level, like:

if len(discoverers) == 0 {
    return nil, errs
}
if errs != nil {
    o.logger.Debugf("Ignoring discoverer errors: %v", errs)
}

The same change is needed in NewForMigDevice.

And I’d also drop the new tests that are just check log messages.