ConfigMap informer panics with DeletedFinalStateUnknown type assertion

[infernus01]

Description: The InformedWatcher in knative/pkg/configmap/informer/informed_watcher.go crashes when handling cache.DeletedFinalStateUnknown objects during ConfigMap deletion events.

Error: panic: interface conversion: interface {} is cache.DeletedFinalStateUnknown, not *v1.ConfigMap

Impact:

• Causes webhook pods to crash and restart repeatedly

Root Cause: Unsafe type assertion in event handlers doesn't handle DeletedFinalStateUnknown objects

Proposed Fix: Add safe type assertion with DeletedFinalStateUnknown handling

[infernus01]

/kind bug

[evankanderson]

Hi @infernus01 ,

Thanks for the bug report! Do you feel up for patching this issue, or are you hoping that someone else will do so?

From the description, it sounds like this may be a new state in Kubernetes. What Knative and Kubernetes versions are you using?

[infernus01]

Thanks for looking into this! Here are the version details: knative.dev/pkg version: v0.0.0-20250415155312-ed3e2158b883 Kubernetes client version: v1.32.2
k8s.io/client-go version: v0.32.7 Tekton Pipelines version: v1.2.0-38-g85152e436

And to answer if I'm gonna patch this issue, it is open to anyone who wants to take the initiative, since I've just started with using knative I might be missing some context. The problem that I suspect is The current knative.dev/pkg/configmap/informer/informed_watcher.go code performs unsafe type assertions:

func (i *InformedWatcher) deleteConfigMapEvent(obj interface{}) {
    configMap := obj.(*corev1.ConfigMap)  // UNSAFE - panics on DeletedFinalStateUnknown
    // ...
}

EDIT: I got to see DeletedFinalStateUnkown in the splunk logs when a pipeline is failed due to some exit code 2.

[github-actions[bot]]

This issue is stale because it has been open for 90 days with no activity. It will automatically close after 30 more days of inactivity. Reopen the issue with /reopen. Mark the issue as fresh by adding the comment /remove-lifecycle stale.

[anjalii-28]

I can take this one and work on a fix. Will post a PR shortly.