Bug: IsMicroShiftCluster: log timeout instead of calling Skip from container node

IsMicroShiftCluster() calls skipper.Skip() when the API poll times out.
When called from a Ginkgo container node context (e.g. InitializeTestFramework
during suite setup), Ginkgo v2 panics because Skip() is only valid inside
leaf nodes. This crashes the test binary and cascades all remaining tests
as failures.

Replace the Skip() call with a log message and let the function fall
through returning (false, nil). This allows framework initialization to
succeed even when the API is unavailable, so that each test's own
BeforeEach health checks (e.g. SkipIfClusterIsNotHealthy) can properly
skip the test instead of the binary crashing.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: fonta-rh

test/extended/util/framework.go

@@ -2341,8 +2341,7 @@ func IsMicroShiftCluster(kubeClient k8sclient.Interface) (bool, error) {
 		e2e.Logf("IsMicroShiftCluster: error accessing microshift-version configmap: %v", err)
 		return false, nil
 	}); err != nil {
-		// Timeout accessing the configmap is a precondition failure
-		skipper.Skip(preconditions.FormatSkipMessage(fmt.Sprintf("microshift-version configmap check timed out after %s: %v", duration, err)))
+		e2e.Logf("IsMicroShiftCluster: timed out after %s: %v (assuming not MicroShift)", duration, err)
 	}
 	if cm == nil {
 		e2e.Logf("IsMicroShiftCluster: microshift-version configmap not found, not MicroShift")