Manager panics when executed without a valid kube context

[mdbooth]

/kind bug

podman run --rm -ti gcr.io/k8s-staging-capi-openstack/capi-openstack-controller:v0.10.1 -- --version    
E0503 16:03:01.692120       1 config.go:133] "unable to load in-cluster config" err="unable to load in-cluster configuration, KUBERNETES_SERVICE_HOST and KUBERNETES_SERVICE_PORT must be defined" logger="controller-runtime.client.config"
E0503 16:03:01.693959       1 main.go:235] "unable to get kubeconfig" err="invalid configuration: no configuration has been provided, try setting KUBERNETES_MASTER environment variable" logger="setup"
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x1d0 pc=0x1ba6cc8]

goroutine 1 [running]:
main.main()
        /workspace/main.go:237 +0x3c8

Looks like we're missing an exit here:

cluster-api-provider-openstack/main.go

Lines 231 to 235 in 2214514

	cfg, err := config.GetConfigWithContext(os.Getenv("KUBECONTEXT"))
	if err != nil {
	setupLog.Error(err, "unable to get kubeconfig")
	}
	cfg.QPS = restConfigQPS

Note that it didn't crash in v0.9.0, but I suspect that was due to luck. I'm guessing this broke when we bumped controller-runtime.