### Introducing Graceful Node Shutdown (Alpha)

- For raw block CSI volumes on Kubernetes, kubelet was incorrectly calling CSI NodeStageVolume for every single "map" (i.e. raw block "mount") operation for a volume already attached to the node. This PR ensures it is only called once per volume per node. ([#112403](https://github.com/kubernetes/kubernetes/pull/112403), [@akankshakumari393](https://github.com/akankshakumari393))...

- For raw block CSI volumes on Kubernetes, kubelet was incorrectly calling CSI NodeStageVolume for every single "map" (i.e. raw block "mount") operation for a volume already attached to the node. This PR ensures it is only called once per volume per node. ([#112403](https://github.com/kubernetes/kubernetes/pull/112403), [@akankshakumari393](https://github.com/akankshakumari393))...

name instead of  "coredns/coredns", thus restoring the behaviour existing before the v1.21 release. Users who rely on nested folder for the coredns image should set the "clusterConfiguration.dns.imageRepository" value including the nested path name (e.g using "registry.company.xyz/coredns" will force kubeadm to use "registry.company.xyz/coredns/coredns" image). No action is needed if using the default registry (k8s.gcr.io). ([#102502](https://github.com/kubernetes/kubernetes/pull/102502), [@ykaka...

func.func @strided_slice_ellipsis_mask(%input: tensor<2x4x8x16x32x64xf32>) {
  // For StridedSlice input[1, ..., 8:, :10, 2:6:2]
  // The ellipsis mask is applied to dim #1, #2, i.e, we get canonicalized
  // slice input[1, :, :, 8:, :10, 2:6:2]

  // The start, limit indices and strides attributes of mhlo.slice would
  // reflect the canonicalized slice.

- For raw block CSI volumes on Kubernetes, kubelet was incorrectly calling CSI NodeStageVolume for every single "map" (i.e. raw block "mount") operation for a volume already attached to the node. This PR ensures it is only called once per volume per node. ([#112403](https://github.com/kubernetes/kubernetes/pull/112403), [@akankshakumari393](https://github.com/akankshakumari393))...

  // CHECK: return %0 : tensor<2xf32>
}

// CHECK-LABEL: ConvertMul1WithBroadcastToIdentity
// If the broadcast doesn't change the dimensions (i.e. constant is smaller than input)
func.func @ConvertMul1WithBroadcastToIdentity(%arg0: tensor<2x2xf32>) -> tensor<2x2xf32> {
  %cst = arith.constant dense<1.0> : tensor<2xf32>

        "properties": {
          "categories": {
            "description": "categories is a list of the grouped resources this resource belongs to (e.g. 'all')",
            "items": {
              "default": "",
              "type": "string"
            },
            "type": "array",
            "x-kubernetes-list-type": "atomic"
          },

	// see https://kubernetes.io/docs/concepts/services-networking/dns-pod-service/#dns-config
	// This does not apply to gateway pods as they typically need a different
	// set of DNS settings than the normal application pods (e.g. in multicluster scenarios).
	PodDNSSearchNamespaces []string `protobuf:"bytes,43,rep,name=podDNSSearchNamespaces,proto3" json:"podDNSSearchNamespaces,omitempty"`

- [`KT-22379`](https://youtrack.jetbrains.com/issue/KT-22379) Condition of while-loop with break can produce unsound smartcast
- [`KT-67021`](https://youtrack.jetbrains.com/issue/KT-67021) K2: Cannot find cached type parameter by FIR symbol: E of the owner: FirRegularClassSymbol Function
- [`KT-67014`](https://youtrack.jetbrains.com/issue/KT-67014) K1/K2 handle when expression as annotation target differently