* Fixes [#49418](https://github.com/kubernetes/kubernetes/pull/49418) where kube-controller-manager can panic on volume.CanSupport methods and enter a crash loop. ([#49420](https://github.com/kubernetes/kubernetes/pull/49420), [@gnufied](https://github.com/gnufied))

## Changelog since v1.8.9

### Other notable changes

* gce: fixes race condition in ServiceController, where nodes weren't updated in the node sync loop, by updating TargetPools in the ensureExternalLoadBalancer call. ([#58368](https://github.com/kubernetes/kubernetes/pull/58368), [@MrHohn](https://github.com/MrHohn))

    * The vSphere cloud provider has a cleanup function with a loop,
    * inside of which a write lock was intended to prevent races against
    * VM creation, but the usage of the sync.RWMutex type is incorrect,
    * plus is placed in a defer call so the locking is effectively
    * nonexistent.  This patch adds an anonymous function call inside
    * the loop to correctly handle acquiring the write lock with deferred

- Fixed 'kubeadm upgrade' infinite loop waiting for pod restart ([#69886](https://github.com/kubernetes/kubernetes/pull/69886), [@bart0sh](https://github.com/bart0sh))

    * The vSphere cloud provider has a cleanup function with a loop,
    * inside of which a write lock was intended to prevent races against
    * VM creation, but the usage of the sync.RWMutex type is incorrect,
    * plus is placed in a defer call so the locking is effectively
    * nonexistent.  This patch adds an anonymous function call inside
    * the loop to correctly handle acquiring the write lock with deferred

source is CSI type or it has migrated annotation, when it's deleted, the PersisentVolume controller won't changes its phase to the Failed state. 
  
  With this patch, the external provisioner can remove the finalizer in next reconcile loop. Unfortunately if the provious existing pv has the Failed state, this patch won't take effort. It requires users to remove finalizer. ([#126043](https://github.com/kubernetes/kubernetes/pull/126043), [@carlory](https://github.com/carlory)) [SIG Apps...

- Fixed bug in CPUManager with race on container map access ([#97427](https://github.com/kubernetes/kubernetes/pull/97427), [@klueska](https://github.com/klueska)) [SIG Node]

* change default azure file mount permission to 0777 ([#69854](https://github.com/kubernetes/kubernetes/pull/69854), [@andyzhangx](https://github.com/andyzhangx))
* Fix a bug in the scheduler that could cause the scheduler to go to an infinite loop when all nodes in a zone are removed. ([#69758](https://github.com/kubernetes/kubernetes/pull/69758), [@bsalamat](https://github.com/bsalamat))

- Fixed a regression since 1.24 in the scheduling framework when overriding MultiPoint plugins (e.g. default plugins).

volume source is CSI type or it has migrated annotation, when it's deleted, the PersisentVolume controller won't changes its phase to the Failed state. 
  
  With this patch, the external provisioner can remove the finalizer in next reconcile loop. Unfortunately if the provious existing pv has the Failed state, this patch won't take effort. It requires users to remove finalizer. ([#126045](https://github.com/kubernetes/kubernetes/pull/126045), [@carlory](https://github.com/carlory)) [SIG Apps...