* attachdetach controller attaches volumes immediately when Pod's PVCs are bound ([#66863](https://github.com/kubernetes/kubernetes/pull/66863), [@cofyc](https://github.com/cofyc))
* This fix prevents a GCE PD volume from being mounted if the udev device link is stale and tries to correct the link. ([#66832](https://github.com/kubernetes/kubernetes/pull/66832), [@msau42](https://github.com/msau42))

- Make sure auto-mounted subpath mount source is already mounted ([#112499](https://github.com/kubernetes/kubernetes/pull/112499), [@andyzhangx](https://github.com/andyzhangx)) [SIG Storage]

## Dependencies

### Added
_Nothing has changed._

### Changed

pkg syscall (freebsd-386-cgo), func Umask(int) int
pkg syscall (freebsd-386-cgo), func Undelete(string) error
pkg syscall (freebsd-386-cgo), func UnixRights(...int) []uint8
pkg syscall (freebsd-386-cgo), func Unmount(string, int) error
pkg syscall (freebsd-386-cgo), func Wait4(int, *WaitStatus, int, *Rusage) (int, error)
pkg syscall (freebsd-386-cgo), func Write(int, []uint8) (int, error)

- Kubelet: fix data race in volume manager's WaitForAllPodsUnmount that could cause errors to be lost during concurrent pod unmount operations. ([#135794](https://github.com/kubernetes/kubernetes/pull/135794), [@AutuSnow](https://github.com/AutuSnow)) [SIG Node and Storage]

* Fix space issue in volumePath with vSphere Cloud Provider ([#38338](https://github.com/kubernetes/kubernetes/pull/38338), [@BaluDontu](https://github.com/BaluDontu))
* Fix issue when attempting to unmount a wrong vSphere volume ([#37413](https://github.com/kubernetes/kubernetes/pull/37413), [@BaluDontu](https://github.com/BaluDontu))

- Fix kubelet restart unmounts volumes of running pods if the referenced PVC is being deleted by the user ([#130335](https://github.com/kubernetes/kubernetes/pull/130335), [@carlory](https://github.com/carlory)) [SIG Node, Storage and Testing]

- Stopped removing the CSI JSON file if the volume was already mounted during subsequent errors. ([#131311](https://github.com/kubernetes/kubernetes/pull/131311), [@gnufied](https://github.com/gnufied))

pkg syscall (darwin-386), func Umask(int) int
pkg syscall (darwin-386), func Undelete(string) error
pkg syscall (darwin-386), func UnixRights(...int) []uint8
pkg syscall (darwin-386), func Unmount(string, int) error
pkg syscall (darwin-386), func Wait4(int, *WaitStatus, int, *Rusage) (int, error)
pkg syscall (darwin-386), func Write(int, []uint8) (int, error)
pkg syscall (darwin-386), method (*Cmsghdr) SetLen(int)

    * Add optional `use_unbounded_threadpool` argument to `map`, to specify that the `map` should use an unbounded threadpool instead of the default pool that is based on the number of cores on the machine. This can improve throughput for map functions which perform IO or otherwise release the CPU.

- Fix kubelet restart unmounts volumes of running pods if the referenced PVC is being deleted by the user ([#130685](https://github.com/kubernetes/kubernetes/pull/130685), [@carlory](https://github.com/carlory)) [SIG Node, Storage and Testing]

## Dependencies