*         -> attackerSwitch
   * ```
   *
   * The attacker serves a set of certificates that yields a too-long chain in our certificate
   * pinner. The served certificates (incorrectly) formed a single chain to the pinner:
   *
   * ```
   *   attackerCa
   *     -> attackerIntermediate
   *         -> pinnedRoot (trusted by CertificatePinner)

 *  Fix: Don't skip cache invalidation if the invalidating response has no body.
 *  Fix: Don't use a cryptographic random number generator for web sockets. Some Android devices
    implement `SecureRandom` incorrectly!
 *  Fix: Correctly canonicalize IPv6 addresses in `HttpUrl`. This prevented OkHttp from trusting
    HTTPS certificates issued to certain IPv6 addresses.
 *  Fix: Don't reuse connections after an unsuccessful `Expect: 100-continue`.

        }
      } finally {
        // Null out the thread field, so that we don't leak a reference to Thread, and so that
        // future `thread == currentThread()` calls from this thread don't incorrectly queue instead
        // of executing. Don't null out the latestTaskQueue field, because the work done here
        // may have scheduled more operations on another thread, and if those operations then

      // 'actualLastElement' is now at 'lastElementAt', and the element that was at 'lastElementAt'
      // is now at the end of queue. If that's the element we wanted to remove in the first place,
      // don't try to (incorrectly) trickle it. Instead, just delete it and we're done.
      queue[size] = null;
      return null;
    }
    E toTrickle = elementData(size);
    queue[size] = null;

- Kubeadm: Fixed a bug where the node registration information for a given node was not fetched correctly during `kubeadm upgrade node` and the node name can end up being incorrect in cases where the node name is not the same as the host name. ([#134319](https://github.com/kubernetes/kubernetes/pull/134319), [@neolit123](https://github.com/neolit123))

    had a severe bug where web sockets were closed on the wrong thread, which caused
    `NullPointerException` crashes in `Deflater`.

 *  Fix: Don't crash after a web socket fails its connection upgrade. We incorrectly released
    the web socket's connections back to the pool before their resources were cleaned up.

 *  Fix: Don't infinite loop when a received web socket message has self-terminating compressed
    data.

## Changelog since v1.27.5

## Changes by Kind

### API Change

- Fixed a bug where CEL expressions in CRD validation rules would incorrectly compute a high estimated cost for functions that return strings, lists or maps.

- Fixed a bug where CEL expressions in CRD validation rules would incorrectly compute a high estimated cost for functions that return strings, lists or maps.

     * ubiquitous protocols. Applications should only call this method to avoid specific
     * compatibility problems, such as web servers that behave incorrectly when HTTP/2 is enabled.
     *
     * The following protocols are currently supported:
     *
     * * [http/1.1][rfc_2616]
     * * [h2][rfc_7540]
     * * [h2 with prior knowledge(cleartext only)][rfc_7540_34]

- 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),...