d.getManagementKey(),
                            "must specify an absolute path but is " + systemPath,
                            d);
                } else if (!sysFile.isFile()) {
                    String msg = "refers to a non-existing file " + sysFile.getAbsolutePath();
                    systemPath = systemPath.replace('/', File.separatorChar).replace('\\', File.separatorChar);

    ReferenceQueue<V> getValueReferenceQueueForTesting() {
      throw new AssertionError();
    }

    /** Unsafely extracts the weak value reference inside of the given {@code entry}. */
    WeakValueReference<K, V, E> getWeakValueReferenceForTesting(InternalEntry<K, V, ?> entry) {
      throw new AssertionError();
    }

    /**

# See https://developer.nvidia.com/cuda-gpus#compute
# `compute_XY` enables PTX embedding in addition to SASS. PTX
# is forward compatible beyond the current compute capability major
# release while SASS is only forward compatible inside the current
# major release. Example: sm_80 kernels can run on sm_89 GPUs but
# not on sm_90 GPUs. compute_80 kernels though can also run on sm_90 GPUs.

If you used to rely on this behavior, now you should create the resources for background tasks inside the background task itself, and use internally only data that doesn't depend on the resources of dependencies with `yield`.

	// ReplicateMRF - audit trail for replication from Most Recent Failures (MRF) queue
	ReplicateMRF = "replicate:mrf"
	// ReplicateIncoming - audit trail of inline replication
	ReplicateIncoming = "replicate:incoming"
	// ReplicateIncomingDelete - audit trail of inline replication of deletes.
	ReplicateIncomingDelete = "replicate:incoming:delete"

	// ReplicateHeal - audit trail for healing of failed/pending replications

			ChangeTime: time.Unix(1389782956, 794414986),
			Xattrs: map[string]string{
				"user.key":  "value",
				"user.key2": "value2",
				// Interestingly, selinux encodes the terminating null inside the xattr
				"security.selinux": "unconfined_u:object_r:default_t:s0\x00",
			},
			PAXRecords: map[string]string{
				"mtime":                         "1386065770.44825232",

    TF_OperationDescription* desc =
        TF_NewOperation(host_graph_, func_name_, func_node_name_);
    for (auto input : inputs) {
      TF_AddInput(desc, input);
    }
    // Set device to CPU because some ops inside the function might not be
    // available on GPU.
    TF_SetDevice(desc, "/cpu:0");
    *op = TF_FinishOperation(desc, s_);
    ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
    ASSERT_NE(*op, nullptr);
  }

$(button).toggleClass(CLASS_NAME_FOCUS, /^focus(in)?$/.test(event.type))\n  })\n\n$(window).on(EVENT_LOAD_DATA_API, () => {\n  // ensure correct active class is set to match the controls' actual values/states\n\n  // find all checkboxes/readio buttons inside data-toggle groups\n  let buttons = [].slice.call(document.querySelectorAll(SELECTOR_DATA_TOGGLES_BUTTONS))\n  for (let i = 0, len = buttons.length; i < len; i++) {\n    const button = buttons[i]\n    const input = button.querySelector(SELECTOR_INPUT)\n...

interacting with the system. Note: the GUI is enabled by default in 1.2 clusters.

<img src="docs/images/newgui.png" width="" alt="Dashboard UI screenshot showing cards that represent applications that run inside a cluster" title="Dashboard UI apps screen">

### Other notable improvements

  * Job was Beta in 1.1 and is GA in 1.2 .

- Mounts `/home/kubernetes/bin/nvidia/vulkan/icd.d` on the host to `/etc/vulkan/icd.d` inside containers requesting GPU. ([#78868](https://github.com/kubernetes/kubernetes/pull/78868), [@chardch](https://github.com/chardch))