* hasNext()} method will return {@code false}.
   *
   * @param iterator the iterator to copy
   * @param type the type of the elements
   * @return a newly-allocated array into which all the elements of the iterator have been copied
   */
  @GwtIncompatible // Array.newInstance(Class, int)
  public static <T extends @Nullable Object> T[] toArray(

- Dynamic resource allocation: when a claim uses "wait for first consumer" allocation (the default), then it will now get deallocated after it was used by a pod. That ensures that the next pod isn't affected by previous scheduling decision and that resources are not kept allocated unless really needed. If keeping a claim allocated is desired, use "immediate allocation." ([#118936](https://github.com/kubernetes/kubernetes/pull/118936), [@pohly](https://github.com/pohly))

- Extended resources requested by initContainers which are allocated using an automatic ResourceClaim now match the behavior of legacy device plugins, reusing the same resources requested by later sidecar initContainers or regular containers when possible, to minimize the total number of devices requested by the pod....

		for _, tt := range compareTests {
			eql := Equal(tt.a, tt.b)
			if eql != (tt.i == 0) {
				t.Errorf(`Equal(%q, %q) = %v`, tt.a, tt.b, eql)
			}
		}
	})
	if allocs > 0 {
		t.Errorf("Equal allocated %v times", allocs)
	}
}

func TestEqualExhaustive(t *testing.T) {
	var size = 128
	if testing.Short() {
		size = 32
	}
	a := make([]byte, size)
	b := make([]byte, size)
	b_init := make([]byte, size)

- Added driver-owned fields in `ResourceClaim.Status` to report device status data for each allocated device. ([#128240](https://github.com/kubernetes/kubernetes/pull/128240), [@LionelJouin](https://github.com/LionelJouin)) [SIG API Machinery, Network, Node and Testing]

  for (int i = 0; i < len; ++i) {
    values[i] = TF_NewBuffer();
    status->status = MessageToBuffer(attr->list().shape(i), values[i]);
    if (!status->status.ok()) {
      // Delete everything allocated to far, the operation has failed.
      for (int j = 0; j <= i; ++j) {
        TF_DeleteBuffer(values[j]);
      }
      return;
    }
  }
}

  }

  fun assertRequests(vararg expectedHosts: String?) {
    assertThat(requestedHosts).containsExactly(*expectedHosts)
    requestedHosts.clear()
  }

  /** Allocates and returns `count` fake IPv4 addresses like [255.0.0.100, 255.0.0.101].  */
  fun allocate(count: Int): List<InetAddress> {
    val from = nextAddress
    nextAddress += count
    return (from until nextAddress)
      .map {

- Dynamic Resource Allocator (DRA): Enhanced validation for the ResourceClaimParametersReference and ResourceClassParametersReference with the following...

- New feature gate, ServiceIPStaticSubrange, to enable the new strategy in the Service IP allocators, so the IP range is subdivided and dynamic allocated ClusterIP addresses for Services are allocated preferently from the upper range. ([#106792](https://github.com/kubernetes/kubernetes/pull/106792), [@aojea](https://github.com/aojea))

- Kube-scheduler: in Kubernetes 1.33, the number of devices that can be allocated per ResourceClaim was accidentally reduced to 16. Now the supported number of devices per ResourceClaim is 32 again. ([#131679](https://github.com/kubernetes/kubernetes/pull/131679), [@mortent](https://github.com/mortent)) [SIG Node]