// Node.  A ClusterCIDR is eligible to be used for a given Node when the node
// selector matches the node in question and has free CIDRs to allocate.  In
// case of multiple matching ClusterCIDR resources, the allocator will attempt
// to break ties using internal heuristics, but any ClusterCIDR whose node
// selector matches the Node may be used.
message ClusterCIDR {
  // Standard object's metadata.

  TF_DeleteStatus(s);
}

void Deallocator(void* data, size_t, void* arg) {
  tensorflow::cpu_allocator()->DeallocateRaw(data);
  *reinterpret_cast<bool*>(arg) = true;
}

TEST(CAPI, Tensor) {
  const int num_bytes = 6 * sizeof(float);
  float* values =
      reinterpret_cast<float*>(tensorflow::cpu_allocator()->AllocateRaw(
          TF_TensorDefaultAlignment(), num_bytes));
  int64_t dims[] = {2, 3};

- Migrated `pkg/controller/nodeipam/ipam/cloud_cidr_allocator.go, pkg/controller/nodeipam/ipam/multi_cidr_range_allocator.go pkg/controller/nodeipam/ipam/range_allocator.go pkg/controller/nodelifecycle/node_lifecycle_controller.go` to structured logging ([#112670](https://github.com/kubernetes/kubernetes/pull/112670), [@yangjunmyfm192085](https://github.com/yangjunmyfm192085))...

                "//tensorflow/lite/kernels/shim:tf_kernel_shim",
                "@local_xla//xla/stream_executor:stream_executor_impl",
                "@local_tsl//tsl/framework:bfc_allocator",
                "@local_tsl//tsl/framework:metrics",
            ] + tf_additional_binary_deps() +
            # TODO(b/259305727): Remove this select and include captured_function in macos builds.
            select({

- Removed the generally available feature gate `ProxyTerminatingEndpoints`. ([#122134](https://github.com/kubernetes/kubernetes/pull/122134), [@ty-dc](https://github.com/ty-dc))

    *   Promote `tf.data.experimental.sample_from_datasets` API to
        `tf.data.Dataset.sample_from_datasets` and deprecate the experimental
        endpoint.
    *   Added `TF_GPU_ALLOCATOR=cuda_malloc_async` that use cudaMallocAsync from
        CUDA 11.2. This could become the default in the future.
*   TF SavedModel:
    *   Custom gradients are now saved by default. See