}
TEST(CAPI, RemoteExecuteSilentCopiesLocalAsyncFuncOrdering) {
  // A remote input may be not ready when we start running a function. Test that
  // the function execution should wait until the remote input is ready.
  TestRemoteExecuteSilentCopiesFunc(/*async=*/true, /*remote=*/false,
                                    /*heavy_load_on_streaming_rpc=*/true);
}

  EXPECT_EQ(cache.CacheSize(), 8);
  TF_EXPECT_OK(ReadCache(&cache, "b", 0, 1, &out));
  EXPECT_EQ(calls, 3);
  // Advance the fake time to `now` + 3, at which point "b" becomes stale.
  env->SetNowSeconds(now + 3);
  // Wait for the pruner to remove "b".
  start = Env::Default()->NowSeconds();
  do {
    Env::Default()->SleepForMicroseconds(100000);
  } while (cache.CacheSize() == 8 && Env::Default()->NowSeconds() - start < 3);

  //
  // 1. Create a variable on `remote_device`, using `ctx_0`.
  TFE_TensorHandle* handle_0 =
      CreateVariable(ctx_0, 1.2, remote_device, /*variable_name=*/"var");

  // 2. Wait for `var` to be created and initialized on the worker.
  TF_Status* status = TF_NewStatus();
  TFE_ContextAsyncWait(ctx_0, status);
  EXPECT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
  TF_DeleteStatus(status);

  // 1. Create a variable on `remote_device`, using `ctx_0`.
  TFE_TensorHandle* handle_0 =
      CreateVariable(ctx_0, 1.2, remote_device, /*variable_name=*/"var2");

  // 2. Wait for `var2` to be created and initialized on the worker.
  TF_Status* status = TF_NewStatus();
  TFE_ContextAsyncWait(ctx_0, status);
  EXPECT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
  TF_DeleteStatus(status);

  StatusPtr first_bad_status(nullptr);

  for (const auto& dt : device_threads_) {
    StatusPtr async_wait_status(TF_NewStatus());
    dt->AsyncWait(async_wait_status.get());
    // Prefer non cancelled errors to uncover real failures.
    if (TF_GetCode(async_wait_status.get()) != TF_OK &&
        (first_bad_status == nullptr ||
         TF_GetCode(first_bad_status.get()) == TF_CANCELLED)) {

        "--quiet", "--wait", "--nocache", \
        "--add", "Microsoft.VisualStudio.Workload.VCTools", \
        "--add", "Microsoft.VisualStudio.Component.VC.Tools.x86.x64", \
        "--add", "Microsoft.VisualStudio.Component.Windows10SDK.19041" -Wait; \
    Start-Process -FilePath C:/TEMP/vs_buildtools.exe -ArgumentList "--installPath", "C:/VS", \
        "--quiet", "--wait", "--nocache", "--includeOptional", \

  proc = subprocess.Popen(
      [environ_cp['PYTHON_BIN_PATH'], paths[0]] + cuda_libraries,
      stdout=subprocess.PIPE,
      env=maybe_encode_env(environ_cp))

  if proc.wait():
    # Errors from find_cuda_config.py were sent to stderr.
    print('Asking for detailed CUDA configuration...\n')
    return False

  config = dict(

    uint64_t timestamp;
    /// Mutex to guard state variable
    absl::Mutex mu;
    /// The state of the block.
    FetchState state ABSL_GUARDED_BY(mu) = FetchState::CREATED;
    /// Wait on cond_var if state is FETCHING.
    absl::CondVar cond_var;
  };

  /// \brief The block map type for the file block cache.
  ///
  /// The block map is an ordered map from Key to Block.

      (https://www.tensorflow.org/api_docs/python/tf/data/experimental/service)
      to write distributed dataset snapshots. The call is non-blocking and
      returns without waiting for the snapshot to finish. Setting `wait=True` to
      `tf.data.Dataset.load` allows the snapshots to be read while they are
      being written.

### Bug Fixes and Other Changes

* <SIMILAR TO ABOVE SECTION, BUT FOR OTHER IMPORTANT CHANGES / BUG FIXES>