TensorHandlePtr value_one(FloatTensorHandle(3., status.get()));
  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());
  TensorHandlePtr value_two(FloatTensorHandle(-2., status.get()));
  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());
  std::array<TFE_TensorHandle*, 2> in_components{value_one.get(),

      "important to you");
  return EmptyWhileParams();
#else
  if (ninputs == 0) {
    status->status =
        InvalidArgument("TF_NewWhile() must be passed at least one input");
    return EmptyWhileParams();
  }

  TF_Graph* cond_graph = TF_NewGraph();
  TF_Graph* body_graph = TF_NewGraph();
  cond_graph->parent = g;
  cond_graph->parent_inputs = inputs;

  //   Outputs:
  std::vector<TensorHandlePtr> op_outputs_ TF_GUARDED_BY(execution_mutex_);
  // TF_Status is an incomplete type and so can't be stack allocated. To avoid
  // unnecessary allocations each Execute call, we keep one heap-allocated
  // version for the thread.
  StatusPtr status_ TF_GUARDED_BY(execution_mutex_);

  const std::string device_;
  ExecutorPtr executor_ TF_GUARDED_BY(execution_mutex_);

//
// TODO(allenl): There are some use-cases that are only supported by copying to
// host at the moment (e.g. debug print on a tensor, .numpy(), etc.). We either
// need to return something here or address these use-cases one by one.
TFE_TensorHandle* CopyTensorFromParallelDevice(TFE_Context* context,
                                               TFE_TensorHandle* tensor,
                                               const char* target_device_name,

  TF_Graph* graph = TF_NewGraph();

  // Make two scalar constants.
  TF_Operation* one = ScalarConst(1, graph, s, "one");
  ASSERT_EQ(TF_OK, TF_GetCode(s)) << TF_Message(s);

  TF_Operation* two = ScalarConst(2, graph, s, "two");
  ASSERT_EQ(TF_OK, TF_GetCode(s)) << TF_Message(s);

  // Add oper.
  TF_Operation* add = Add(one, two, graph, s, "add");
  ASSERT_EQ(TF_OK, TF_GetCode(s)) << TF_Message(s);

  const char task1_name[] = "/job:localhost/replica:0/task:1/device:CPU:0";
  const char task2_name[] = "/job:localhost/replica:0/task:2/device:CPU:0";

  // Create one variable per task.
  TFE_TensorHandle* h0 = TestVariable(ctx, 1.0, task1_name);
  TFE_TensorHandle* h1 = TestVariable(ctx, 2.0, task2_name);
  TFE_TensorHandle* h2 = TestVariable(ctx, 3.0, task0_name);

// accessing local paths). This test suite tests exactly one scheme for each
// invocation. By default, we are testing all schemes available but this can be
// restricted by using `--schemes` to specify a set of schemes to test.
//
// Example invocation:
//  bazel test //tensorflow/c/experimental/filesystem:modular_filesystem_test \\
//  --test_arg=--dso=/path/to/one.so --test_arg=--dso=/path/to/another.so \\

  TFE_DeleteContext(ctx);
  EXPECT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);

  const int num_devices = TF_DeviceListCount(devices);
  EXPECT_GE(num_devices, 1) << "At least one CPU device should exist";
  for (int i = 0; i < num_devices; ++i) {
    EXPECT_NE("", TF_DeviceListName(devices, i, status)) << i;
    EXPECT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
  }

  auto core_ops = std::make_unique<T>();
  memset(core_ops.get(), 0, sizeof(T));
  memcpy(core_ops.get(), plugin_ops, copy_size);
  return core_ops;
}

// Registers one filesystem from the plugin.
//
// Must be called only with `index` a valid index in `info->ops`.
static Status RegisterFileSystem(const TF_FilesystemPluginInfo* info,
                                 int index) {

                             params->body_graph, s_, "add1");
    ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
    TF_Operation* one = ScalarConst(1, params->body_graph, s_);
    ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
    TF_Operation* add2 = Add(add1, one, params->body_graph, s_, "add2");
    ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
    params->body_outputs[0] = {add2, 0};