TF_GUARDED_BY(mu);
  bool delete_requested TF_GUARDED_BY(mu);  // set true by TF_DeleteGraph

  // Used to link graphs contained in TF_WhileParams to the parent graph that
  // will eventually contain the full while loop.
  TF_Graph* parent;
  TF_Output* parent_inputs;
};

struct TF_OperationDescription {
  TF_OperationDescription(TF_Graph* g, const char* op_type,
                          const char* node_name)

    if (new_node == nullptr) continue;
    parent->name_map[new_node->name()] = new_node;
  }

  // Populate 'outputs'.
  DCHECK_LE(loop_outputs.size(), num_loop_vars);
  for (int i = 0; i < loop_outputs.size(); ++i) {
    outputs[i] = {ToOperation(loop_outputs[i].node()), loop_outputs[i].index()};
  }
}

}  // namespace
#endif  // !defined(IS_MOBILE_PLATFORM) && !defined(IS_SLIM_BUILD)

  VerifyFDefOutputs(fdef, M({{"test_loop_exit"}}));
  VerifyFDefEdges(fdef,
                  {{"feed1", "test_loop/Enter:0"},
                   {"test_loop/Enter:output:0", "test_loop/Merge:0"},
                   {"test_loop/Merge:output:0", "test_loop/Switch:0"},
                   {"test_loop/Switch:output_false:0", "test_loop/Exit:0"},
                   {"test_loop/Exit:output:0", "test_loop_exit"}},
                  {}, false);
}

  """
  var = environ_cp.get(var_name)
  if not var:
    var = get_input(ask_for_var)
    print('\n')
  if not var:
    var = var_default
  return var


def prompt_loop_or_load_from_env(environ_cp,
                                 var_name,
                                 var_default,
                                 ask_for_var,
                                 check_success,

  TensorHandlePtr combined_value = CreatePerDeviceValues(
      context.get(), in_components, device_name, status.get());
  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());

  // Loop to make synchronization failures more deterministic
  for (int i = 0; i < 100; ++i) {
    TensorHandlePtr multiply_result(
        Multiply(context.get(), combined_value.get(), combined_value.get(),

        "//tensorflow/core:protos_all_cc",
        "//tensorflow/core:test",
        "//tensorflow/core:test_main",
    ],
)

tf_cc_test(
    name = "while_loop_test",
    size = "medium",
    srcs = ["while_loop_test.cc"],
    deps = [
        ":c_api",
        ":c_test_util",
        "//tensorflow/core:lib",
        "//tensorflow/core:test",
        "//tensorflow/core:test_main",

  TF_Output cond_output;

  // The loop body graph. The inputs are the current values of the loop
  // variables. The outputs are the updated values of the loop variables.
  TF_Graph* const body_graph;
  const TF_Output* const body_inputs;
  TF_Output* const body_outputs;

  // Unique null-terminated name for this while loop. This is used as a prefix
  // for created operations.
  const char* name;
} TF_WhileParams;

                                                        TF_Status* status);

// Some TF ops need a step container to be set to limit the lifetime of some
// resources (mostly TensorArray and Stack, used in while loop gradients in
// graph mode). Calling this on a context tells it to start a step.
TF_CAPI_EXPORT extern void TFE_ContextStartStep(TFE_Context* ctx);

// Ends a step. When there is no active step (that is, every started step has

## Known Caveats

*   TF Core:
    *   A longstanding bug in `tf.while_loop`, which caused it to execute
        sequentially, even when `parallel_iterations>1`, has now been fixed.
        However, the increased parallelism may result in increased memory use.
        Users who experience unwanted regressions should reset their
        `while_loop`'s `parallel_iterations` value to 1, which is consistent
        with prior behavior.

// upstream gradients.
//
// TODO(apassos) provide concrete template instantiations for TFE_TensorHandle
// specialization, which is blocked by quite a few things needing to loop back
// into python now.
template <typename Gradient, typename BackwardFunction, typename TapeTensor>
class VSpace {
 public:
  virtual ~VSpace() {}

  // Returns the number of elements in the gradient tensor.