}

Status ImmediateExecutionTensorHandle::SummarizeValue(
    std::string& summary) const {
  Status status;
  AbstractTensorPtr resolved(
      // TODO(allenl): Resolve should be const, and the caches that get updated
      // marked mutable.
      const_cast<ImmediateExecutionTensorHandle*>(this)->Resolve(&status));
  if (!status.ok()) {
    return status;
  }
  summary = resolved->SummarizeValue();
  return absl::OkStatus();

      "ApiDefMap is not supported on mobile.");
  return nullptr;
#else
  mutex_lock l(api_def_map->lock);
  if (!api_def_map->update_docs_called) {
    api_def_map->api_def_map.UpdateDocs();
    api_def_map->update_docs_called = true;
  }
  string name_str(name, name_len);
  const auto* api_def = api_def_map->api_def_map.GetApiDef(name_str);
  if (api_def == nullptr) {

  worker_server.release();

  // Update the server def with a new set of names (worker instead of
  // localhost).
  tensorflow::ServerDef updated_server_def = GetServerDef("worker", 2);
  serialized = updated_server_def.SerializeAsString();

  updated_server_def.set_task_index(1);
  tensorflow::Status s = tensorflow::GrpcServer::Create(
      updated_server_def, tensorflow::Env::Default(), &worker_server);

      output_gradients, result, /*build_default_zeros_grads*/ false));
  return absl::OkStatus();
}

// Helper functions which delegate to `AbstractOperation`, update
// the state of the ForwardOperation and call the tape as appropriate.
// These APIs are mainly to facilitate testing and are subject to change.
namespace internal {
Status Reset(AbstractOperation* op_, const char* op,

                         tensorflow::FunctionLibraryDefinition* flib_def,
                         std::vector<std::string>* control_ret_node_names,
                         bool* control_rets_updated) override {
    // Inject failure to function instantiation if finding a node that contains
    // the given node name (error_node_) and requested device (error_device_).
    for (const auto node : graph->get()->nodes()) {

  ASSERT_TRUE(worker_server1->Start().ok());

  // 5a. Update `ctx_0` with updated `server_def_0`.
  {
    server_def_0.set_task_index(0);
    string serialized_update = server_def_0.SerializeAsString();
    TF_Status* status = TF_NewStatus();
    TFE_ContextUpdateServerDefWithTimeout(ctx_0, 0, serialized_update.data(),
                                          serialized_update.size(),

      if (!status->status.ok()) return;
      input_tensors_vector[i] = &input_tensor;
    }
  }

  // Create an inference context with dummy values, which will be updated later.
  InferenceContext c(TF_GRAPH_DEF_VERSION, node_def, op_reg_data->op_def,
                     std::vector<ShapeHandle>(num_inputs), input_tensors_vector,
                     {},

  ASSERT_TRUE(worker_server1->Start().ok());

  // 5a. Update `ctx_0` with updated `server_def_0`.
  {
    server_def_0.set_task_index(0);
    string serialized_update = server_def_0.SerializeAsString();
    TF_Status* status = TF_NewStatus();
    TFE_ContextUpdateServerDef(ctx_0, 0, serialized_update.data(),
                               serialized_update.size(), status);

    } else {
      // Remove the stale block and continue.
      RemoveFile_Locked(key.first);
    }
  }

  // Insert a new empty block, setting the bookkeeping to sentinel values
  // in order to update them as appropriate.
  auto new_entry = std::make_shared<Block>();
  lru_list_.push_front(key);
  lra_list_.push_front(key);
  new_entry->lru_iterator = lru_list_.begin();
  new_entry->lra_iterator = lra_list_.begin();

  tf_gcs_filesystem::ExpiringLRUCache<int> cache(
      1, 0, [&env]() { return env->NowSeconds(); });
  env->SetNowSeconds(1);
  // Verify that replacement of an existing element works, and updates the
  // timestamp of the entry.
  cache.Insert(key, 41);
  env->SetNowSeconds(2);
  cache.Insert(key, 42);
  // 1 second after the most recent insertion, the entry is still valid.
  env->SetNowSeconds(3);