std::vector<string> input_names(ninputs);
  std::vector<string> output_names(noutputs);
  std::vector<string> target_oper_names(ntargets);
  for (int i = 0; i < ninputs; ++i) {
    input_names[i] = c_input_names[i];
  }
  for (int i = 0; i < noutputs; ++i) {
    output_names[i] = c_output_names[i];
  }
  for (int i = 0; i < ntargets; ++i) {
    target_oper_names[i] = c_target_oper_names[i];
  }

    data = [
        ":test_op1.so",
        "//tensorflow/cc/saved_model:saved_model_half_plus_two",
    ],
    linkopts = select({
        "//tensorflow:macos": ["-headerpad_max_install_names"],
        "//conditions:default": [],
    }),
    tags = [
        "no_cuda_asan",  # TODO(b/181771536)
        "no_windows",  # TODO(b/155444728)
    ],

// device_info is an opaque caller-defined type stored with the custom device
// which is passed to the functions referenced in the TFE_CustomDevice struct
// `device` (execute, delete_device, etc.). It can for example contain the
// names of wrapped devices.
//
// There are currently no graph semantics implemented for registered custom
// devices, so executing tf.functions which contain operations placed on the
// custom devices will fail.
//

  // Verify that function and its grad are in host graph's GraphDef
  GraphDef gdef;
  GetGraphDef(host_graph_, &gdef);
  std::vector<string> func_names = GetFuncNames(gdef);
  ASSERT_EQ(2, func_names.size());
  ASSERT_EQ(func_name_, func_names[0]);
  ASSERT_EQ("MyGrad", func_names[1]);
  std::vector<std::pair<string, string>> grads = GetGradDefs(gdef);
  ASSERT_EQ(1, grads.size());
  ASSERT_EQ(func_name_, grads[0].first);

  const std::vector<std::string>& op_names =
      tensorflow::unwrap(ctx)->GetLoggedOpsTestonly();

  std::ostringstream op_names_oss;
  for (const auto& op : op_names) {
    op_names_oss << op << ", ";
  }
  const std::string& op_names_str = op_names_oss.str();
  void* data = tensorflow::port::Malloc(op_names_str.length());
  op_names_str.copy(static_cast<char*>(data), op_names_str.length(), 0);
  buf->data = data;

# This is the official list of TensorFlow authors for copyright purposes.
# This file is distinct from the CONTRIBUTORS files.
# See the latter for an explanation.

# Names should be added to this file as:
# Name or Organization <email address>
# The email address is not required for organizations.

Google Inc.
Yuan Tang <******@****.***>

      the terms of any separate license agreement you may have executed
      with Licensor regarding such Contributions.

   6. Trademarks. This License does not grant permission to use the trade
      names, trademarks, service marks, or product names of the Licensor,
      except as required for reasonable and customary use in describing the
      origin of the Work and reproducing the content of the NOTICE file.

  TF_Status* status = TF_NewStatus();
  std::vector<std::string> device_names;
  TF_DeviceList* devices = TFE_ContextListDevices(ctx, status);
  EXPECT_EQ(TF_GetCode(status), TF_OK) << TF_Message(status);
  const int num_devices = TF_DeviceListCount(devices);
  for (int i = 0; i < num_devices; ++i) {
    device_names.emplace_back(TF_DeviceListName(devices, i, status));

// graph_def.library().gradient()
std::vector<std::pair<string, string>> GetGradDefs(
    const tensorflow::GraphDef& graph_def);

// Returns a sorted vector of names contained in `grad_def`
std::vector<string> GetFuncNames(const tensorflow::GraphDef& graph_def);

class CSession {
 public:
  CSession(TF_Graph* graph, TF_Status* s, bool use_XLA = false);

/// modular world. Once all TensorFlow filesystems are converted to use the
/// plugin based approach, this file will replace the one in core/platform and
/// the names will lose the `Modular` part. Until that point, the `Modular*`
/// classes here are experimental and subject to breaking changes.
/// For documentation on these methods, consult `core/platform/filesystem.h`.