// 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);

**Standalone code to reproduce the issue** 
Provide a reproducible test case that is the bare minimum necessary to generate
the problem. If possible, please share a link to Colab/Jupyter/any notebook.

Also, please include a link to a GraphDef or the model if possible.

**Any other info / logs**

Include any logs or source code that would be helpful to diagnose the problem.
If including tracebacks, please include the full traceback. Large logs and files

    const tensorflow::GraphDef& graph_def) {
  std::vector<std::pair<string, string>> grads;
  for (const tensorflow::GradientDef& grad : graph_def.library().gradient()) {
    grads.emplace_back(grad.function_name(), grad.gradient_func());
  }
  std::sort(grads.begin(), grads.end());
  return grads;
}

std::vector<string> GetFuncNames(const tensorflow::GraphDef& graph_def) {
  std::vector<string> names;

  delete s;
}

void TF_ExtendGraph(TF_DeprecatedSession* s, const void* proto,
                    size_t proto_len, TF_Status* status) {
  GraphDef g;
  if (!tensorflow::ParseProtoUnlimited(&g, proto, proto_len)) {
    status->status = InvalidArgument("Invalid GraphDef");
    return;
  }
  status->status = s->session->Extend(g);
}

}  // end extern "C"

            TF_OperationOutputConsumers(TF_Output{three, 0}, &three_port, 1));
  EXPECT_EQ(add, three_port.oper);
  EXPECT_EQ(1, three_port.index);

  // Serialize to GraphDef.
  GraphDef graph_def;
  ASSERT_TRUE(GetGraphDef(graph, &graph_def));

  // Validate GraphDef is what we expect.
  bool found_placeholder = false;
  bool found_scalar_const = false;
  bool found_add = false;
  for (const auto& n : graph_def.node()) {

    const char* text_proto,
    std::function<void(FunctionDef*)>* mutate_proto_func, TF_Status* status) {
  tensorflow::GraphDef gdef;
  if (!tensorflow::protobuf::TextFormat::ParseFromString(text_proto, &gdef)) {
    status->status = tensorflow::errors::Internal(
        "Invalid text proto for GraphDef: ", text_proto);
    return {};
  }
  const auto& fdef_lib = gdef.library();
  if (fdef_lib.gradient_size() > 0) {

    TF_LOCKS_EXCLUDED(session->graph->mu, session->mu);

std::string getTF_OutputDebugString(TF_Output node);

// Set whether to propagate assigned device information when constructing a new
// Graph from a GraphDef. By default assigned device information is not copied
// and is re-computed by the runtime.
inline void TF_ImportGraphDefOptionsSetPropagateDeviceSpec(
    TF_ImportGraphDefOptions* opts, unsigned char propagate_device_spec) {

  TF_GraphCopyFunction(host_graph_, func_, grad_func, s_);
  ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);

  // 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]);

                                  TF_Status* status);

// Has the same behavior as TF_GraphImportGraphDefWithResults, but instead of
// taking in a serialized tensorflow::GraphDef, it takes in a *pointer* to the
// C++ *in memory representation* of the GraphDef, stored in `graph_def->data`
TF_CAPI_EXPORT extern TF_ImportGraphDefResults*
TF_GraphImportGraphDefWithResultsNoSerialization(
    TF_Graph* graph, const TF_Buffer* graph_def,

should be used with caution when working with untrusted models.

### Saved graphs and checkpoints

When loading untrusted serialized computation graphs (in form of a `GraphDef`,
`SavedModel`, or equivalent on-disk format), the set of computation primitives
available to TensorFlow is powerful enough that you should assume that the
TensorFlow process effectively executes arbitrary code.