: forward_outputs_(f_outputs) {
    for (auto output : forward_outputs_) {
      if (output) {
        output->Ref();
      }
    }
  }

  Status Compute(AbstractContext* ctx,
                 absl::Span<AbstractTensorHandle* const> grad_outputs,
                 absl::Span<AbstractTensorHandle*> grad_inputs) override {
    AbstractTensorHandle* upstream_grad = grad_outputs[0];

  EXPECT_EQ(num_compute_calls, 5);
  cache2.LookupOrCompute("c", &value, compute_func, &status);
  TF_EXPECT_OK(status.status);
  EXPECT_EQ(value, 2);
  EXPECT_EQ(num_compute_calls, 5);
  cache2.LookupOrCompute("d", &value, compute_func, &status);
  TF_EXPECT_OK(status.status);
  EXPECT_EQ(value, 3);
  EXPECT_EQ(num_compute_calls, 5);

  // Re-read "a", ensure it is re-computed.
  cache2.LookupOrCompute("a", &value, compute_func, &status);

                 TF_Message(first_bad_status.get()));
    return result;
  }
  // For each output of the original operation, pack the per-device
  // TensorHandles we've computed into a single parallel TensorHandle.
  std::vector<std::unique_ptr<ParallelTensor>> per_device_outputs;
  per_device_outputs.reserve(first_op_output_count);
  for (int i = 0; i < first_op_output_count; ++i) {

    grad_inputs[0] = grad_outputs[0];
    if (grad_inputs[0]) {
      grad_inputs[0]->Ref();
    }
    return absl::OkStatus();
  }
};

// Computes:
//
// @tf.custom_gradient
// def f(input):
//   def grad(grads):
//     return grads[0]
//   return tf.exp(input), grad
// outputs = [f(inputs[0])]
Status ExpWithPassThroughGrad(AbstractContext* ctx,

#include "tensorflow/c/eager/abstract_context.h"

namespace tensorflow {
namespace gradients {
namespace {
class IdentityNGradientFunction : public GradientFunction {
 public:
  Status Compute(AbstractContext* ctx,
                 absl::Span<AbstractTensorHandle* const> grad_outputs,
                 absl::Span<AbstractTensorHandle*> grad_inputs) override {
    for (int i = 0; i < grad_outputs.size(); i++) {

      control_output_names_vec.push_back(string(control_output_names[i]));
    }
  }

  // Compute body nodes.
  std::vector<const Node*> body_nodes;
  status->status = tensorflow::ComputeBodyNodes(
      fn_body, fn_name, num_opers, opers, input_nodes, &body_nodes);
  if (TF_GetCode(status) != TF_OK) return nullptr;

  // Compute body nodes.
  std::vector<const Node*> control_output_nodes;

// Copyright 2013 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.

// This .cc file will be automatically compiled by the go tool and
// included in the package.

#include <string>
#include "main.h"

std::string Caller::call() {
	if (callback_ != 0)
		return callback_->run();
	return "";

  TF_DeleteBuffer(op_list_buf);
}

class DummyKernel : public tensorflow::OpKernel {
 public:
  explicit DummyKernel(tensorflow::OpKernelConstruction* context)
      : OpKernel(context) {}
  void Compute(tensorflow::OpKernelContext* context) override {}
};

// Test we can query kernels
REGISTER_OP("TestOpWithSingleKernel")
    .Input("a: float")
    .Input("b: float")
    .Output("o: float");

class TestUnavailableErrorOp : public tensorflow::OpKernel {
 public:
  explicit TestUnavailableErrorOp(tensorflow::OpKernelConstruction* ctx)
      : tensorflow::OpKernel(ctx) {}
  void Compute(tensorflow::OpKernelContext* ctx) override {
    ctx->SetStatus(tensorflow::errors::Unavailable("Test error."));
  }
};
REGISTER_KERNEL_BUILDER(
    Name("TestNonCommUnavailable").Device(tensorflow::DEVICE_DEFAULT),

==============================================================================*/
#include "tensorflow/c/experimental/gradients/not_differentiable.h"

namespace tensorflow {
namespace gradients {
Status NotDifferentiableGradientFunction::Compute(
    AbstractContext* ctx, absl::Span<AbstractTensorHandle* const> grad_outputs,
    absl::Span<AbstractTensorHandle*> grad_inputs) {
  for (int i = 0; i < grad_inputs.size(); i++) {
    grad_inputs[i] = nullptr;
  }