limitations under the License.
==============================================================================*/

#ifndef TENSORFLOW_C_EAGER_CUSTOM_DEVICE_TESTUTIL_H_
#define TENSORFLOW_C_EAGER_CUSTOM_DEVICE_TESTUTIL_H_

// A simple logging device to test custom device registration.
#include <memory>

#include "tensorflow/c/c_api.h"
#include "tensorflow/c/eager/c_api.h"
#include "tensorflow/c/eager/c_api_experimental.h"

#include "tensorflow/c/c_api.h"
#include "tensorflow/c/experimental/grappler/grappler.h"
#include "tensorflow/core/framework/graph.pb.h"
#include "tensorflow/core/grappler/optimizers/custom_graph_optimizer.h"
#include "tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.h"
#include "tensorflow/core/platform/status.h"
#include "tensorflow/core/protobuf/rewriter_config.pb.h"

namespace tensorflow {
namespace grappler {

-   type: input id: tfversion attributes: label: Tensorflow Version description:
    placeholder: ex,. tf 2.8 validations: required: true

-   type: dropdown id: Code attributes: label: Custom Code description:
    options: - "Yes" - "No" validations: required: true

-   type: input id: OS attributes: label: OS Platform and Distribution
    description: placeholder: e.g., Linux Ubuntu 16.04 validations: required:

  TF_DeleteStatus(status);
  TFE_DeleteOp(var_op);
  TFE_DeleteOp(var_op_2);
  TFE_DeleteContext(ctx);
}

// Needs to work with a const TFE_Op since custom devices should not modify the
// op they are called with.
TFE_Op* CloneOp(const TFE_Op* other) {
  TF_Status* status = TF_NewStatus();
  TFE_Context* context = TFE_OpGetContext(other, status);

  ASSERT_TRUE(TF_GetCode(status.get()) == TF_OK) << TF_Message(status.get());

  const char* custom0 = "/job:localhost/replica:0/task:0/device:CUSTOM:0";
  const char* custom1 = "/job:localhost/replica:0/task:0/device:CUSTOM:1";
  bool arrived = false;
  bool executed = false;
  RegisterLoggingDevice(context.get(), custom0,
                        /*strict_scope_placement=*/false, &arrived, &executed,

    CHECK_EQ(grad_inputs.size(), 1);
    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,

        ":not_differentiable",
        "//tensorflow/c/eager:abstract_context",
        "//tensorflow/c/eager:gradients_internal",
    ],
)

tf_cuda_cc_test(
    name = "custom_gradient_test",
    size = "small",
    srcs = [
        "custom_gradient_test.cc",
    ],
    args = ["--heap_check="],  # TODO(b/174752220): Remove
    tags = tf_cuda_tests_tags(),
    deps = [
        "//tensorflow/c:tf_status_helper",

computation, shared state, and the operations that mutate that state. It maps the nodes of a dataflow graph across many machines in a cluster, and within a machine across multiple computational devices, including multicore CPUs, general purpose GPUs, and custom-designed ASICs known as Tensor Processing Units (TPUs). This architecture gives flexibility to the application developer, whereas in previous “parameter server” designs the management of shared state is built into the system, TensorFlow enables developers...

# setup.python.sh: Install a specific Python version and packages for it.
# Usage: setup.python.sh <pyversion> <requirements.txt>

# Sets up custom apt sources for our TF images.

# Prevent apt install tzinfo from asking our location (assumes UTC)
export DEBIAN_FRONTEND=noninteractive

# Set up shared custom sources
apt-get update
apt-get install -y gnupg ca-certificates

# Deadsnakes: https://launchpad.net/~deadsnakes/+archive/ubuntu/ppa

  // be an empty string if no device was explicitly requested, but will
  // otherwise be the name of this custom device. Ops are placed onto a custom
  // device if any of their inputs are on that custom device, but custom devices
  // are free to set a bad status in order to require explicit placement.
  void (*execute)(const TFE_Op* op, int* num_outputs,