ASSERT_EQ(errors::OK, status_.code()) << status_.message();

    {
      AbstractContext* ctx_raw = nullptr;
      status_ =
          BuildImmediateExecutionContext(std::get<1>(GetParam()), &ctx_raw);
      ASSERT_EQ(errors::OK, status_.code()) << status_.message();
      immediate_execution_ctx_.reset(ctx_raw);
    }

    // Computing numerical gradients with TensorFloat-32 is numerically

  {
    AbstractContext* ctx_raw = nullptr;
    Status s =
        BuildImmediateExecutionContext(std::get<1>(GetParam()), &ctx_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();
    ctx.reset(ctx_raw);
  }

  AbstractTensorHandlePtr x;
  {
    AbstractTensorHandle* x_raw = nullptr;
    Status s = TestScalarTensorHandle<float, TF_FLOAT>(ctx.get(), 2.0f, &x_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();

  AbstractTensorHandlePtr x1;
  {
    AbstractTensorHandle* x1_raw = nullptr;
    status_ = TestScalarTensorHandle<float, TF_FLOAT>(
        immediate_execution_ctx_.get(), 1.0f, &x1_raw);
    ASSERT_EQ(errors::OK, status_.code()) << status_.message();
    x1.reset(x1_raw);
  }

  AbstractTensorHandlePtr x2;
  {
    AbstractTensorHandle* x2_raw = nullptr;
    status_ = TestScalarTensorHandle<float, TF_FLOAT>(

  {
    AbstractContext* ctx_raw = nullptr;
    Status s =
        BuildImmediateExecutionContext(std::get<1>(GetParam()), &ctx_raw);
    ASSERT_EQ(errors::OK, s.code()) << s.message();
    ctx.reset(ctx_raw);
  }

  AbstractTensorHandlePtr x;
  {
    AbstractTensorHandle* x_raw = nullptr;
    Status s = TestScalarTensorHandle<float, TF_FLOAT>(ctx.get(), 1.0f, &x_raw);

    AbstractTensorHandlePtr numerical_grad;
    {
      AbstractTensorHandle* numerical_grad_raw;
      s = CalcNumericalGrad(ctx, model, inputs,
                            /*input_index=*/i, use_function,
                            &numerical_grad_raw);
      ASSERT_EQ(errors::OK, s.code()) << s.message();
      numerical_grad.reset(numerical_grad_raw);
    }

    TF_Tensor* numerical_tensor;

  void Release() override { delete this; }
  Status Reset(const char* op, const char* raw_device_name) override {
    if (op_) {
      return errors::FailedPrecondition("Reset called on already built op.");
    }
    if (raw_device_name) {
      device_name_ = raw_device_name;
    }
    op_type_ = op;
    return absl::OkStatus();
  }

*   Fixes a security vulnerability caused by lack of validation in
    `tf.raw_ops.DataFormatVecPermute` and `tf.raw_ops.DataFormatDimMap`
    ([CVE-2020-26267](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-26267))
*   Fixes a vulnerability caused by attempting to write to immutable memory
    region in `tf.raw_ops.ImmutableConst`
    ([CVE-2020-26268](https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-26268)

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

#ifdef __cplusplus
extern "C" {
#endif

// Resets `op_to_reset` with `op_or_function_name` and `raw_device_name`. This
// is for performance optimization by reusing an exiting unused op rather than
// creating a new op every time. If `raw_device_name` is `NULL` or empty, it
// does not set the device name. If it's not `NULL`, then it attempts to parse

        output->Unref();
      }
      TF_RETURN_IF_ERROR(ctx->RegisterFunction(func));
    }

    AbstractOperationPtr fn_op(ctx->CreateOperation());
    TF_RETURN_IF_ERROR(fn_op->Reset(fn_name, /*raw_device_name=*/nullptr));
    for (auto input : inputs) {
      TF_RETURN_IF_ERROR(fn_op->AddInput(input));
    }
    int retvals = outputs.size() - null_indices.size();

void TFE_OpReset(TFE_Op* op_to_reset, const char* op_or_function_name,
                 const char* raw_device_name, TF_Status* status) {
  if (op_to_reset) {
    tensorflow::ImmediateExecutionOperation* op =
        tensorflow::unwrap(op_to_reset);
    op->Clear();
    status->status = op->Reset(op_or_function_name, raw_device_name);
  } else {
    TF_SetStatus(status, TF_INVALID_ARGUMENT,