close(src_fd);
  close(dst_fd);
  if (error_code != 0) {
    errno = error_code;
    return -1;
  } else {
    return 0;
  }
}

int RemoveSpecialDirectoryEntries(const struct dirent* entry) {
  return strcmp(entry->d_name, ".") != 0 && strcmp(entry->d_name, "..") != 0;
}

                   TF_Status* tf_status) {}

void PopulateDefaultParam(TP_OptimizerRegistrationParams* params) {
  params->struct_size = TP_OPTIMIZER_REGISTRATION_PARAMS_STRUCT_SIZE;
  params->optimizer_configs->struct_size = TP_OPTIMIZER_CONFIGS_STRUCT_SIZE;
  params->optimizer->struct_size = TP_OPTIMIZER_STRUCT_SIZE;
  params->optimizer->create_func = nullptr;
  params->optimizer->optimize_func = optimize_func;

// SECTION 1. Implementation for `TF_RandomAccessFile`
// ----------------------------------------------------------------------------
namespace tf_random_access_file {

typedef struct PosixFile {
  const char* filename;
  int fd;
} PosixFile;

static void Cleanup(TF_RandomAccessFile* file) {
  auto posix_file = static_cast<PosixFile*>(file->plugin_file);
  close(posix_file->fd);

  bool* executed_flag;
  // If true, only explicit op placements are accepted. If false, uses
  // type-based dispatch.
  bool strict_scope_placement;
};

struct LoggedTensor {
  TFE_TensorHandle* tensor;
  LoggedTensor() = delete;
  explicit LoggedTensor(TFE_TensorHandle* tensor) : tensor(tensor) {}
  ~LoggedTensor() { TFE_DeleteTensorHandle(tensor); }
};

int64_t LoggedTensorDim(void* data, int dim_index, TF_Status* status) {

      session->graph,
      tensorflow::strings::StrCat("fifo_queue_enqueue_", tensor_id).c_str());
  if (enqueue_op == nullptr) {
    status->status = tensorflow::errors::Internal(
        "Unable to find the enqueue node in the TF graph.");
    return;
  }

  TF_Operation* placeholder_op = TF_GraphOperationByName(
      session->graph,
      tensorflow::strings::StrCat("arg_tensor_enqueue_", tensor_id).c_str());

  } else {
    *result_plugin_resource = nullptr;
  }
  status->status = cc_status;
}

// -------------------------  VariableInfo  ------------------------------------
struct TF_VariableInfo {
  TF_VariableInfo() = delete;
  // TF_VariableInfo is constructed here by TensorFlow, and will be passed to
  // plugin as a opaque pointer. Plugin will need to call C APIs below to

  // The resulting TensorHandle owns an opaque pointer to "device memory", which
  // for a ParallelDevice is really a ParallelTensor. When the TensorHandle is
  // deleted, it will call ParallelTensorDeallocator to free the struct.
  ParallelTensor* t_released = t.release();
  TFE_CustomDeviceTensorHandleMethods handle_methods;
  handle_methods.num_dims = &ParallelTensorNumDims;
  handle_methods.dim = &ParallelTensorDim;

#include "tsl/platform/errors.h"

namespace {

#define VALIDATE_STRUCT_SIZE(STRUCT_NAME, STRUCT_OBJ, SIZE_VALUE_NAME) \
  do {                                                                 \
    if (STRUCT_OBJ.struct_size == 0) {                                 \
      return absl::Status(absl::StatusCode::kFailedPrecondition,       \
                          "struct_size field in " #STRUCT_NAME         \

  auto allocator = std::make_unique<AsyncValueAllocator>();
  tensorflow::Tensor tensor(allocator.get(), DT_FLOAT, {1});

  EXPECT_THAT(
      GetPjRtCBufferFromTensor(&tensor),
      StatusIs(error::INTERNAL, HasSubstr(absl::StrCat(
                                    "Input tensor does not have PjRtBuffer"))));
}

TEST(TensorPjRtBufferUtilTest, GetPjRtCBufferFromTensorIncoorectType) {
  auto allocator = std::make_unique<AsyncValueAllocator>();

  Status s = Shape(&shape);
  std::string shape_string;
  if (!s.ok()) {
    shape_string = "<error computing shape>";
  } else {
    shape_string = shape.DebugString();
  }
  return absl::StrCat("TensorHandle(shape=", shape_string,
                      ", dtype=", DataType_Name(DataType()),
                      ", type=", FullType().DebugString(), ")");
}

Status AbstractTensorHandle::TensorHandleStatus() const {