*status = handle.NumDims(&rank);
  if (!status->ok()) {
    return shape;
  }
  shape.reserve(rank);
  for (int i = 0; i < rank; ++i) {
    int64_t dim;
    *status = handle.Dim(i, &dim);
    if (!status->ok()) {
      return shape;
    }
    shape.push_back(dim);
  }
  return shape;
}

}  // namespace

extern "C" {

TF_CAPI_EXPORT extern TFE_TensorDebugInfo* TFE_TensorHandleTensorDebugInfo(

  // ImmediateExecutionTensorHandles, but comes up for example when computing
  // the shape of a parallel tensor with component shapes differing across
  // devices.
  virtual absl::Status Dim(int dim_index, int64_t* dim) const = 0;

  // Returns the device which created the handle.
  virtual const char* DeviceName(absl::Status* status) const = 0;
  // Returns the device where the tensor was placed.

  if (h == nullptr) {
    status->status = tensorflow::errors::InvalidArgument("Invalid handle");
    return -1;
  }

  int64_t dim = -1;
  status->status = tensorflow::unwrap(h)->Dim(dim_index, &dim);
  return dim;
}

const char* TFE_TensorHandleDeviceName(TFE_TensorHandle* h, TF_Status* status) {
  if (h == nullptr) {

      shape.dims = nullptr;
      continue;
    }
    shape.dims = new int64_t[shape.num_dims];
    for (size_t j = 0; j < shape.num_dims; ++j) {
      shape.dims[j] = c.Value(c.Dim(shape_handle, j));
    }
  }
  if (output_shapes != nullptr) *output_shapes = output_shapes_result;

  // TODO(bgogul): Set output_resource_shapes_and_types.
}

  ParallelTensor* t_released = t.release();
  TFE_CustomDeviceTensorHandleMethods handle_methods;
  handle_methods.num_dims = &ParallelTensorNumDims;
  handle_methods.dim = &ParallelTensorDim;
  handle_methods.deallocator = &ParallelTensorDeallocator;
  handle_methods.summarize = &ParallelTensorSummarize;
  return TensorHandlePtr(TFE_NewCustomDeviceTensorHandle(

    return;
  }

  // Rank is greater than 0, so fill in the values, if known, and
  // -1 for unknown values.
  for (int i = 0; i < num_dims; ++i) {
    auto dim = ic->Dim(shape, i);
    int64_t value = -1;
    if (ic->ValueKnown(dim)) {
      value = ic->Value(dim);
    }
    dims[i] = value;
  }
}

// TF_OperationDescription functions ------------------------------------------

extern "C" {

#undef EXPECT_TF_META

TEST(CAPI, TestTensorAligned) {
  int64_t dim = 7;
  size_t tensor_size_bytes = dim * TF_DataTypeSize(TF_FLOAT);
  TF_Tensor* a = TF_AllocateTensor(
      /*dtype=*/TF_FLOAT, /*dims=*/&dim, /*num_dims=*/1,
      /*len=*/tensor_size_bytes);
  float* data = reinterpret_cast<float*>(TF_TensorData(a));
  for (int i = 0; i < dim; ++i) {
    data[i] = 0;
  }
  if (TF_TensorDefaultAlignment() > 0) {

t-Samhain","An D\xf9bhlachd"],monthsShort:["Faoi","Gear","M\xe0rt","Gibl","C\xe8it","\xd2gmh","Iuch","L\xf9n","Sult","D\xe0mh","Samh","D\xf9bh"],monthsParseExact:!0,weekdays:["Did\xf2mhnaich","Diluain","Dim\xe0irt","Diciadain","Diardaoin","Dihaoine","Disathairne"],weekdaysShort:["Did","Dil","Dim","Dic","Dia","Dih","Dis"],weekdaysMin:["D\xf2","Lu","M\xe0","Ci","Ar","Ha","Sa"],longDateFormat:{LT:"HH:mm",LTS:"HH:mm:ss",L:"DD/MM/YYYY",LL:"D MMMM YYYY",LLL:"D MMMM YYYY HH:mm",LLLL:"dddd, D MMMM YYYY ...

* `tf.argmax`: `dimension` becomes `axis` * `tf.argmin`: `dimension` becomes
`axis` * `tf.count_nonzero`: `reduction_indices` becomes `axis` *
`tf.expand_dims`: `dim` becomes `axis` * `tf.reduce_all`: `reduction_indices`
becomes `axis` * `tf.reduce_any`: `reduction_indices` becomes `axis` *
`tf.reduce_join`: `reduction_indices` becomes `axis` * `tf.reduce_logsumexp`: