TF_CAPI_EXPORT extern void TFE_ContextAsyncWait(TFE_Context* ctx,
                                                TF_Status* status);

// This function will block till the operation that produces `h` has
// completed. This is only valid on local TFE_TensorHandles. The pointer
// returned will be on the device in which the TFE_TensorHandle resides (so e.g.
// for a GPU tensor this will return a pointer to GPU memory). The pointer is

  std::vector<TensorHandlePtr> tensors_;
  // Parallel tensors are immutable but compute their shape lazily unless it is
  // provided on construction. The optional has a value if the lazy computation
  // has been completed or the shape was provided on construction.
  mutable absl::optional<std::vector<int64_t>> shape_;
  const TF_DataType dtype_;
};

template <typename DataType>

// will block till the operation that produces `h` has completed.
TF_CAPI_EXPORT extern const char* TFE_TensorHandleDeviceName(
    TFE_TensorHandle* h, TF_Status* status);

// Returns the name of the device in whose memory `h` resides.
//
// This function will block till the operation that produces `h` has completed.
TF_CAPI_EXPORT extern const char* TFE_TensorHandleBackingDeviceName(