TFE_Execute(matmul, &retvals[0], &num_retvals, status);
    CHECK_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
    if (state.iterations() >= state.max_iterations && async) {
      TFE_Executor* executor = TFE_ContextGetExecutorForThread(ctx);
      TFE_ExecutorWaitForAllPendingNodes(executor, status);
      ASSERT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
      TFE_DeleteExecutor(executor);
    }
  }
  TFE_DeleteOp(matmul);