41 lines
1.6 KiB
C++
41 lines
1.6 KiB
C++
// Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
#pragma once
|
|
|
|
#include <vector>
|
|
|
|
#include "paddle/phi/core/dense_tensor.h"
|
|
|
|
namespace phi {
|
|
|
|
// WHY add this op?
|
|
// This op is used for convert fused_all_reduce_op_handle in Graph to Program.
|
|
// i.e, fused_all_reduce_op_handle = check_memory_continue + all_reduce_sum
|
|
// There are two reasons that check_memory_continue is added:
|
|
// 1. all_reduce_sum takes single tensor as input, while
|
|
// fused_all_reduce_op_handle takes tensor array as input, so we need a op to
|
|
// convert tensor array into a single tensor
|
|
// 2. fused_all_reduce_op_handle has a premise that all tensor's addresses are
|
|
// continue, so we need a op to do the check.
|
|
|
|
// see details in fused_all_reduce_op_handle.cc
|
|
template <typename T, typename Context>
|
|
void CheckMemoryContinueKernel(const Context &dev_ctx,
|
|
const std::vector<const DenseTensor *> &input,
|
|
DenseTensor *output,
|
|
std::vector<DenseTensor *> xout);
|
|
|
|
} // namespace phi
|