memory.cc 4.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.

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. */

#include "paddle/memory/memory.h"
L
liaogang 已提交
16 17
#include "paddle/memory/detail/buddy_allocator.h"
#include "paddle/memory/detail/system_allocator.h"
18

L
liaogang 已提交
19 20
#include <stdlib.h>  // for memcpy

21 22 23
namespace paddle {
namespace memory {

L
liaogang 已提交
24 25 26 27 28 29 30
detail::BuddyAllocator* GetCPUBuddyAllocator() {
  static detail::BuddyAllocator* a = nullptr;
  if (a == nullptr) {
    a = new detail::BuddyAllocator(new detail::CPUAllocator,
                                   platform::CpuMinChunkSize(),
                                   platform::CpuMaxChunkSize());
  }
L
liaogang 已提交
31
  return a;
L
liaogang 已提交
32 33
}

L
liaogang 已提交
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
template <>
void* Alloc<platform::CPUPlace>(platform::CPUPlace place, size_t size) {
  return GetCPUBuddyAllocator()->Alloc(size);
}

template <>
void Free<platform::CPUPlace>(platform::CPUPlace place, void* p) {
  GetCPUBuddyAllocator()->Free(p);
}

template <>
size_t Used<platform::CPUPlace>(platform::CPUPlace place) {
  return GetCPUBuddyAllocator()->Used();
}

L
liaogang 已提交
49 50
template <>
void Copy<platform::CPUPlace, platform::CPUPlace>(platform::CPUPlace, void* dst,
L
liaogang 已提交
51 52 53
                                                  platform::CPUPlace,
                                                  const void* src, size_t num) {
  memcpy(dst, src, num);
L
liaogang 已提交
54 55
}

L
liaogang 已提交
56
#ifndef PADDLE_ONLY_CPU
L
liaogang 已提交
57 58 59 60

detail::BuddyAllocator* GetGPUBuddyAllocator(int gpu_id) {
  static detail::BuddyAllocator** as = NULL;
  if (as == NULL) {
L
liaogang 已提交
61
    int gpu_num = platform::GetDeviceCount();
L
liaogang 已提交
62 63 64 65 66 67 68 69 70 71 72
    as = new detail::BuddyAllocator*[gpu_num];
    for (int gpu = 0; gpu < gpu_num; gpu++) {
      platform::SetDeviceId(gpu);
      as[gpu] = new detail::BuddyAllocator(new detail::GPUAllocator,
                                           platform::GpuMinChunkSize(),
                                           platform::GpuMaxChunkSize());
    }
  }
  return as[gpu_id];
}

L
liaogang 已提交
73 74 75 76
template <>
void* Alloc<platform::GPUPlace>(platform::GPUPlace place, size_t size) {
  return GetGPUBuddyAllocator(place.device)->Alloc(size);
}
L
liaogang 已提交
77

L
liaogang 已提交
78 79 80
template <>
void Free<platform::GPUPlace>(platform::GPUPlace place, void* p) {
  GetGPUBuddyAllocator(place.device)->Free(p);
81 82
}

L
liaogang 已提交
83 84 85
template <>
size_t Used<platform::GPUPlace>(platform::GPUPlace place) {
  return GetGPUBuddyAllocator(place.device)->Used();
86 87
}

L
liaogang 已提交
88 89 90 91 92
template <>
void Copy<platform::CPUPlace, platform::GPUPlace>(platform::CPUPlace, void* dst,
                                                  platform::GPUPlace,
                                                  const void* src, size_t num,
                                                  cudaStream_t stream) {
L
liaogang 已提交
93
  platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyDeviceToHost, stream);
L
liaogang 已提交
94 95 96 97 98 99 100
}

template <>
void Copy<platform::GPUPlace, platform::CPUPlace>(platform::GPUPlace, void* dst,
                                                  platform::CPUPlace,
                                                  const void* src, size_t num,
                                                  cudaStream_t stream) {
L
liaogang 已提交
101
  platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyHostToDevice, stream);
L
liaogang 已提交
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
}

template <>
void Copy<platform::GPUPlace, platform::GPUPlace>(platform::GPUPlace dst_place,
                                                  void* dst,
                                                  platform::GPUPlace src_place,
                                                  const void* src, size_t num,
                                                  cudaStream_t stream) {
  if (dst_place == src_place) {
    platform::GpuMemcpyAsync(dst, src, num, cudaMemcpyDeviceToDevice, stream);
  } else {
    platform::GpuMemcpyPeer(dst, dst_place.device, src, src_place.device, num,
                            stream);
  }
}

118
#endif  // PADDLE_ONLY_CPU
119 120 121

}  // namespace memory
}  // namespace paddle