helper.h 4.3 KB
Newer Older
T
tensor-tang 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
/* Copyright (c) 2018 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 <string>
#include <vector>
#include "paddle/fluid/operators/jit/gen_base.h"
#include "paddle/fluid/operators/jit/kernel_base.h"
#include "paddle/fluid/operators/jit/kernel_key.h"
#include "paddle/fluid/operators/jit/kernel_pool.h"
#include "paddle/fluid/platform/place.h"

namespace paddle {
namespace operators {
namespace jit {

T
tensor-tang 已提交
29 30 31 32
#define SIGMOID_THRESHOLD_MIN -40.0
#define SIGMOID_THRESHOLD_MAX 13.0
#define EXP_MAX_INPUT 40.0

T
tensor-tang 已提交
33
template <KernelType KT, typename KernelTuples, typename PlaceType>
T
tensor-tang 已提交
34 35 36 37 38
inline typename std::enable_if<
    std::is_same<typename KernelTuples::data_type, float>::value &&
        std::is_same<PlaceType, platform::CPUPlace>::value,
    typename KernelTuples::func_type>::type
GetJitCode(typename KernelTuples::attr_type attr) {
T
tensor-tang 已提交
39 40
  using Func = typename KernelTuples::func_type;
  using Attr = typename KernelTuples::attr_type;
T
tensor-tang 已提交
41 42 43 44 45 46
  size_t key = JitCodeKey<Attr>(attr);
  auto& codes = JitCodePool<KT>().Instance();
  if (codes.Has(key)) {
    return codes.AllKernels().at(key)->template getCode<Func>();
  }

T
tensor-tang 已提交
47
  // creator is not related with attr, so can use KernelKey as key
T
tensor-tang 已提交
48
  KernelKey kkey(KT, PlaceType());
T
tensor-tang 已提交
49 50 51 52 53 54 55 56 57 58 59 60 61
  // pool: (KernelKey(type, place), vector<GenCreatorPtr>)
  auto& creator_map = JitCodeCreatorPool().Instance().AllCreators();
  auto iter = creator_map.find(kkey);
  if (iter != creator_map.end()) {
    auto& creators = iter->second;
    for (auto& cur : creators) {
      auto i = dynamic_cast<const JitCodeCreator<Attr>*>(cur.get());
      if (i && i->UseMe(attr)) {
        auto p = i->CreateJitCode(attr);
        if (p) {
          auto f = p->template getCode<Func>();
          codes.Insert(key, std::move(p));
          return f;
T
tensor-tang 已提交
62 63 64 65
        }
      }
    }
  }
T
tensor-tang 已提交
66 67 68
  return nullptr;
}

T
tensor-tang 已提交
69 70 71 72 73 74 75 76 77
template <KernelType KT, typename KernelTuples, typename PlaceType>
inline typename std::enable_if<
    !std::is_same<typename KernelTuples::data_type, float>::value ||
        !std::is_same<PlaceType, platform::CPUPlace>::value,
    typename KernelTuples::func_type>::type
GetJitCode(typename KernelTuples::attr_type attr) {
  return nullptr;
}

T
tensor-tang 已提交
78 79
// Refer code do not related with attr, which is just for cast
// Refer is always on CPUPlace
T
tensor-tang 已提交
80 81
template <KernelType KT, typename KernelTuples>
inline typename KernelTuples::func_type GetRefer() {
T
tensor-tang 已提交
82 83 84 85 86 87 88
  auto& ref_pool = ReferKernelPool().Instance().AllKernels();
  KernelKey kkey(KT, platform::CPUPlace());
  auto ref_iter = ref_pool.find(kkey);
  PADDLE_ENFORCE(ref_iter != ref_pool.end(),
                 "Every Kernel should have reference function.");
  auto& ref_impls = ref_iter->second;
  for (auto& impl : ref_impls) {
T
tensor-tang 已提交
89
    auto i = dynamic_cast<const ReferKernel<KernelTuples>*>(impl.get());
T
tensor-tang 已提交
90 91 92 93 94 95 96
    if (i) {
      return i->GetFunc();
    }
  }
  return nullptr;
}

T
tensor-tang 已提交
97
template <KernelType KT, typename KernelTuples,
T
tensor-tang 已提交
98
          typename PlaceType = platform::CPUPlace>
T
tensor-tang 已提交
99 100
typename KernelTuples::func_type Get(typename KernelTuples::attr_type attr) {
  auto jitfunc = GetJitCode<KT, KernelTuples, PlaceType>(attr);
T
tensor-tang 已提交
101 102 103
  if (jitfunc) {
    return jitfunc;
  }
T
tensor-tang 已提交
104 105

  // pool: (KernelKey(type, place), vector<KernelPtr>)
T
tensor-tang 已提交
106
  KernelKey kkey(KT, PlaceType());
T
tensor-tang 已提交
107 108 109 110 111
  auto& pool = KernelPool().Instance().AllKernels();
  auto iter = pool.find(kkey);
  if (iter != pool.end()) {
    auto& impls = iter->second;
    for (auto& impl : impls) {
T
tensor-tang 已提交
112
      auto i = dynamic_cast<const KernelImpl<KernelTuples>*>(impl.get());
T
tensor-tang 已提交
113 114 115 116 117 118 119
      if (i && i->UseMe(attr)) {
        return i->GetFunc();
      }
    }
  }

  // The last implementation should be reference function on CPUPlace.
T
tensor-tang 已提交
120
  return GetRefer<KT, KernelTuples>();
T
tensor-tang 已提交
121 122
}

123 124
const char* to_string(KernelType kt);

T
tensor-tang 已提交
125 126
KernelType to_kerneltype(const std::string& act);

T
tensor-tang 已提交
127 128 129
}  // namespace jit
}  // namespace operators
}  // namespace paddle