op_registry.h 11.5 KB
Newer Older
1 2
#pragma once

3
#include <algorithm>
Y
Yu Yang 已提交
4
#include <type_traits>
5 6
#include <unordered_map>
#include <unordered_set>
Q
Qiao Longfei 已提交
7
#include "paddle/framework/attr_checker.h"
8 9
#include "paddle/framework/op_desc.pb.h"
#include "paddle/framework/op_proto.pb.h"
Q
Qiao Longfei 已提交
10
#include "paddle/framework/operator.h"
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63

namespace paddle {
namespace framework {

// helper class to set attribute type
struct AttrTypeHelper {
  template <typename T>
  static void SetAttrType(AttrProto* attr);

  static Attribute GetAttrValue(const AttrDesc& attr_desc) {
    switch (attr_desc.type()) {
      case paddle::framework::AttrType::INT: {
        return attr_desc.i();
      }
      case paddle::framework::AttrType::FLOAT: {
        return attr_desc.f();
      }
      case paddle::framework::AttrType::STRING: {
        return attr_desc.s();
      }
      case paddle::framework::AttrType::INTS: {
        std::vector<int> val(attr_desc.ints_size());
        for (int i = 0; i < attr_desc.ints_size(); ++i) {
          val[i] = attr_desc.ints(i);
        }
        return val;
      }
      case paddle::framework::AttrType::FLOATS: {
        std::vector<float> val(attr_desc.floats_size());
        for (int i = 0; i < attr_desc.floats_size(); ++i) {
          val[i] = attr_desc.floats(i);
        }
        return val;
      }
      case paddle::framework::AttrType::STRINGS: {
        std::vector<std::string> val(attr_desc.strings_size());
        for (int i = 0; i < attr_desc.strings_size(); ++i) {
          val[i] = attr_desc.strings(i);
        }
        return val;
      }
    }
    PADDLE_ENFORCE(false, "Unknown OpDesc::AttrDesc::type !");
    return boost::blank();
  }
};

// this class not only make proto but also init attribute checkers.
class OpProtoAndCheckerMaker {
 public:
  OpProtoAndCheckerMaker(OpProto* proto, OpAttrChecker* op_checker)
      : proto_(proto), op_checker_(op_checker) {}

64 65 66 67 68 69 70 71
  ~OpProtoAndCheckerMaker() {
    PADDLE_ENFORCE(validated_, "should call Validate after build");
  }

  void Validate() {
    validated_ = true;
    CheckNoDuplicatedInOutAttrs();
  }
72

73
 protected:
74 75
  void AddInput(const std::string& name, const std::string& comment,
                bool multiple = false) {
76
    auto input = proto_->mutable_inputs()->Add();
77 78
    *input->mutable_name() = name;
    *input->mutable_comment() = comment;
79 80 81 82 83 84 85 86
    input->set_multiple(multiple);
    if (multiple) {
      SetHasMultipleInput();
    }
  }

  void AddInputs(const std::string& name, const std::string& comment) {
    AddInput(name, comment, true);
87 88
  }

89 90
  void AddOutput(const std::string& name, const std::string& comment,
                 bool temporary = false, bool multiple = false) {
91
    auto output = proto_->mutable_outputs()->Add();
92 93
    *output->mutable_name() = name;
    *output->mutable_comment() = comment;
94 95 96 97 98 99 100 101 102 103 104 105 106
    output->set_multiple(multiple);
    if (multiple) {
      SetHasMultipleOutput();
    }
    output->set_temporary(temporary);
    if (temporary) {
      SetHasTemporaryOutput();
    }
  }

  void AddOutputs(const std::string& name, const std::string& comment,
                  bool temporary = false) {
    AddOutput(name, comment, temporary, true);
107 108 109 110
  }

  template <typename T>
  TypedAttrChecker<T>& AddAttr(const std::string& name,
111 112
                               const std::string& comment,
                               bool generated = false) {
113
    auto attr = proto_->mutable_attrs()->Add();
114 115
    *attr->mutable_name() = name;
    *attr->mutable_comment() = comment;
116
    attr->set_generated(generated);
117 118 119 120 121 122 123 124
    AttrTypeHelper::SetAttrType<T>(attr);
    return op_checker_->AddAttrChecker<T>(name);
  }

  void AddComment(const std::string& comment) {
    *(proto_->mutable_comment()) = comment;
  }

125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
 private:
  void SetHasMultiple(const std::string& in_out, bool* flag) {
    if (!*flag) {
      AddAttr<std::vector<int>>(in_out + "_format",
                                "The multiple index of " + in_out +
                                    "\n"
                                    R"DOC(
This attribute is used by Paddle core framework. Paddle's Op support each input
or output could be a list of variable. This attribute is used to show how that
list organized.

e.g.
  input = ["a", "b", "c", "d", "e", "f"]
  input_format = [0, 4, 5, 6]

means
  The number of all input variables this op is six, and they are segmented into
  three inputs.

  The first input is input[0:4], second is input[4:5], third is input[5:6].
)DOC",
                                /*generated*/ true);
      *flag = true;
    }
  }

  void SetHasMultipleInput() { SetHasMultiple("input", &has_multiple_input_); }
  void SetHasMultipleOutput() {
    SetHasMultiple("output", &has_multiple_output_);
  }

  void SetHasTemporaryOutput() {
    if (!has_temporary_output_) {
      AddAttr<std::vector<int>>("temporary_index",
                                R"DOC(The temporary index of output.

Not all output of Paddle Op is used by user. For faster computation, each op
could output some its internal state to other op, other op could take that
output to make compute faster.

Add a mark to which output is temporary is helpful for future optimization.
)DOC",
                                /*generated*/ true)
          .SetDefault(std::vector<int>());
      has_temporary_output_ = true;
    }
  }

173
  void CheckNoDuplicatedInOutAttrs() {
174
    std::unordered_set<std::string> names;
175 176 177 178
    auto checker = [&](const std::string& name) {
      PADDLE_ENFORCE(!names.count(name), "[%s] is duplicated", name);
      names.insert(name);
    };
179
    for (auto& attr : proto_->attrs()) {
180 181 182 183 184 185 186
      checker(attr.name());
    }
    for (auto& input : proto_->inputs()) {
      checker(input.name());
    }
    for (auto& output : proto_->outputs()) {
      checker(output.name());
187 188 189
    }
  }

190 191
  OpProto* proto_;
  OpAttrChecker* op_checker_;
192
  bool validated_{false};
193 194 195
  bool has_multiple_input_{false};
  bool has_multiple_output_{false};
  bool has_temporary_output_{false};
196 197 198
};

class OpRegistry {
Q
Qiao Longfei 已提交
199
  using OpCreator = std::function<OperatorBase*()>;
200 201 202 203

 public:
  template <typename OpType, typename ProtoMakerType>
  static void RegisterOp(const std::string& op_type) {
204 205 206
    creators()[op_type] = [] { return new OpType; };
    OpProto& op_proto = protos()[op_type];
    OpAttrChecker& op_checker = op_checkers()[op_type];
207 208
    auto maker = ProtoMakerType(&op_proto, &op_checker);
    maker.Validate();
Y
Yu Yang 已提交
209 210 211 212 213
    *op_proto.mutable_type() = op_type;
    PADDLE_ENFORCE(
        op_proto.IsInitialized(),
        "Fail to initialize %s's OpProto, because %s is not initialized",
        op_type, op_proto.InitializationErrorString());
214 215
  }

Q
Qiao Longfei 已提交
216
  static OperatorPtr CreateOp(const OpDesc& op_desc) {
217
    std::string op_type = op_desc.type();
Q
Qiao Longfei 已提交
218
    OperatorPtr op(creators().at(op_type)());
Q
Qiao Longfei 已提交
219
    op->type_ = op_desc.type();
220 221 222 223 224 225 226
    op->inputs_.reserve((size_t)op_desc.inputs_size());
    std::copy(op_desc.inputs().begin(), op_desc.inputs().end(),
              std::back_inserter(op->inputs_));
    op->outputs_.reserve((size_t)op_desc.outputs_size());
    std::copy(op_desc.outputs().begin(), op_desc.outputs().end(),
              std::back_inserter(op->outputs_));
    for (auto& attr : op_desc.attrs()) {
Q
Qiao Longfei 已提交
227
      op->attrs_[attr.name()] = AttrTypeHelper::GetAttrValue(attr);
228
    }
Q
Qiao Longfei 已提交
229
    op_checkers().at(op_type).Check(op->attrs_);
Q
Qiao Longfei 已提交
230
    op->Init();
231 232 233
    return op;
  }

Y
Yu Yang 已提交
234 235 236 237 238
  static std::unordered_map<std::string, OpProto>& protos() {
    static std::unordered_map<std::string, OpProto> protos_;
    return protos_;
  };

239
 private:
240 241 242 243
  static std::unordered_map<std::string, OpCreator>& creators() {
    static std::unordered_map<std::string, OpCreator> creators_;
    return creators_;
  }
244

245 246 247 248 249
  static std::unordered_map<std::string, OpAttrChecker>& op_checkers() {
    static std::unordered_map<std::string, OpAttrChecker> op_checkers_;
    return op_checkers_;
  };
};
250 251 252 253

template <typename OpType, typename ProtoMakerType>
class OpRegisterHelper {
 public:
Y
Yu Yang 已提交
254
  OpRegisterHelper(const char* op_type) {
255 256 257 258
    OpRegistry::RegisterOp<OpType, ProtoMakerType>(op_type);
  }
};

259 260 261
/**
 * check if MACRO is used in GLOBAL NAMESPACE.
 */
Y
Yu Yang 已提交
262 263 264 265 266 267
#define STATIC_ASSERT_GLOBAL_NAMESPACE(uniq_name, msg)                        \
  struct __test_global_namespace_##uniq_name##__ {};                          \
  static_assert(std::is_same<::__test_global_namespace_##uniq_name##__,       \
                             __test_global_namespace_##uniq_name##__>::value, \
                msg)

268 269 270
/**
 * Macro to Register Operator.
 */
Y
Yu Yang 已提交
271 272 273 274 275 276 277
#define REGISTER_OP(__op_type, __op_class, __op_maker_class)                 \
  STATIC_ASSERT_GLOBAL_NAMESPACE(__reg_op__##__op_type,                      \
                                 "REGISTER_OP must be in global namespace"); \
  static ::paddle::framework::OpRegisterHelper<__op_class, __op_maker_class> \
      __op_register_##__op_type##__(#__op_type);                             \
  int __op_register_##__op_type##_handle__() { return 0; }

278 279 280 281
/**
 * Macro to Register OperatorKernel.
 */
#define REGISTER_OP_KERNEL(type, DEVICE_TYPE, PlaceType, KernelType)      \
Y
Yu Yang 已提交
282
  STATIC_ASSERT_GLOBAL_NAMESPACE(                                         \
283
      __reg_op_kernel_##type##_##DEVICE_TYPE##__,                         \
Y
Yu Yang 已提交
284 285 286 287 288 289 290 291 292 293
      "REGISTER_OP_KERNEL must be in global namespace");                  \
  struct __op_kernel_register__##type##__ {                               \
    __op_kernel_register__##type##__() {                                  \
      ::paddle::framework::OperatorWithKernel::OpKernelKey key;           \
      key.place_ = PlaceType();                                           \
      ::paddle::framework::OperatorWithKernel::AllOpKernels()[#type][key] \
          .reset(new KernelType());                                       \
    }                                                                     \
  };                                                                      \
  static __op_kernel_register__##type##__ __reg_kernel_##type##__;        \
294
  int __op_kernel_register_##type##_handle_##DEVICE_TYPE##__() { return 0; }
Y
Yu Yang 已提交
295 296 297 298 299 300 301

#define REGISTER_OP_GPU_KERNEL(type, KernelType) \
  REGISTER_OP_KERNEL(type, GPU, ::paddle::platform::GPUPlace, KernelType)

#define REGISTER_OP_CPU_KERNEL(type, KernelType) \
  REGISTER_OP_KERNEL(type, CPU, ::paddle::platform::CPUPlace, KernelType)

302 303 304 305
/**
 * Macro to mark what Operator and Kernel we will use and tell the compiler to
 * link them into target.
 */
Y
Yu Yang 已提交
306 307 308 309 310 311 312 313
#define USE_OP_WITHOUT_KERNEL(op_type)                      \
  STATIC_ASSERT_GLOBAL_NAMESPACE(                           \
      __use_op_without_kernel_##op_type,                    \
      "USE_OP_WITHOUT_KERNEL must be in global namespace"); \
  extern int __op_register_##op_type##_handle__();          \
  static int __use_op_ptr_##op_type##_without_kernel__      \
      __attribute__((unused)) = __op_register_##op_type##_handle__()

Y
Yu Yang 已提交
314 315 316 317 318 319 320 321
#define USE_OP_KERNEL(op_type, DEVICE_TYPE)                               \
  STATIC_ASSERT_GLOBAL_NAMESPACE(                                         \
      __use_op_kernel_##op_type##_##DEVICE_TYPE##__,                      \
      "USE_OP_KERNEL must be in global namespace");                       \
  extern int __op_kernel_register_##op_type##_handle_##DEVICE_TYPE##__(); \
  static int __use_op_ptr_##op_type##_##DEVICE_TYPE##_kernel__            \
      __attribute__((unused)) =                                           \
          __op_kernel_register_##op_type##_handle_##DEVICE_TYPE##__()
Y
Yu Yang 已提交
322

323 324
// use Operator with only cpu kernel.
#define USE_OP_CPU(op_type)       \
Y
Yu Yang 已提交
325
  USE_OP_WITHOUT_KERNEL(op_type); \
326
  USE_OP_KERNEL(op_type, CPU)
Y
Yu Yang 已提交
327

328 329
#ifdef PADDLE_ONLY_CPU
#define USE_OP(op_type) USE_OP_CPU(op_type)
Y
Yu Yang 已提交
330
#else
331 332
#define USE_OP(op_type) \
  USE_OP_CPU(op_type);  \
Y
Yu Yang 已提交
333 334
  USE_OP_KERNEL(op_type, GPU)
#endif
335 336 337

}  // namespace framework
}  // namespace paddle