operator.cc 19.7 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
Q
Qiao Longfei 已提交
2 3 4 5 6 7 8 9 10 11 12 13

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. */
D
dzhwinter 已提交
14
#include <gflags/gflags.h>
D
dzhwinter 已提交
15
#include <glog/logging.h>
Q
Qiao Longfei 已提交
16

17
#include <algorithm>
D
dzhwinter 已提交
18

Y
Yi Wang 已提交
19 20 21 22 23
#include "paddle/fluid/framework/data_transform.h"
#include "paddle/fluid/framework/executor.h"
#include "paddle/fluid/framework/operator.h"
#include "paddle/fluid/framework/shape_inference.h"
#include "paddle/fluid/framework/var_type.h"
24
#include "paddle/fluid/platform/profiler.h"
Q
Qiao Longfei 已提交
25

D
dzhwinter 已提交
26
DECLARE_bool(benchmark);
D
dzhwinter 已提交
27

Q
Qiao Longfei 已提交
28 29 30
namespace paddle {
namespace framework {

31 32 33 34 35 36
std::vector<std::tuple<platform::Place, LibraryType>> kKernelPriority = {
    std::make_tuple(platform::CUDAPlace(0), LibraryType::kCUDNN),
    std::make_tuple(platform::CUDAPlace(0), LibraryType::kPlain),
    std::make_tuple(platform::CPUPlace(), LibraryType::kMKLDNN),
    std::make_tuple(platform::CPUPlace(), LibraryType::kPlain),
};
D
dzhwinter 已提交
37

38 39
static DDim GetDims(const Scope& scope, const std::string& name) {
  Variable* var = scope.FindVar(name);
Q
qiaolongfei 已提交
40 41
  if (var == nullptr) {
    return DDim({-1});
Q
Qiao Longfei 已提交
42 43 44
  }

  if (var->IsType<LoDTensor>()) {
45 46 47 48 49 50 51 52
    return var->Get<LoDTensor>().dims();
  } else if (var->IsType<SelectedRows>()) {
    return var->Get<SelectedRows>().GetCompleteDims();
  } else {
    return DDim({-1});
  }
}

Q
Qiao Longfei 已提交
53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
static LoD GetLoD(const Scope& scope, const std::string& name) {
  Variable* var = scope.FindVar(name);
  auto default_lod = LoD({{}});

  if (var == nullptr) {
    return default_lod;
  }

  if (var->IsType<LoDTensor>()) {
    return var->Get<LoDTensor>().lod();
  } else {
    return default_lod;
  }
}

68 69 70 71 72 73 74 75 76 77 78 79
void OperatorBase::Run(const Scope& scope, const platform::Place& place) {
  if (platform::is_gpu_place(place)) {
#ifndef PADDLE_WITH_CUDA
    PADDLE_THROW("Cannot run operator on place %s", place);
#else
    auto dev_id = boost::get<platform::CUDAPlace>(place).device;
    platform::SetDeviceId(dev_id);
#endif
  }
  RunImpl(scope, place);
}

80
std::string OperatorBase::Input(const std::string& name) const {
Y
Yu Yang 已提交
81
  auto& ins = Inputs(name);
Y
Yu Yang 已提交
82
  PADDLE_ENFORCE_LE(ins.size(), 1UL,
83 84
                    "Operator %s's input %s should contain only one variable.",
                    type_, name);
Y
Yu Yang 已提交
85
  return ins.empty() ? kEmptyVarName : ins[0];
Y
Yan Chunwei 已提交
86 87
}

Y
Yu Yang 已提交
88 89
const std::vector<std::string>& OperatorBase::Inputs(
    const std::string& name) const {
Y
Yu Yang 已提交
90
  auto it = inputs_.find(name);
91 92
  PADDLE_ENFORCE(it != inputs_.end(), "Operator %s does not have the input %s.",
                 type_, name);
Y
Yu Yang 已提交
93
  return it->second;
Y
Yan Chunwei 已提交
94 95
}

96
std::string OperatorBase::Output(const std::string& name) const {
Y
Yu Yang 已提交
97
  auto& outs = Outputs(name);
Y
Yu Yang 已提交
98
  PADDLE_ENFORCE_LE(outs.size(), 1UL,
99 100
                    "Operator %s's output %s should contain only one variable.",
                    type_, name);
Y
Yu Yang 已提交
101
  return outs.empty() ? kEmptyVarName : outs[0];
Y
Yan Chunwei 已提交
102 103
}

Y
Yu Yang 已提交
104 105
const std::vector<std::string>& OperatorBase::Outputs(
    const std::string& name) const {
Y
Yu Yang 已提交
106
  auto it = outputs_.find(name);
107 108
  PADDLE_ENFORCE(it != outputs_.end(),
                 "Operator %s does not have an output called %s.", type_, name);
Y
Yu Yang 已提交
109
  return it->second;
Y
Yan Chunwei 已提交
110 111
}

112
std::string OperatorBase::DebugStringEx(const Scope* scope) const {
Q
Qiao Longfei 已提交
113
  std::stringstream ss;
Y
Yu Yang 已提交
114
  ss << "Op(" << type_ << "), inputs:{";
Y
Yu Yang 已提交
115 116
  for (auto it = inputs_.begin(); it != inputs_.end();) {
    auto& input = *it;
Y
Yu Yang 已提交
117 118 119
    ss << input.first << "[";
    for (size_t i = 0; i < input.second.size(); ++i) {
      ss << input.second[i];
120
      if (scope) {
Q
Qiao Longfei 已提交
121 122
        ss << "[" << GetDims(*scope, input.second[i]) << "]";
        ss << "(" << GetLoD(*scope, input.second[i]) << ")";
123
      }
Y
Yu Yang 已提交
124 125 126
      if (i != input.second.size() - 1) {
        ss << ", ";
      }
127
    }
Y
Yu Yang 已提交
128
    ss << "]";
Y
Yu Yang 已提交
129 130
    ++it;
    if (it != inputs_.end()) {
131 132
      ss << ", ";
    }
Q
Qiao Longfei 已提交
133
  }
Y
Yu Yang 已提交
134
  ss << "}, outputs:{";
Y
Yu Yang 已提交
135 136
  for (auto it = outputs_.begin(); it != outputs_.end();) {
    auto& output = *it;
Y
Yu Yang 已提交
137 138 139
    ss << output.first << "[";
    for (size_t i = 0; i < output.second.size(); ++i) {
      ss << output.second[i];
140
      if (scope) {
Q
Qiao Longfei 已提交
141 142
        ss << "[" << GetDims(*scope, output.second[i]) << "]";
        ss << "(" << GetLoD(*scope, output.second[i]) << ")";
143
      }
Y
Yu Yang 已提交
144 145 146
      if (i != output.second.size() - 1) {
        ss << ", ";
      }
147
    }
Y
Yu Yang 已提交
148
    ss << "]";
Y
Yu Yang 已提交
149 150
    ++it;
    if (it != outputs_.end()) {
151 152
      ss << ", ";
    }
Q
Qiao Longfei 已提交
153
  }
Y
Yu Yang 已提交
154
  ss << "}.";
Q
Qiao Longfei 已提交
155 156 157
  return ss.str();
}

D
dongzhihong 已提交
158 159
void OperatorBase::Rename(const std::string& old_name,
                          const std::string& new_name) {
Y
Yu Yang 已提交
160 161 162 163 164 165 166
  for (auto& input : inputs_) {
    std::replace(input.second.begin(), input.second.end(), old_name, new_name);
  }
  for (auto& output : outputs_) {
    std::replace(output.second.begin(), output.second.end(), old_name,
                 new_name);
  }
D
dongzhihong 已提交
167 168
}

Y
Yu Yang 已提交
169
OperatorBase::OperatorBase(const std::string& type,
Y
Yu Yang 已提交
170 171
                           const VariableNameMap& inputs,
                           const VariableNameMap& outputs,
Y
Yu Yang 已提交
172 173
                           const AttributeMap& attrs)
    : type_(type), inputs_(inputs), outputs_(outputs), attrs_(attrs) {
174 175
  GenerateTemporaryNames();
  CheckAllInputOutputSet();
Y
Yu Yang 已提交
176
}
177

Q
qijun 已提交
178 179
std::vector<std::string> OperatorBase::InputVars() const {
  std::vector<std::string> ret_val;
Y
Yu Yang 已提交
180
  for (auto& o : inputs_) {
Q
qijun 已提交
181 182 183 184 185 186
    ret_val.reserve(ret_val.size() + o.second.size());
    ret_val.insert(ret_val.end(), o.second.begin(), o.second.end());
  }
  return ret_val;
}

Y
Yu Yang 已提交
187 188 189 190 191 192 193 194 195 196
std::vector<std::string> OperatorBase::OutputVars(bool has_intermediate) const {
  std::vector<std::string> ret_val;
  if (has_intermediate) {
    // push all outputs into ret_val
    for (auto& o : outputs_) {
      ret_val.reserve(ret_val.size() + o.second.size());
      ret_val.insert(ret_val.end(), o.second.begin(), o.second.end());
    }
    return ret_val;
  }
Y
Yu Yang 已提交
197
  auto& info = OpInfoMap::Instance().Get(Type());
Y
Yu Yang 已提交
198 199

  // get all OpProto::Var for outputs
Y
Yu Yang 已提交
200
  for (auto& o : info.Proto().outputs()) {
Y
Yu Yang 已提交
201 202 203 204 205 206 207 208 209
    // ignore all intermediate output
    if (o.intermediate()) continue;
    auto out = outputs_.find(o.name());
    if (out != outputs_.end()) {
      ret_val.reserve(ret_val.size() + out->second.size());
      ret_val.insert(ret_val.end(), out->second.begin(), out->second.end());
    }
  }
  return ret_val;
D
dongzhihong 已提交
210 211
}

212 213 214
void OperatorBase::CheckAllInputOutputSet() const {
  auto& info_map = OpInfoMap::Instance();
  auto* op_info = info_map.GetNullable(Type());
Y
Yu Yang 已提交
215
  if (op_info == nullptr || op_info->proto_ == nullptr) return;
216 217 218

  for (auto& in : op_info->Proto().inputs()) {
    PADDLE_ENFORCE(inputs_.find(in.name()) != inputs_.end(),
Y
Yu Yang 已提交
219
                   "Type %s's input %s is not set", Type(), in.name());
220 221 222 223
  }

  for (auto& out : op_info->Proto().outputs()) {
    PADDLE_ENFORCE(outputs_.find(out.name()) != outputs_.end(),
Y
Yu Yang 已提交
224
                   "Type %s's output %s is not set", Type(), out.name());
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
  }
}

void OperatorBase::GenerateTemporaryNames() {
  static std::atomic<size_t> gUniqId(0UL);
  for (auto& output : outputs_) {
    for (auto& output_name : output.second) {
      if (output_name == kTempVarName) {
        output_name += type_;
        output_name += "@";
        output_name += std::to_string(gUniqId.fetch_add(1));
      }
    }
  }
}

241 242 243 244
static bool VarIsTensor(const Variable* var) {
  return var->IsType<LoDTensor>() || var->IsType<SelectedRows>();
}

245
static const Tensor* GetTensorFromVar(Variable* var) {
Q
QI JUN 已提交
246
  if (var->IsType<LoDTensor>()) {
247
    return var->GetMutable<LoDTensor>();
Q
QI JUN 已提交
248
  } else if (var->IsType<SelectedRows>()) {
249
    return var->GetMutable<SelectedRows>()->mutable_value();
Q
QI JUN 已提交
250
  } else {
Y
Yang Yang 已提交
251 252
    PADDLE_THROW("Variable type_id %s, expect LoDTensor/SelectedRows.",
                 var->Type().name());
Q
QI JUN 已提交
253 254 255 256 257
  }
}

static Tensor* GetMutableTensorFromVar(Variable* var) {
  if (var->IsType<LoDTensor>()) {
258
    return var->GetMutable<LoDTensor>();
Q
QI JUN 已提交
259
  } else if (var->IsType<SelectedRows>()) {
260
    return var->GetMutable<SelectedRows>()->mutable_value();
Q
QI JUN 已提交
261
  } else {
Y
Yang Yang 已提交
262 263
    PADDLE_THROW("Variable type_id %s, expect LoDTensor/SelectedRows.",
                 var->Type().name());
Q
QI JUN 已提交
264 265 266
  }
}

267
template <>
268
const Tensor* ExecutionContext::Input<Tensor>(const std::string& name) const {
269
  auto* var = InputVar(name);
270 271
  return var == nullptr ? nullptr
                        : GetTensorFromVar(const_cast<Variable*>(var));
272 273 274
}

template <>
275
const std::vector<const Tensor*> ExecutionContext::MultiInput<Tensor>(
276 277 278 279
    const std::string& name) const {
  auto names = op().Inputs(name);
  std::vector<const Tensor*> res;
  res.reserve(names.size());
280 281 282 283 284
  std::transform(names.begin(), names.end(), std::back_inserter(res),
                 [&](const std::string& sub_name) {
                   auto var = scope_.FindVar(sub_name);
                   return var == nullptr ? nullptr : GetTensorFromVar(var);
                 });
285 286 287 288
  return res;
}

template <>
289
Tensor* ExecutionContext::Output<Tensor>(const std::string& name) const {
290
  auto var = OutputVar(name);
Q
QI JUN 已提交
291
  return var == nullptr ? nullptr : GetMutableTensorFromVar(var);
292 293 294
}

template <>
295
std::vector<Tensor*> ExecutionContext::MultiOutput<Tensor>(
296 297 298 299
    const std::string& name) const {
  auto names = op().Outputs(name);
  std::vector<Tensor*> res;
  res.reserve(names.size());
300 301
  std::transform(names.begin(), names.end(), std::back_inserter(res),
                 [&](const std::string& sub_name) {
302 303
                   auto var = scope_.FindVar(sub_name);
                   return var == nullptr ? nullptr
Q
QI JUN 已提交
304
                                         : GetMutableTensorFromVar(var);
305
                 });
306 307 308
  return res;
}

Y
Yu Yang 已提交
309 310 311 312 313
bool OpSupportGPU(const std::string& op_type) {
  auto& all_kernels = OperatorWithKernel::AllOpKernels();
  auto it = all_kernels.find(op_type);
  if (it == all_kernels.end()) {
    // All control operator must support GPU
314

Y
Yu Yang 已提交
315 316 317 318 319 320 321 322 323 324
    return true;
  }
  for (auto& kern_pair : it->second) {
    if (platform::is_gpu_place(kern_pair.first.place_)) {
      return true;
    }
  }
  return false;
}

325 326 327 328 329 330 331 332 333 334 335
class RuntimeInferShapeContext : public InferShapeContext {
 public:
  RuntimeInferShapeContext(const OperatorBase& op, const Scope& scope)
      : op_(op), scope_(scope) {}

  bool HasInput(const std::string& name) const override {
    auto& ins = Inputs(name);
    size_t length = ins.size();
    if (length == 0) {
      return false;
    }
F
fengjiayi 已提交
336 337
    PADDLE_ENFORCE_EQ(length, 1UL,
                      "Input %s should not have more than one inputs", name);
338 339 340 341 342 343 344 345 346 347 348
    auto ipt = ins[0];
    auto* var = ipt == kEmptyVarName ? nullptr : scope_.FindVar(ipt);
    return var != nullptr;
  }

  bool HasOutput(const std::string& name) const override {
    auto& outs = Outputs(name);
    size_t length = outs.size();
    if (length == 0) {
      return false;
    }
F
fengjiayi 已提交
349 350
    PADDLE_ENFORCE_EQ(length, 1UL,
                      "Output %s should not have more than one inputs", name);
351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393
    auto ipt = outs[0];
    auto* var = ipt == kEmptyVarName ? nullptr : scope_.FindVar(ipt);
    return var != nullptr;
  }

  bool HasInputs(const std::string& name) const override {
    auto inputs = op_.Inputs(name);
    if (inputs.empty()) {
      return false;
    }
    for (auto& input : inputs) {
      if (scope_.FindVar(input) == nullptr) {
        return false;
      }
    }
    return true;
  }

  bool HasOutputs(const std::string& name) const override {
    auto outputs = op_.Outputs(name);
    if (outputs.empty()) {
      return false;
    }
    for (auto& output : outputs) {
      if (scope_.FindVar(output) == nullptr) {
        return false;
      }
    }
    return true;
  }

  AttrReader Attrs() const override { return AttrReader(op_.Attrs()); }

  const std::vector<std::string>& Inputs(
      const std::string& name) const override {
    return op_.Inputs(name);
  }

  const std::vector<std::string>& Outputs(
      const std::string& name) const override {
    return op_.Outputs(name);
  }

Q
Qiao Longfei 已提交
394 395 396 397 398 399 400 401 402 403 404 405
  void ShareLoD(const std::string& in, const std::string& out, size_t i = 0,
                size_t j = 0) const override {
    PADDLE_ENFORCE_LT(i, Inputs(in).size());
    PADDLE_ENFORCE_LT(j, Outputs(out).size());
    Variable* in_var = scope_.FindVar(Inputs(in)[i]);
    Variable* out_var = scope_.FindVar(Outputs(out)[j]);
    if (!in_var->IsType<LoDTensor>()) return;
    PADDLE_ENFORCE(out_var->IsType<LoDTensor>(),
                   "The %d-th output of Output(%s) must be LoDTensor.", j, out);
    auto in_tensor = in_var->Get<LoDTensor>();
    auto* out_tensor = out_var->GetMutable<LoDTensor>();
    out_tensor->set_lod(in_tensor.lod());
D
dzhwinter 已提交
406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424

    // TODO(dzhwinter) : reuse ShareLoD in most operators.
    // Need to call ShareLayout explicitly in sequence related ops.
    // Shall we have a better method to shared info between in/out Tensor?
    out_tensor->set_layout(in_tensor.layout());
  }

  void ShareLayout(const std::string& in, const std::string& out, size_t i = 0,
                   size_t j = 0) const {
    PADDLE_ENFORCE_LT(i, Inputs(in).size());
    PADDLE_ENFORCE_LT(j, Outputs(out).size());
    Variable* in_var = scope_.FindVar(Inputs(in)[i]);
    Variable* out_var = scope_.FindVar(Outputs(out)[j]);
    if (!in_var->IsType<LoDTensor>()) return;
    PADDLE_ENFORCE(out_var->IsType<LoDTensor>(),
                   "The %d-th output of Output(%s) must be LoDTensor.", j, out);
    auto in_tensor = in_var->Get<LoDTensor>();
    auto* out_tensor = out_var->GetMutable<LoDTensor>();
    out_tensor->set_layout(in_tensor.layout());
Q
Qiao Longfei 已提交
425 426
  }

427 428 429
  bool IsRuntime() const override { return true; }

 protected:
430 431 432 433 434 435 436
  DDim GetDim(const std::string& name) const override {
    Variable* var = scope_.FindVar(name);
    if (var->IsType<LoDTensor>()) {
      return var->Get<LoDTensor>().dims();
    } else if (var->IsType<SelectedRows>()) {
      return var->Get<SelectedRows>().GetCompleteDims();
    } else {
F
fengjiayi 已提交
437 438 439 440 441 442 443
      PADDLE_THROW(
          "Only LoDTensor/SelectedRows support 'GetDim', but Variable %s's "
          "type_id is %s.",
          name, var->Type().name());
    }
  }

F
fengjiayi 已提交
444
  std::vector<DDim> GetRepeatedDims(const std::string& name) const override {
F
fengjiayi 已提交
445 446 447 448 449
    Variable* var = scope_.FindVar(name);
    if (var->IsType<ReaderHolder>()) {
      return var->Get<ReaderHolder>().shapes();
    } else {
      PADDLE_THROW(
F
fengjiayi 已提交
450
          "Only ReaderHolder support 'GetRepeatedDims', but Variable %s's "
F
fengjiayi 已提交
451 452
          "type_id is %s.",
          name, var->Type().name());
453 454 455 456 457 458 459 460 461 462
    }
  }

  void SetDim(const std::string& name, const DDim& dim) override {
    Variable* var = scope_.FindVar(name);
    if (var->IsType<LoDTensor>()) {
      var->GetMutable<LoDTensor>()->Resize(dim);
    } else if (var->IsType<SelectedRows>()) {
      var->GetMutable<SelectedRows>()->set_height(dim[0]);
    } else {
Y
Yang Yang 已提交
463 464
      PADDLE_THROW("Variable %s type_id %s, expect LoDTensor/SelectedRows.",
                   name, var->Type().name());
465 466 467
    }
  }

F
fengjiayi 已提交
468 469 470 471 472 473 474 475 476 477 478 479 480
  void SetRepeatedDims(const std::string& name,
                       const std::vector<DDim>& dims) override {
    Variable* var = scope_.FindVar(name);
    if (var->IsType<ReaderHolder>()) {
      var->GetMutable<ReaderHolder>()->set_shapes(dims);
    } else {
      PADDLE_THROW(
          "Only ReaderHolder support 'SetRepeatedDims', but Variable %s's "
          "type_id is %s.",
          name, var->Type().name());
    }
  }

481
  proto::VarType::Type GetVarType(const std::string& name) const override {
482 483 484 485
    auto* var = scope_.FindVar(name);
    return ToVarType(var->Type());
  }

F
fengjiayi 已提交
486 487 488 489
  InferShapeVarPtr GetVarPtr(const std::string& name) override {
    return scope_.FindVar(name);
  }

490
 private:
491 492 493 494
  const OperatorBase& op_;
  const Scope& scope_;
};

495 496
void OperatorWithKernel::RunImpl(const Scope& scope,
                                 const platform::Place& place) const {
497 498
  RuntimeInferShapeContext infer_shape_ctx(*this, scope);
  this->InferShape(&infer_shape_ctx);
Y
Yu Yang 已提交
499 500
  platform::DeviceContextPool& pool = platform::DeviceContextPool::Instance();
  auto dev_ctx = pool.Get(place);
501 502
  // profile
  platform::RecordEvent record_event(Type(), dev_ctx);
503 504 505 506
  // check if op[type] has kernel registered.
  auto& all_op_kernels = AllOpKernels();
  auto kernels_iter = all_op_kernels.find(type_);
  if (kernels_iter == all_op_kernels.end()) {
Y
Yu Yang 已提交
507 508
    PADDLE_THROW(
        "There are no kernels which are registered in the %s operator.", type_);
509 510
  }

D
dzhwinter 已提交
511
  ExecutionContext ctx(*this, scope, *dev_ctx);
512

Q
qiaolongfei 已提交
513 514
  OpKernelMap& kernels = kernels_iter->second;

515 516
  // TODO(dzhwinter) : kernel fallback mechanism will be added when all the
  // transform functions are ready.
Q
qiaolongfei 已提交
517

518 519 520 521 522
  // for (auto& candidate : kKernelPriority) {
  //   Do selection
  // }

  auto expected_kernel_key = this->GetExpectedKernelType(ctx);
Q
qiaolongfei 已提交
523 524
  VLOG(3) << "expected_kernel_key:" << expected_kernel_key;

525 526 527 528 529 530 531
  auto kernel_iter = kernels.find(expected_kernel_key);
  if (kernel_iter == kernels.end()) {
    PADDLE_THROW("op %s does not have kernel for %s", type_,
                 KernelTypeToString(expected_kernel_key));
  }

  // do data transform
532 533 534 535 536 537 538 539 540 541
  Scope& new_scope = scope.NewScope();

  for (auto& var_name_item : this->Inputs()) {
    for (auto& var_name : var_name_item.second) {
      auto* var = scope.FindVar(var_name);
      if (var && VarIsTensor(var)) {
        auto* tensor_in = GetTensorFromVar(var);
        if (tensor_in->IsInitialized()) {
          auto kernel_type_for_var = this->GetKernelTypeForVar(
              var_name_item.first, *tensor_in, expected_kernel_key);
542
          if (TransFromNeeded(kernel_type_for_var, expected_kernel_key)) {
543 544 545 546 547 548 549 550
            auto out_var_names = OutputVars(true);
            if (std::find(out_var_names.begin(), out_var_names.end(),
                          var_name) != out_var_names.end()) {
              PADDLE_THROW(
                  "var %s is both input and output, "
                  "does not support transform",
                  var_name);
            }
551 552
            VLOG(3) << "Transform Variable " << var_name << " from "
                    << kernel_type_for_var << " to " << expected_kernel_key;
553
            auto* trans_var = new_scope.Var(var_name);
554 555 556 557
            std::shared_ptr<Tensor> out(new Tensor);
            DataTransform(expected_kernel_key, kernel_type_for_var, *tensor_in,
                          out.get());
            CopyVariableWithTensor(*var, *(out.get()), *trans_var);
558
          }
Q
QI JUN 已提交
559 560
        }
      }
561 562
    }
  }
Q
QI JUN 已提交
563

D
dzhwinter 已提交
564 565 566 567 568
  auto* new_dev_ctx = pool.Get(expected_kernel_key.place_);
  kernel_iter->second->Compute(
      ExecutionContext(*this, new_scope, *new_dev_ctx));

  /*For profiling/benchmark only*/
D
dzhwinter 已提交
569
  if (FLAGS_benchmark) {
D
dzhwinter 已提交
570 571
    new_dev_ctx->Wait();
  }
Q
Qiao Longfei 已提交
572 573
}

574
proto::VarType::Type OperatorWithKernel::IndicateDataType(
Y
Yu Yang 已提交
575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599
    const ExecutionContext& ctx) const {
  auto& scope = ctx.scope();
  int data_type = -1;
  for (auto& input : this->inputs_) {
    for (auto& ipt_name : input.second) {
      auto* var = scope.FindVar(ipt_name);
      if (var != nullptr) {
        const Tensor* t = nullptr;
        if (var->IsType<Tensor>()) {
          t = &var->Get<Tensor>();
        } else if (var->IsType<LoDTensor>()) {
          t = &var->Get<LoDTensor>();
        } else if (var->IsType<SelectedRows>()) {
          t = &(var->Get<SelectedRows>().value());
        }
        if (t != nullptr) {
          int tmp = static_cast<int>(ToDataType(t->type()));
          PADDLE_ENFORCE(tmp == data_type || data_type == -1,
                         "DataType of Paddle Op %s must be the same.", Type());
          data_type = tmp;
        }
      }
    }
  }
  PADDLE_ENFORCE(data_type != -1, "DataType should be indicated by input");
600
  return static_cast<proto::VarType::Type>(data_type);
Y
Yu Yang 已提交
601
}
602

603 604 605 606 607 608 609 610 611 612 613
OpKernelType OperatorWithKernel::GetExpectedKernelType(
    const ExecutionContext& ctx) const {
  return OpKernelType(IndicateDataType(ctx), ctx.GetPlace());
}

OpKernelType OperatorWithKernel::GetKernelTypeForVar(
    const std::string& var_name, const Tensor& tensor,
    const OpKernelType& expected_kernel_type) const {
  return OpKernelType(expected_kernel_type.data_type_, tensor.place());
}

Q
Qiao Longfei 已提交
614
}  // namespace framework
L
liaogang 已提交
615
}  // namespace paddle