graph.cc 5.7 KB
Newer Older
X
Xin Pan 已提交
1
/* Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
X
start  
Xin Pan 已提交
2 3 4 5 6 7 8 9 10 11 12 13 14

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. */

X
Xin Pan 已提交
15
#include <algorithm>
Z
Zhen Wang 已提交
16
#include <unordered_map>
X
Xin Pan 已提交
17

X
start  
Xin Pan 已提交
18
#include "paddle/fluid/framework/ir/graph.h"
X
Xin Pan 已提交
19
#include "paddle/fluid/framework/op_proto_maker.h"
20 21
#include "paddle/fluid/framework/program_desc.h"
#include "paddle/fluid/framework/var_desc.h"
X
start  
Xin Pan 已提交
22 23

namespace paddle {
X
Xin Pan 已提交
24
namespace framework {
X
Xin Pan 已提交
25
namespace ir {
X
Xin Pan 已提交
26

X
clean  
Xin Pan 已提交
27
Graph::Graph(const ProgramDesc &program) : program_(program) {
28 29 30
  auto var_nodes = InitFromProgram(program_);
  ResolveHazard(var_nodes);
}
31

32 33
std::map<std::string, std::vector<ir::Node *>> Graph::InitFromProgram(
    const ProgramDesc &program) {
M
minqiyang 已提交
34
  VLOG(3) << "block in program:" << program_.Size();
35
  std::unordered_map<std::string, VarDesc *> all_vars;
36 37
  // var nodes for each var name, will have multiple versions in SSA
  std::map<std::string, std::vector<ir::Node *>> var_nodes;
38 39 40 41 42
  for (auto *var : program.Block(0).AllVars()) {
    all_vars.emplace(var->Name(), var);
  }

  for (auto *op : program.Block(0).AllOps()) {
X
clean  
Xin Pan 已提交
43
    ir::Node *node = CreateOpNode(op);
X
Xin Pan 已提交
44 45
    // For input args, reuse the same var name if it was created before.
    // Otherwise, create a new one.
46 47
    for (auto &each_var_name : op->InputArgumentNames()) {
      ir::Node *var = nullptr;
X
Xin Pan 已提交
48
      if (var_nodes.find(each_var_name) != var_nodes.end()) {
X
Xin Pan 已提交
49
        var = var_nodes.at(each_var_name).back();
X
Xin Pan 已提交
50
      } else if (all_vars.count(each_var_name) != 0) {
X
clean  
Xin Pan 已提交
51
        var = CreateVarNode(all_vars.at(each_var_name));
X
Xin Pan 已提交
52
        var_nodes[each_var_name].push_back(var);
53
      } else {
X
Xin Pan 已提交
54 55 56
        // Operation input var can be optional (dispensable). Which means
        // the operation doesn't really need the var at runtime. In this
        // case, the no-existed var is ready at the beginning.
X
polish  
Xin Pan 已提交
57
        var = CreateEmptyNode(each_var_name, ir::Node::Type::kVariable);
X
Xin Pan 已提交
58
        var_nodes[each_var_name].push_back(var);
59 60 61 62
      }
      node->inputs.push_back(var);
      var->outputs.push_back(node);
    }
X
Xin Pan 已提交
63
    // For output args, always create a new var.
64
    for (auto &each_var_name : op->OutputArgumentNames()) {
X
Xin Pan 已提交
65 66 67 68 69 70 71 72 73
      ir::Node *var = nullptr;
      if (all_vars.count(each_var_name) != 0) {
        var = CreateVarNode(all_vars.at(each_var_name));
      } else {
        // Operation output vars can be @EMPTY@. For example, while_grad
        // can have multi @EMPTY@ outputs with no VarDesc.
        // TODO(panyx0718): Add a test.
        var = CreateEmptyNode(each_var_name, ir::Node::Type::kVariable);
      }
X
Xin Pan 已提交
74
      var_nodes[each_var_name].push_back(var);
75 76 77 78
      node->outputs.push_back(var);
      var->inputs.push_back(node);
    }
  }
X
polish  
Xin Pan 已提交
79
  Set<const std::vector<OpDesc *>>(
X
Xin Pan 已提交
80
      details::kStaleProgramOpDescs,
X
polish  
Xin Pan 已提交
81
      new std::vector<OpDesc *>(program.Block(0).AllOps()));
G
Gabor Buella 已提交
82
  return var_nodes;
83
}
X
Xin Pan 已提交
84

85 86
void Graph::ResolveHazard(
    const std::map<std::string, std::vector<ir::Node *>> &var_nodes) {
X
polish  
Xin Pan 已提交
87
  /**
88 89 90 91 92
   * We should handle write after read(WAR) and write after write(WAW) here.
   * Because some of the operators of the program can be executed parallelly.
   * So, to make the program running in the right order, we should add the
   * dependence of WAR and WAW.
   *
X
polish  
Xin Pan 已提交
93 94 95
   *
   * https://en.wikipedia.org/wiki/Hazard_(computer_architecture)#Write_after_read_(WAR)
   */
X
Xin Pan 已提交
96

X
Xin Pan 已提交
97 98 99 100 101 102 103 104
  for (auto &var : var_nodes) {
    auto &versions = var.second;
    if (versions.size() <= 1) continue;

    auto it_new = versions.rbegin();
    auto it_old = versions.rbegin();
    ++it_old;
    for (; it_old != versions.rend(); it_new = it_old, ++it_old) {
M
minqiyang 已提交
105
      VLOG(3) << "deal with var: " << (*it_new)->Name();
X
Xin Pan 已提交
106 107 108 109
      ir::Node *write_op =
          (*it_new)->inputs.empty() ? nullptr : (*it_new)->inputs[0];
      const auto &read_ops = (*it_old)->outputs;

D
dzhwinter 已提交
110 111 112 113
      PADDLE_ENFORCE(
          write_op,
          string::Sprintf("The write_op of var %s should not be empty.",
                          (*it_new)->Name()));
114 115 116 117

      // Add write after write dependence
      ir::Node *upstream_op =
          (*it_old)->inputs.empty() ? nullptr : (*it_old)->inputs[0];
X
Xin Pan 已提交
118 119
      // TODO(zcd): Add a test.
      if (upstream_op && upstream_op != write_op) {
120 121 122 123 124 125 126
        ir::Node *dep_var = CreateControlDepVar();
        write_op->inputs.push_back(dep_var);
        upstream_op->outputs.push_back(dep_var);
        dep_var->outputs.push_back(write_op);
        dep_var->inputs.push_back(upstream_op);
      }

X
Xin Pan 已提交
127 128 129 130 131 132
      for (auto *read_op : read_ops) {
        // Manually add a dependency var from read_op to write_op;
        if (read_op == write_op) {
          // Read Write is the same op.
          continue;
        }
X
Xin Pan 已提交
133 134 135 136 137 138 139 140 141 142 143
        // 2 ops might have been connected via other vars.
        bool has_dep = false;
        for (ir::Node *r_out : read_op->outputs) {
          for (ir::Node *w_in : write_op->inputs) {
            if (r_out == w_in) {
              has_dep = true;
              break;
            }
          }
        }
        if (has_dep) continue;
X
Xin Pan 已提交
144

X
Xin Pan 已提交
145
        ir::Node *dep_var = CreateControlDepVar();
X
Xin Pan 已提交
146 147 148 149 150 151 152
        read_op->outputs.push_back(dep_var);
        dep_var->inputs.push_back(read_op);
        write_op->inputs.push_back(dep_var);
        dep_var->outputs.push_back(write_op);
      }
    }
  }
X
better  
Xin Pan 已提交
153
}
X
Xin Pan 已提交
154 155 156 157

bool IsControlDepVar(const ir::Node &var) {
  return var.Name().find(ir::Node::kControlDepVarName) != std::string::npos;
}
X
Xin Pan 已提交
158
}  // namespace ir
X
Xin Pan 已提交
159
}  // namespace framework
X
start  
Xin Pan 已提交
160
}  // namespace paddle