optimization_pass_runner.cc

/* Copyright 2018 The TensorFlow 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.
==============================================================================*/
// This file creates a library that can run any registered optimization pass.
// The binary that uses this will be run in a form similar to:
// ./optimization_pass_runner  --input_file_path=/tmp/input.pbtxt
// --output_file_path=/tmp/output.pbtxt
// --optimization_pass=NameOfGraphOptimizationPass
#include "tensorflow/tools/optimization/optimization_pass_runner.h"

#include <memory>
#include <string>
#include <vector>

#include "tensorflow/core/common_runtime/device.h"
#include "tensorflow/core/common_runtime/device_set.h"
#include "tensorflow/core/common_runtime/graph_constructor.h"
#include "tensorflow/core/common_runtime/optimization_registry.h"
#include "tensorflow/core/framework/allocator.h"
#include "tensorflow/core/framework/device_attributes.pb.h"
#include "tensorflow/core/framework/function.h"
#include "tensorflow/core/framework/function.pb.h"
#include "tensorflow/core/framework/graph.pb.h"
#include "tensorflow/core/framework/op.h"
#include "tensorflow/core/framework/types.h"
#include "tensorflow/core/graph/graph.h"
#include "tensorflow/core/lib/core/errors.h"
#include "tensorflow/core/lib/core/status.h"
#include "tensorflow/core/platform/env.h"
#include "tensorflow/core/protobuf/config.pb.h"
#include "tensorflow/core/public/session_options.h"

namespace tensorflow {
namespace {
// A fake device used to populate a DeviceSet.
class FakeDevice : public Device {
 private:
  explicit FakeDevice(const DeviceAttributes& device_attributes)
      : Device(nullptr, device_attributes) {}

 public:
  Status Sync() override;
  static std::unique_ptr<Device> Make(const string& name, const string& type);
};

Status FakeDevice::Sync() {
  return errors::Unimplemented("FakeDevice::Sync()");
}

std::unique_ptr<Device> FakeDevice::Make(const string& name,
                                         const string& type) {
  DeviceAttributes device_attributes;
  device_attributes.set_name(name);
  device_attributes.set_device_type(DeviceType(type).type());
  return std::unique_ptr<Device>(new FakeDevice(device_attributes));
}

Status FindPassWithName(absl::string_view name,
                        GraphOptimizationPass** result) {
  *result = nullptr;
  // Run the optimization pass specified by the command line flag.
  for (const auto& groups_and_passes :
       OptimizationPassRegistry::Global()->groups()) {
    for (const auto& phase_and_passes : groups_and_passes.second) {
      for (const auto& pass : phase_and_passes.second) {
        if (pass->name() == name) {
          if (*result) {
            return errors::Internal("Found more than one pass with name ",
                                    name);
          }
          *result = pass.get();
        }
      }
    }
  }

  return *result == nullptr
             ? errors::Internal("Could not find pass with name ", name)
             : OkStatus();
}
}  // namespace

Status OptimizationPassRunner::Run(absl::string_view pass_to_run,
                                   GraphDef input, GraphDef* result) {
  auto session_options = absl::make_unique<SessionOptions>();
  session_options->config.mutable_graph_options()
      ->mutable_optimizer_options()
      ->set_global_jit_level(jit_level_);
  FunctionDefLibrary flib;
  std::unique_ptr<Graph> graph = absl::make_unique<Graph>(OpRegistry::Global());

  GraphOptimizationPassOptions options;
  options.session_options = session_options.get();
  options.graph = &graph;
  std::unique_ptr<FunctionLibraryDefinition> flib_def(
      new FunctionLibraryDefinition((*options.graph)->op_registry(), flib));
  options.flib_def = flib_def.get();

  // Grab the data
  GraphConstructorOptions graph_opts;
  graph_opts.expect_device_spec = true;
  graph_opts.allow_internal_ops = true;
  TF_RETURN_IF_ERROR(ConvertGraphDefToGraph(graph_opts, std::move(input),
                                            options.graph->get()));

  // Add all devices that were previously configured with AddDevice.
  DeviceSet device_set;
  for (auto& device : devices_) {
    device_set.AddDevice(device.get());
  }
  options.device_set = &device_set;

  GraphOptimizationPass* pass;
  TF_RETURN_IF_ERROR(FindPassWithName(pass_to_run, &pass));
  TF_RETURN_IF_ERROR(pass->Run(options));

  options.graph->get()->ToGraphDef(result);
  return OkStatus();
}

Status OptimizationPassRunner::SetJitLevel(
    OptimizerOptions::GlobalJitLevel jit_level) {
  jit_level_ = jit_level;
  return OkStatus();
}

Status OptimizationPassRunner::AddDevices(absl::string_view type, int count) {
  for (int i = 0; i < count; i++) {
    devices_.push_back(FakeDevice::Make(
        absl::StrCat("/job:localhost/replica:0/task:0/device:", type, ":", i),
        absl::StrCat(type)));
    devices_.push_back(FakeDevice::Make(
        absl::StrCat("/job:localhost/replica:0/task:0/device:XLA_", type, ":",
                     i),
        absl::StrCat(type)));
  }

  return OkStatus();
}
}  // namespace tensorflow