Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into fix_grpc_destroy_bug

Dockerfile

@@ -23,7 +23,7 @@ ENV HOME /root
 COPY ./paddle/scripts/docker/root/ /root/
 
 RUN apt-get update && \
-    apt-get install -y --allow-downgrades \
+    apt-get install -y --allow-downgrades patchelf \
     git python-pip python-dev python-opencv openssh-server bison \
     libnccl2=2.1.2-1+cuda8.0 libnccl-dev=2.1.2-1+cuda8.0 \
     wget unzip unrar tar xz-utils bzip2 gzip coreutils ntp \

doc/fluid/api/transpiler.rst

@@ -14,6 +14,15 @@ DistributeTranspiler
     :members:
     :noindex:
 
+.. _api_fluid_transpiler_InferenceTranspiler:
+
+InferenceTranspiler
+-------------------
+
+..  autoclass:: paddle.fluid.transpiler.InferenceTranspiler
+    :members:
+    :noindex:
+
 .. _api_fluid_transpiler_memory_optimize:
 
 memory_optimize

paddle/fluid/framework/details/build_strategy.h

@@ -34,7 +34,7 @@ struct BuildStrategy {
 
   std::string debug_graphviz_path_{""};
 
-  bool enable_data_balance_{true};
+  bool enable_data_balance_{false};
 };
 
 }  // namespace details

paddle/fluid/framework/details/data_balance_op_handle.cc

@@ -86,9 +86,9 @@ std::vector<std::array<int, 3>> DataBalanceOpHandle::GetBalancePlan(
 }
 
 void DataBalanceOpHandle::RunImpl() {
-  if (places_.size() == 1) {
-    return;
-  }
+  PADDLE_ENFORCE_GT(places_.size(), 1,
+                    "Data balance can only be enabled when the number of "
+                    "places to run larger than 1.");
   auto in_var_handles = DynamicCast<VarHandle>(inputs_);
   auto out_var_handles = DynamicCast<VarHandle>(outputs_);
   PADDLE_ENFORCE(in_var_handles.size() % places_.size() == 0);

paddle/fluid/framework/details/multi_devices_graph_builder.cc

@@ -59,6 +59,11 @@ MultiDevSSAGraphBuilder::MultiDevSSAGraphBuilder(
     grad_names_.insert(GradVarName(p));
   }
   balance_vars_.resize(places_.size(), 0);
+  if (strategy_.enable_data_balance_ && places_.size() == 1) {
+    LOG(WARNING) << "It is no need to enable data balance when there is only "
+                    "one place. enable_data_balance is set to False.";
+    strategy_.enable_data_balance_ = false;
+  }
 }
 
 void MultiDevSSAGraphBuilder::CreateOpHandleIOs(SSAGraph *result,

paddle/fluid/framework/op_registry.h

@@ -186,13 +186,7 @@ struct OpKernelRegistrarFunctorEx<PlaceType, false, I,
   STATIC_ASSERT_GLOBAL_NAMESPACE(                                        \
       __reg_op__##op_type,                                               \
       "REGISTER_OPERATOR must be called in global namespace");           \
-  class _OpClass_##op_type##_ : public op_class {                      \
-   public:                                                             \
-    DEFINE_OP_CLONE_METHOD(_OpClass_##op_type##_);                     \
-    DEFINE_OP_CONSTRUCTOR(_OpClass_##op_type##_, op_class);            \
-  };                                                                   \
-  static ::paddle::framework::OperatorRegistrar<_OpClass_##op_type##_, \
-                                                ##__VA_ARGS__>         \
+  static ::paddle::framework::OperatorRegistrar<op_class, ##__VA_ARGS__> \
       __op_registrar_##op_type##__(#op_type);                            \
   int TouchOpRegistrar_##op_type() {                                     \
     __op_registrar_##op_type##__.Touch();                                \

paddle/fluid/framework/op_registry_test.cc

@@ -193,15 +193,10 @@ TEST(OpRegistry, CustomChecker) {
   ASSERT_EQ(test_attr, 4);
 }
 
-class CosineOpComplete : public paddle::framework::CosineOp {
- public:
-  DEFINE_OP_CONSTRUCTOR(CosineOpComplete, paddle::framework::CosineOp);
-  DEFINE_OP_CLONE_METHOD(CosineOpComplete);
-};
-
 TEST(OperatorRegistrar, Test) {
   paddle::framework::OperatorRegistrar<
-      CosineOpComplete, paddle::framework::CosineOpProtoAndCheckerMaker>
+      paddle::framework::CosineOp,
+      paddle::framework::CosineOpProtoAndCheckerMaker>
       reg("cos");
 }
 

paddle/fluid/framework/operator.h

@@ -121,10 +121,6 @@ class OperatorBase {
   //! Get all outputs variable names
   virtual std::vector<std::string> OutputVars(bool has_intermediate) const;
 
-  // Return a new operator instance, which is as same as this.
-  // Use unique_ptr to prevent caller forget to delete this pointer.
-  virtual std::unique_ptr<OperatorBase> Clone() const = 0;
-
  protected:
   std::string type_;
   // NOTE: in case of OpGrad, inputs_ contains:
@@ -145,37 +141,6 @@ class OperatorBase {
                        const platform::Place& place) const = 0;
 };
 
-// Macro for define a clone method.
-// If you are writing an kernel operator, `Clone` will be defined when you
-// register it. i.e. `Clone` method is not needed to define by yourself.
-#define DEFINE_OP_CLONE_METHOD(cls)                                            \
-  std::unique_ptr<::paddle::framework::OperatorBase> Clone() const final {     \
-    return std::unique_ptr<::paddle::framework::OperatorBase>(new cls(*this)); \
-  }
-
-// Macro for define a default constructor for Operator.
-// You can also use
-//   using PARENT_CLASS::PARENT_CLASS;
-// to use parent's constructor.
-#define DEFINE_OP_CONSTRUCTOR(cls, parent_cls)             \
-  cls(const std::string& type,                             \
-      const ::paddle::framework::VariableNameMap& inputs,  \
-      const ::paddle::framework::VariableNameMap& outputs, \
-      const paddle::framework::AttributeMap& attrs)        \
-      : parent_cls(type, inputs, outputs, attrs) {}
-
-class NOP : public OperatorBase {
- public:
-  using OperatorBase::OperatorBase;
-  std::unique_ptr<OperatorBase> Clone() const override {
-    return std::unique_ptr<OperatorBase>(new NOP(*this));
-  }
-
- private:
-  void RunImpl(const Scope& scope,
-               const platform::Place& place) const override {}
-};
-
 class ExecutionContext {
  public:
   ExecutionContext(const OperatorBase& op, const Scope& scope,

paddle/fluid/framework/operator_test.cc

@@ -247,26 +247,3 @@ TEST(OpKernel, multi_inputs) {
   auto op = paddle::framework::OpRegistry::CreateOp(op_desc);
   op->Run(scope, cpu_place);
 }
-
-class OperatorClone : public paddle::framework::OperatorBase {
- public:
-  DEFINE_OP_CLONE_METHOD(OperatorClone);
-  OperatorClone(const std::string& type,
-                const paddle::framework::VariableNameMap& inputs,
-                const paddle::framework::VariableNameMap& outputs,
-                const paddle::framework::AttributeMap& attrs)
-      : OperatorBase(type, inputs, outputs, attrs) {}
-
- private:
-  void RunImpl(const paddle::framework::Scope& scope,
-               const paddle::platform::Place& place) const override {}
-};
-
-TEST(Operator, Clone) {
-  paddle::framework::InitDevices(true);
-  OperatorClone a("ABC", paddle::framework::VariableNameMap{},
-                  paddle::framework::VariableNameMap{},
-                  paddle::framework::AttributeMap{});
-  auto b = a.Clone();
-  ASSERT_EQ(a.Type(), b->Type());
-}

paddle/fluid/framework/var_type_inference_test.cc

@@ -22,6 +22,17 @@ limitations under the License. */
 namespace paddle {
 namespace framework {
 
+class NOP : public OperatorBase {
+ public:
+  NOP(const std::string &type, const VariableNameMap &inputs,
+      const VariableNameMap &outputs, const AttributeMap &attrs)
+      : OperatorBase(type, inputs, outputs, attrs) {}
+
+ private:
+  void RunImpl(const Scope &scope,
+               const platform::Place &place) const override {}
+};
+
 class SumOpMaker : public OpProtoAndCheckerMaker {
  public:
   void Make() {

paddle/fluid/operators/read_op.cc

@@ -92,9 +92,13 @@ class ReadOpMaker : public framework::OpProtoAndCheckerMaker {
   void Make() override {
     AddInput("Reader", "(ReaderHolder) The executed reader.");
     AddOutput("Out", "(LoDTensor) The output data.").AsDuplicable();
-    AddAttr<bool>("throw_eof_exp",
+    AddAttr<bool>(
+        "throw_eof_exp",
         "If set true, an exception will be thrown when the Reader "
-                  "yields empty (which means there is no next data).")
+        "yields empty (which means there is no next data).\n"
+        "NOTES: This flag must be true always. It will be set to false"
+        " only when the data-balance is enabled in ParallelExecutor"
+        " and it is set by ParallelExecutor instance, not users.")
         .SetDefault(true);
     AddComment(R"DOC(
       Read Operator

paddle/legacy/gserver/gradientmachines/MultiGradientMachine.cpp

@@ -532,6 +532,7 @@ void TrainerThread::computeThread() {
         break;
     }
   }
+  hl_fini();
 }
 
 void TrainerThread::prefetch() {
@@ -651,6 +652,7 @@ void TrainerThread::copyGradToBufferThread() {
     }
     partnerThread->notifyGradientCollect(pid);
   }
+  hl_fini();
 }
 
 void TrainerThread::gradCollectThread() {
@@ -693,6 +695,7 @@ void TrainerThread::gradCollectThread() {
       notifyCopyGradToBuffer(pid);
     }
   }
+  hl_fini();
 }
 
 void TrainerThread::doCallback(int pid) {
@@ -741,6 +744,7 @@ void TrainerThread::valueDispatchThread() {
 
     thread->notifyValueReady(pid);
   }
+  hl_fini();
 }
 
 void TrainerThread::notifyValueReady(int paramId) {

paddle/legacy/gserver/gradientmachines/ParallelNeuralNetwork.cpp

@@ -197,6 +197,7 @@ void ParallelThread::computeThread() {
       job_work.layer_->markAllInputGrad();
     }
   }
+  hl_fini();
 }
 
 void ParallelThread::start() {

python/paddle/fluid/tests/unittests/test_data_balance.py

@@ -103,8 +103,12 @@ class TestDataBalance(unittest.TestCase):
             exe = fluid.Executor(place)
             exe.run(startup_prog)
 
+            build_strategy = fluid.BuildStrategy()
+            build_strategy.enable_data_balance = True
             parallel_exe = fluid.ParallelExecutor(
-                use_cuda=self.use_cuda, main_program=main_prog)
+                use_cuda=self.use_cuda,
+                main_program=main_prog,
+                build_strategy=build_strategy)
 
             if (parallel_exe.device_count > self.batch_size):
                 print("WARNING: Unittest TestDataBalance skipped. \
@@ -145,9 +149,12 @@ class TestDataBalance(unittest.TestCase):
             place = fluid.CUDAPlace(0) if self.use_cuda else fluid.CPUPlace()
             exe = fluid.Executor(place)
             exe.run(startup_prog)
-
+            build_strategy = fluid.BuildStrategy()
+            build_strategy.enable_data_balance = True
             parallel_exe = fluid.ParallelExecutor(
-                use_cuda=self.use_cuda, main_program=main_prog)
+                use_cuda=self.use_cuda,
+                main_program=main_prog,
+                build_strategy=build_strategy)
 
             if (parallel_exe.device_count > self.batch_size):
                 print("WARNING: Unittest TestDataBalance skipped. \

python/paddle/fluid/transpiler/inference_transpiler.py

@@ -19,7 +19,7 @@ from ..framework import Program
 from ..executor import global_scope
 
 
-class InferenceTranspiler:
+class InferenceTranspiler(object):
     '''
     Convert the fluid program to optimized inference program.
 

python/paddle/libs/__init__.py

+# 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.
+
+# used for setup.py.in to store the thirdparty shared libraries

python/setup.py.in

 from setuptools import setup, Distribution, Extension
 import subprocess
+import shutil
+import os
 class BinaryDistribution(Distribution):
     def has_ext_modules(foo):
         return True
@@ -62,6 +64,7 @@ write_version_py(filename='@PADDLE_BINARY_DIR@/python/paddle/version.py')
 
 
 packages=['paddle',
+          'paddle.libs',
           'paddle.utils',
           'paddle.dataset',
           'paddle.reader',
@@ -114,11 +117,34 @@ package_dir={
 if '${WITH_FLUID_ONLY}'== 'OFF':
     package_dir['py_paddle']='${PADDLE_BINARY_DIR}/python/py_paddle'
 
-
-paddle_rt_lib_dir = 'lib'
-paddle_rt_libs = ['${WARPCTC_LIBRARIES}']
-if '${MKL_SHARED_LIBS}'!= '':
-  paddle_rt_libs += '${MKL_SHARED_LIBS}'.split(';')
+# put all thirdparty libraries in paddle.libs
+package_data['paddle.libs']=['libwarpctc.so']
+libs_path='${PADDLE_BINARY_DIR}/python/paddle/libs'
+shutil.copy('${WARPCTC_LIBRARIES}', libs_path)
+if '${WITH_MKL}' == 'ON':
+    shutil.copy('${MKLML_LIB}', libs_path)
+    shutil.copy('${MKLML_IOMP_LIB}', libs_path)
+    package_data['paddle.libs']+=['libmklml_intel.so','libiomp5.so']
+if '${WITH_MKLDNN}' == 'ON':
+    # change rpath of libmkldnn.so.0, add $ORIGIN/ to it.
+    # The reason is that all thirdparty libraries in the same directory,
+    # thus, libmkldnn.so.0 will find libmklml_intel.so and libiomp5.so.
+    command = "patchelf --set-rpath '$ORIGIN/' ${MKLDNN_SHARED_LIB}"
+    if os.system(command) != 0:
+        raise Exception("patchelf --set-rpath for libmkldnn.so.0 fails")
+    package_data['paddle.libs']+=['libmkldnn.so.0']
+    shutil.copy('${MKLDNN_SHARED_LIB}', libs_path)
+# remove unused paddle/libs/__init__.py
+os.remove(libs_path+'/__init__.py')
+package_dir['paddle.libs']=libs_path
+
+# change rpath of core.so, add $ORIGIN/../libs/ to it.
+# The reason is that libwarpctc.so, libiomp5.so etc are in paddle.libs, and
+# core.so is in paddle.fluid, thus paddle/fluid/../libs will pointer to above libraries.
+# This operation will fix https://github.com/PaddlePaddle/Paddle/issues/3213
+command = "patchelf --set-rpath '$ORIGIN/../libs/' ${PADDLE_BINARY_DIR}/python/paddle/fluid/core.so"
+if os.system(command) != 0:
+    raise Exception("patchelf --set-rpath for core.so fails")
 
 setup(name='${PACKAGE_NAME}',
       version='${PADDLE_VERSION}',
@@ -128,6 +154,5 @@ setup(name='${PACKAGE_NAME}',
       ext_modules=[Extension('_foo', ['stub.cc'])],
       package_data=package_data,
       package_dir=package_dir,
-      scripts=paddle_bins,
-      data_files=[(paddle_rt_lib_dir, paddle_rt_libs)]
+      scripts=paddle_bins
 )