ubuntu16.04下安装caffe以及错误汇总

 这篇文章主要讲安装和编译caffe过程以及遇到的问题，还有解决方法。关于如何安装nvidia驱动，cuda，cudnn，以及opencv3 等网上已经有足够多的教程了，可以直接在网上找。

1. 系统配置：

OS: Ubuntu16.04
python Version: python3
CUDA Version: 10.0

2. caffe安装

2.1  安装依赖库：

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev

2.2 下载caffe源码

下载地址：git clone https://github.com/BVLC/caffe.git

 

3. 修改文件

3.1 修改Makefile.config文件

• 进入caffe源码主目录，拷贝一个Makefile.config作为编译配置选项

cp Makefile.config.example Makefile.config

• 修改编译文件Makefile.config 

# USE_CUDNN := 1  
# USE_OPENCV := 0
# OPENCV_VERSION := 3
# WITH_PYTHON_LAYER := 1

# 改为： 

USE_CUDNN := 1
USE_OPENCV := 1
OPENCV_VERSION := 3
WITH_PYTHON_LAYER := 1

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib

改为：

INCLUDE_DIRS := $(PYTHON_INCLUDE) /usr/local/include /usr/include/hdf5/serial
LIBRARY_DIRS := $(PYTHON_LIB) /usr/local/lib /usr/lib /usr/lib/x86_64-linux-gnu /usr/lib/x86_64-linux-gnu/hdf5/serial

使用python3.5版本，把python2.7代码加注释，python3.5前的注释去掉 

#PYTHON_INCLUDE := /usr/include/python2.7 \
                  /usr/lib/python2.7/dist-packages/numpy/core/include

PYTHON_INCLUDE := /usr/include/python3.5m \
                 /usr/lib/python3.5/dist-packages/numpy/core/include

3.2 修改Makefile文件

NVCCFLAGS += -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)
改为：
NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)


LIBRARIES += glog gflags protobuf boost_system boost_filesystem m
改为：
LIBRARIES += glog gflags protobuf boost_system boost_filesystem m hdf5_serial_hl hdf5_serial

4. 编译

make all -j8

 

遇到错误：

1. 编译选项问题 

/usr/include/c++/4.9/bits/c++0x_warning.h:32:2: error: #error This file requires compiler and library support for the ISO C++ 2011 standard. This support is currently experimental, and must be enabled with the -std=c++11 or -std=gnu++11 compiler options.
 #error This file requires compiler and library support for the 

......

/usr/local/include/google/protobuf/arena.h:324:36: note:   template argument deduction/substitution failed:
Makefile:600: recipe for target '.build_release/src/caffe/proto/caffe.pb.o' failed
make: *** [.build_release/src/caffe/proto/caffe.pb.o] Error 1

解决方法：请看这位大神是如何操作的https://blog.csdn.net/pkwcxy/article/details/82351669?utm_source=blogxgwz8

在Makefile文件中找到：

##############################
# Configure build
##############################

找到：

# Linux

添加：CXXFLAGS += -std=c++11

安装上面的方法一开始没报错，但过了一段时间后还是报同样的错误，再进一步如下修改，打开Makefile文件，给CXXFLAGS、NVCCFLAGS、LINKFLAGS都使用-std=c++11标准编译：

CXXFLAGS += -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS) 
改为：
CXXFLAGS += -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS) -std=c++11 


NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)
改为：
NVCCFLAGS += -D_FORCE_INLINES -ccbin=$(CXX) -Xcompiler -fPIC $(COMMON_FLAGS)  -std=c++11


LINKFLAGS += -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS)
改为：
LINKFLAGS += -pthread -fPIC $(COMMON_FLAGS) $(WARNINGS) -std=c++11

 

2. CUDA版本问题

重新编译还是有错误：

NVCC src/caffe/solvers/adagrad_solver.cu
nvcc fatal   : Unsupported gpu architecture 'compute_20'
NVCC src/caffe/solvers/sgd_solver.cu
nvcc fatal   : Unsupported gpu architecture 'compute_20'
Makefile:607: recipe for target '.build_release/cuda/src/caffe/solvers/rmsprop_solver.o' failed

原因是不支持GPU架构'compute_20'，在Makefile.config里面的找到

CUDA_ARCH := -gencode arch=compute_20,code=sm_20
                  -gencode arch=compute_20,code=sm_21 \
                  -gencode arch=compute_30,code=sm_30 \
                  -gencode arch=compute_35,code=sm_35 \
                  -gencode arch=compute_50,code=sm_50 \
                  -gencode arch=compute_52,code=sm_52 \
                  -gencode arch=compute_60,code=sm_60 \
                  -gencode arch=compute_61,code=sm_61 \
                  -gencode arch=compute_61,code=compute_61

把前面compute_20相关的那2行删除
-gencode arch=compute_20,code=sm_20
-gencode arch=compute_20,code=sm_21
 

3. g++版本问题

.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
collect2: error: ld returned 1 exit status
Makefile:637: recipe for target '.build_release/tools/upgrade_solver_proto_text.bin' failed
make: *** [.build_release/tools/upgrade_solver_proto_text.bin] Error 1
make: *** 正在等待未完成的任务....
......

这个g++版本问题，解决方法这里参考：https://blog.csdn.net/chenshuibiao/article/details/78734957

先查看g++版本：

gcc --version     # 或 g++ --version

gcc (Ubuntu 4.9.3-13ubuntu2) 4.9.3
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

发现我的是4.9.3版本，需要升级gcc版本。查看我Ubuntu系统发现里面已经存在更高级的5版本，就不用再安装了，直接强制软链接一下即可：

sudo ln -s /usr/bin/gcc-5 /usr/bin/gcc -f
sudo ln -s /usr/bin/g++-5 /usr/bin/g++ -f

再次查看g++版本发现已经变成5.4.0了

 g++ (Ubuntu 5.4.0-6ubuntu1~16.04.11) 5.4.0 20160609
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

再次编译caffe源码，发现没问题了。

4. python版本问题

如果在Makefile.config文件里面Python相关的配置有问题，可能会遇到以下的问题： 

CXX/LD -o .build_release/tools/upgrade_net_proto_text.bin
CXX/LD -o .build_release/examples/siamese/convert_mnist_siamese_data.bin
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/tools/convert_imageset.o：在函数‘std::string* google::MakeCheckOpString(unsigned long const&, int const&, char const*)’中：
convert_imageset.cpp:(.text._ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/convert_imageset.o：在函数‘main’中：
convert_imageset.cpp:(.text.startup+0x347)：对‘google::SetUsageMessage(std::string const&)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/tools/caffe.o：在函数‘std::string* google::MakeCheckOpString(cudaError const&, cudaError const&, char const*)’中：
caffe.cpp:(.text._ZN6google17MakeCheckOpStringI9cudaErrorS1_EEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringI9cudaErrorS1_EEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/caffe.o：在函数‘std::string* google::MakeCheckOpString(unsigned long const&, int const&, char const*)’中：
caffe.cpp:(.text._ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/caffe.o：在函数‘main’中：
caffe.cpp:(.text.startup+0x53)：对‘google::SetVersionString(std::string const&)’未定义的引用
caffe.cpp:(.text.startup+0x86)：对‘google::SetUsageMessage(std::string const&)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/tools/extract_features.o：在函数‘std::string* google::MakeCheckOpString(int const&, int const&, char const*)’中：
extract_features.cpp:(.text._ZN6google17MakeCheckOpStringIiiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringIiiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/extract_features.o：在函数‘std::string* google::MakeCheckOpString(unsigned long const&, unsigned long const&, char const*)’中：
extract_features.cpp:(.text._ZN6google17MakeCheckOpStringImmEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringImmEEPSsRKT_RKT0_PKc]+0x51)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/tools/compute_image_mean.o：在函数‘std::string* google::MakeCheckOpString(int const&, int const&, char const*)’中：
compute_image_mean.cpp:(.text._ZN6google17MakeCheckOpStringIiiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringIiiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/compute_image_mean.o：在函数‘std::string* google::MakeCheckOpString(unsigned long const&, int const&, char const*)’中：
compute_image_mean.cpp:(.text._ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringImiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/tools/compute_image_mean.o：在函数‘main’中：
compute_image_mean.cpp:(.text.startup+0x172)：对‘google::SetUsageMessage(std::string const&)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/lib/libcaffe.so：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
.build_release/examples/siamese/convert_mnist_siamese_data.o：在函数‘convert_dataset(char const*, char const*, char const*)’中：
convert_mnist_siamese_data.cpp:(.text+0x4cf)：对‘leveldb::DB::Open(leveldb::Options const&, std::string const&, leveldb::DB**)’未定义的引用
.build_release/examples/siamese/convert_mnist_siamese_data.o：在函数‘std::string* google::MakeCheckOpString(unsigned int const&, int const&, char const*)’中：
convert_mnist_siamese_data.cpp:(.text._ZN6google17MakeCheckOpStringIjiEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringIjiEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/examples/siamese/convert_mnist_siamese_data.o：在函数‘std::string* google::MakeCheckOpString(unsigned int const&, unsigned int const&, char const*)’中：
convert_mnist_siamese_data.cpp:(.text._ZN6google17MakeCheckOpStringIjjEEPSsRKT_RKT0_PKc[_ZN6google17MakeCheckOpStringIjjEEPSsRKT_RKT0_PKc]+0x50)：对‘google::base::CheckOpMessageBuilder::NewString()’未定义的引用
.build_release/lib/libcaffe.so：对‘leveldb::Status::ToString() const’未定义的引用
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
Makefile:637: recipe for target '.build_release/tools/compute_image_mean.bin' failed
collect2: error: ld returned 1 exit status
collect2: error: ld returned 1 exit status
make: *** [.build_release/tools/compute_image_mean.bin] Error 1
make: *** 正在等待未完成的任务....
Makefile:637: recipe for target '.build_release/tools/caffe.bin' failed
make: *** [.build_release/tools/caffe.bin] Error 1
Makefile:637: recipe for target '.build_release/tools/extract_features.bin' failed
make: *** [.build_release/tools/extract_features.bin] Error 1
Makefile:637: recipe for target '.build_release/tools/upgrade_solver_proto_text.bin' failed
make: *** [.build_release/tools/upgrade_solver_proto_text.bin] Error 1
Makefile:637: recipe for target '.build_release/tools/upgrade_net_proto_binary.bin' failed
make: *** [.build_release/tools/upgrade_net_proto_binary.bin] Error 1
Makefile:637: recipe for target '.build_release/tools/convert_imageset.bin' failed
make: *** [.build_release/tools/convert_imageset.bin] Error 1
Makefile:637: recipe for target '.build_release/tools/upgrade_net_proto_text.bin' failed
make: *** [.build_release/tools/upgrade_net_proto_text.bin] Error 1
Makefile:642: recipe for target '.build_release/examples/siamese/convert_mnist_siamese_data.bin' failed
make: *** [.build_release/examples/siamese/convert_mnist_siamese_data.bin] Error 1
 

经过自己研究发现，如果使用Python2版本就会有以上的问题，如果我改为 python3版本，即

PYTHON_INCLUDE := /usr/include/python3.5m \
                   /usr/lib/python3.5/dist-packages/numpy/core/include

就发现没有问题了。我系统里面两个版本都安装有的，可能是默认使用python3的问题吧。

 

5. 找不到 -lboost_python3

CXX examples/cifar10/convert_cifar_data.cpp
AR -o .build_release/lib/libcaffe.a
LD -o .build_release/lib/libcaffe.so.1.0.0
/usr/bin/ld: 找不到 -lboost_python3
collect2: error: ld returned 1 exit status
Makefile:585: recipe for target '.build_release/lib/libcaffe.so.1.0.0' failed
make: *** [.build_release/lib/libcaffe.so.1.0.0] Error 1

解决方法：

在/usr/lib/x86_64-linux-gnu/ 下查看有没有libboost_python-py3.5.so，如果有链接一下就可以了。

sudo ln -s libboost_python-py35.so libboost_python3.so

如果你使用的是python3.6版本，发现在 /usr/lib/x86_64-linux-gnu/ 下没有libboost_python-py3.6.so 的，那你就的自己编译一个了，可以参考我的另一篇博文：https://blog.csdn.net/u012505617/article/details/88556621

 

 

---

 参考：

https://blog.csdn.net/sinat_27240143/article/details/82701285

https://blog.csdn.net/fanhenghui/article/details/80092131