caffe配置
caffe配置
配置工作基本參考上面兩個鏈接。
編譯環境爲：win7，VS2013，cuda6.5，opencv3.0
【10.28】到目前爲止，編譯還沒有成功，存在一些錯誤。
在此列出我在編譯過程中出現的問題以及未解決的錯誤，希望可以給大家一些參考，也希望大家幫我看看現在出現的錯誤，以及如何解決。
1、【出錯的具體文件我不記得了==。】錯誤 3 error C1083: 無法打開包括文件: “sys/ioctl.h”: No such file or directory F:\new user\caffe\caffe-master\src\caffe\parallel.cpp 6 1 caffetest
5 IntelliSense: 無法打開 源 文件 “sys/mman.h” f:\new user\caffe\caffe-master\src\caffe\parallel.cpp 7 1 caffetest
解決方法：直接註釋掉
2、layer文件夾：
1）【base_data_layer.cpp】 錯誤 21 error C4703: 使用了可能未初始化的本地指針變量“stream” f:\new user\caffe\caffe-master\src\caffe\layers\base_data_layer.cpp 98 1 caffetest
解決辦法：在項目屬性->配置屬性->C/C++->SDL檢查，選測否
2）【bnll_layer.cu】 錯誤 23 error : expected a “)” F:\new user\caffe\caffe-master\src\caffe\layers\bnll_layer.cu 36 1 caffetest
錯誤 31 error MSB3721: 命令“”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\bin\nvcc.exe” -gencode=arch=compute_20,code=\”sm_20,compute_20\” –use-local-env –cl-version 2013 -ccbin “D:\Program Files\Microsoft Visual Studio 12.0\VC\bin\x86_amd64” -I”D:\Program Files\boost_1_56_0” -I”D:\Program Files\opencv\build\include” -I”D:\Program Files\opencv\build\include\opencv” -I”D:\Program Files\opencv\build\include\opencv2” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\openblas” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\hdf5” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\lmdb” -I”F:\new user\caffe\caffe-master\include” -I”F:\new user\caffe\caffe-master\src” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -G –keep-dir x64\Debug -maxrregcount=0 –machine 64 –compile -cudart static -g -DWIN32 -D_DEBUG -D_CONSOLE -D_LIB -D_CRT_SECURE_NO_WARNINGS -D_SCL_SECURE_NO_WARNINGS -D_UNICODE -DUNICODE -Xcompiler “/EHsc /W3 /nologo /Od /Zi /RTC1 /MDd ” -o x64\Debug\bnll_layer.cu.obj “F:\new user\caffe\caffe-master\src\caffe\layers\bnll_layer.cu””已退出，返回代碼爲 2。 C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\BuildCustomizations\CUDA 6.5.targets 593 9 caffetest
解決方法： 代碼中的kBNLL_THRESHOLD 替換爲50.0
以下是還未解決的錯誤：
1、【constrastive_loss_layer.cpp】 錯誤 22 error C2784: “_Ty std::max(std::initializer_list<_Elem>,_Pr)”: 未能從“float”爲“std::initializer_list<_Elem>”推導 模板 參數 F:\new user\caffe\caffe-master\src\caffe\layers\contrastive_loss_layer.cpp 56 1 caffetest
錯誤 23 error C2780: “const _Ty &std::max(const _Ty &,const _Ty &,_Pr)”: 應輸入 3 個參數，卻提供了 2 個 F:\new user\caffe\caffe-master\src\caffe\layers\contrastive_loss_layer.cpp 56 1 caffetest
錯誤 24 error C2780: “_Ty std::max(std::initializer_list<_Elem>)”: 應輸入 1 個參數，卻提供了 2 個 F:\new user\caffe\caffe-master\src\caffe\layers\contrastive_loss_layer.cpp 56 1 caffetest
錯誤 25 error C2782: “const _Ty &std::max(const _Ty &,const _Ty &)”: 模板 參數“_Ty”不明確 F:\new user\caffe\caffe-master\src\caffe\layers\contrastive_loss_layer.cpp 56 1 caffetest
2、【conv_layer.cu】 錯誤 78 error : too few arguments in function call F:\new user\caffe\caffe-master\src\caffe\layers\conv_layer.cu 19 1 caffetest
錯誤 79 error MSB3721: 命令“”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\bin\nvcc.exe” -gencode=arch=compute_20,code=\”sm_20,compute_20\” –use-local-env –cl-version 2013 -ccbin “D:\Program Files\Microsoft Visual Studio 12.0\VC\bin\x86_amd64” -I”D:\Program Files\boost_1_56_0” -I”D:\Program Files\opencv\build\include” -I”D:\Program Files\opencv\build\include\opencv” -I”D:\Program Files\opencv\build\include\opencv2” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\openblas” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\hdf5” -I”F:\new user\caffe\caffe-master\3rdparty\3rdparty\include\lmdb” -I”F:\new user\caffe\caffe-master\include” -I”F:\new user\caffe\caffe-master\src” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -I”C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5\include” -G –keep-dir x64\Debug -maxrregcount=0 –machine 64 –compile -cudart static -g -DWIN32 -D_DEBUG -D_CONSOLE -D_LIB -D_CRT_SECURE_NO_WARNINGS -D_SCL_SECURE_NO_WARNINGS -D_UNICODE -DUNICODE -Xcompiler “/EHsc /W3 /nologo /Od /Zi /RTC1 /MDd ” -o x64\Debug\conv_layer.cu.obj “F:\new user\caffe\caffe-master\src\caffe\layers\conv_layer.cu””已退出，返回代碼爲 2。 C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\BuildCustomizations\CUDA 6.5.targets 593 9 caffetest
3、【db.cpp 】 錯誤 18 error C4716: “caffe::db::GetDB”: 必須返回一個值 f:\new user\caffe\caffe-master\src\caffe\util\db.cpp 42 1 caffetest
4、【hdf5.cpp 】 錯誤 19 error C2360: “occurrences_32”的初始化操作由“case”標籤跳過 F:\new user\caffe\caffe-master\src\caffe\util\hdf5.cpp 34 1 caffetest
5、【Signal_handler.cpp】 錯誤 17 error C2065: “SIGHUP”: 未聲明的標識符 F:\new user\caffe\caffe-master\src\caffe\util\signal_handler.cpp 16 1 caffetest
錯誤太多了=。=