[英]CMake build Tensorflow C++ on Windows 10 Error : Test COMPILER_OPT_ARCH_NATIVE_SUPPORTED - failed
我想使用CMake在Windows 10上构建Tensorflow C ++。 但是,我无法通过COMPILER_OPT_ARCH_NATIVE_SUPPORTED测试,这导致构建尝试失败。 这是我在tf github repo上的公开问题: https : //github.com/tensorflow/tensorflow/issues/24076 ,它没有得到任何回应,因此我为什么要在stackoverflow上提问,希望这里有人能启发我。 我的想法不多了。 我已经尝试使用Python 3.5和3.6,Visual Studio 2017和2015。所有方法均未成功。
系统信息
操作系统平台和发行版(例如,Linux Ubuntu 16.04):Windows 10
从(源或二进制)安装TensorFlow:源
描述问题,我遵循https://github.com/tensorflow/tensorflow/tree/master/tensorflow/contrib/cmake中的“逐步构建Windows”。 但是,在第3步中,我似乎无法通过Performing Test COMPILER_OPT_ARCH_NATIVE_SUPPORTED
,从而导致构建过程失败。
提供您在遇到问题之前执行的命令/步骤的确切顺序
PS C:\Users\bw\tensorflow\tensorflow\contrib\cmake\build> cmake .. -A x64 -Thost=x64 -DCMAKE_BUILD_TYPE=Release `
>> -DSWIG_EXECUTABLE='C:\Program Files\swigwin-3.0.12\swig.exe' `
>> -DPYTHON_EXECUTABLE='C:\Program Files (x86)\Microsoft Visual Studio\Shared\Python36_64\python.exe' `
>> -DPYTHON_LIBRARIES='C:\Program Files (x86)\Microsoft Visual Studio\Shared\Python36_64\libs\python36.lib'
-- Building for: Visual Studio 15 2017
CMake Warning at CMakeLists.txt:9 (message):
Your current cmake generator is set to use 32 bit toolset architecture.
This may cause "compiler out of heap space" errors when building. Consider
using the flag -Thost=x64 when running cmake.
-- The C compiler identification is MSVC 19.16.27024.1
-- The CXX compiler identification is MSVC 19.16.27024.1
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/VC/Tools/MSVC/14.16.27023/bin/Hostx64/x64/cl.exe
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/VC/Tools/MSVC/14.16.27023/bin/Hostx64/x64/cl.exe -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/VC/Tools/MSVC/14.16.27023/bin/Hostx64/x64/cl.exe
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/VC/Tools/MSVC/14.16.27023/bin/Hostx64/x64/cl.exe -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Performing Test COMPILER_OPT_ARCH_NATIVE_SUPPORTED
-- Performing Test COMPILER_OPT_ARCH_NATIVE_SUPPORTED - Failed
-- Performing Test MSVC_OPENMP_SUPPORT
-- Performing Test MSVC_OPENMP_SUPPORT - Success
-- Found PythonInterp: C:/Program Files (x86)/Microsoft Visual Studio/Shared/Python36_64/python.exe (found version "3.6.7")
-- Found PythonLibs: optimized;C:/Program Files (x86)/Microsoft Visual Studio/Shared/Python36_64/libs/python36.lib;debug;C:/Program Files (x86)/Microsoft Visual Studio/Shared/Python36_64/libs/python36_d.lib (found version "3.6.7")
-- Found SWIG: C:/Program Files/swigwin-3.0.12/swig.exe (found version "3.0.12")
CMake Error at tf_python.cmake:811 (string):
string sub-command REPLACE requires at least four arguments.
Call Stack (most recent call first):
CMakeLists.txt:583 (include)
CMake Error at tf_python.cmake:812 (string):
string sub-command REPLACE requires at least four arguments.
Call Stack (most recent call first):
CMakeLists.txt:583 (include)
CMake Error at tf_python.cmake:813 (string):
string sub-command REPLACE requires at least four arguments.
Call Stack (most recent call first):
CMakeLists.txt:583 (include)
-- Configuring incomplete, errors occurred!
See also "C:/Users/bw/tensorflow/tensorflow/contrib/cmake/build/CMakeFiles/CMakeOutput.log".
See also "C:/Users/bw/tensorflow/tensorflow/contrib/cmake/build/CMakeFiles/CMakeError.log".
任何其他信息/日志
CMakeOutput.log: https ://www.dropbox.com/s/7fweyunxdbmxa1k/CMakeOutput.log ? dl =0 CMakeError.log: https ://www.dropbox.com/s/tucx0tl6346kdpd/CMakeError.log ? dl =0
指南的第一行指出:
TensorFlow不建议使用CMAKE构建。 请使用bazel为所有平台构建TF。 有关详细信息,请参见TensorFlow安装指南。
显然,您需要使用一个已知可以在Windows上通过CMAKE正确构建的发行版本。 如果要构建最新的代码,可能应该切换到推荐的构建系统。
查看此项目的CMakeLists.txt文件,可以看到Performing Test COMPILER_OPT_ARCH_NATIVE_SUPPORTED - Failed
是检查编译器是否接受-march=native
标志的步骤。 并非如此,请参阅此修复程序https://github.com/tensorflow/tensorflow/issues/8724#issuecomment-289326917 。
实际错误是由于tf_python.cmake
中的脚本问题tf_python.cmake
无法正确解析其中一个源文件。 在我看来,这些问题无关紧要。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.