search

grimoire / amirstan_plugin

Useful tensorrt plugin. For pytorch and mmdetection model conversion.
MIT License
159 stars 38 forks source link

Does it support Windows such as Win10? #13

Open yxliang opened 3 years ago

yxliang commented 3 years ago

Hi, is it any possible to make this awesome lib to support windows?

grimoire commented 3 years ago

Hi, I haven't tested this repo on windows, but as long as nvidia did provide TensorRT windows package, that should be possible.

yxliang commented 3 years ago

Many thanks.

biao321 commented 3 years ago

Hi, is it any possible to make this awesome lib to support windows?

Have you successed to build the awesome libs in Windows? I come to the error " The CUDA compiler 'C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.1/bin/nvcc.exe' is not able to compile a simple test program." when adopting CMake tools to build the libs.

The error logs are as follows: Selecting Windows SDK version 10.0.17763.0 to target Windows 10.0.19042. The C compiler identification is MSVC 19.16.27045.0 The CXX compiler identification is MSVC 19.16.27045.0 Detecting C compiler ABI info Detecting C compiler ABI info - done Check for working C compiler: D:/SoftWare/VS2017/VC/Tools/MSVC/14.16.27023/bin/Hostx86/x64/cl.exe - skipped Detecting C compile features Detecting C compile features - done Detecting CXX compiler ABI info Detecting CXX compiler ABI info - done Check for working CXX compiler: D:/SoftWare/VS2017/VC/Tools/MSVC/14.16.27023/bin/Hostx86/x64/cl.exe - skipped Detecting CXX compile features Detecting CXX compile features - done Found CUDA: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.1 (found version "11.1") Found TensorRT headers at D:/mmdet/TensorRT-8.0.1.6.Windows10.x86_64.cuda-11.3.cudnn8.2/TensorRT-8.0.1.6/include Find TensorRT libs at D:/mmdet/TensorRT-8.0.1.6.Windows10.x86_64.cuda-11.3.cudnn8.2/TensorRT-8.0.1.6/lib/nvinfer.lib;D:/mmdet/TensorRT-8.0.1.6.Windows10.x86_64.cuda-11.3.cudnn8.2/TensorRT-8.0.1.6/lib/nvparsers.lib;D:/mmdet/TensorRT-8.0.1.6.Windows10.x86_64.cuda-11.3.cudnn8.2/TensorRT-8.0.1.6/lib/nvinfer_plugin.lib Found TENSORRT: D:/mmdet/TensorRT-8.0.1.6.Windows10.x86_64.cuda-11.3.cudnn8.2/TensorRT-8.0.1.6/include
WITH_DEEPSTREAM: false GPU_ARCHS is not defined. Generating CUDA code for default SMs: 35;53;61;70;75;80 The CUDA compiler identification is NVIDIA 11.1.74 Detecting CUDA compiler ABI info Detecting CUDA compiler ABI info - failed Check for working CUDA compiler: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.1/bin/nvcc.exe Check for working CUDA compiler: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.1/bin/nvcc.exe - broken CMake Error at C:/Program Files/CMake/share/cmake-3.21/Modules/CMakeTestCUDACompiler.cmake:56 (message): The CUDA compiler

"C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v11.1/bin/nvcc.exe"

is not able to compile a simple test program.

It fails with the following output:

Change Dir: D:/mmdet/amirstan_plugin/build/CMakeFiles/CMakeTmp

Run Build Command(s):D:/SoftWare/VS2017/MSBuild/15.0/Bin/MSBuild.exe cmTC_0362e.vcxproj /p:Configuration=Debug /p:Platform=x64 /p:VisualStudioVersion=15.0 /v:m && 用于 .NET Framework 的 Microsoft (R) 生成引擎版本 15.9.21+g9802d43bc3

  Compiling CUDA source file main.cu...

  D:\mmdet\amirstan_plugin\build\CMakeFiles\CMakeTmp>"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\bin\nvcc.exe" -gencode=arch=compute_52,code=\"sm_52,compute_52\" --use-local-env -ccbin "D:\SoftWare\VS2017\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64" -x cu    -I"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\include"     --keep-dir x64\Debug -maxrregcount=0  --machine 64 --compile -cudart static -Xcompiler="-Wno-deprecated-declarations -Zi -Ob0" -g   -D"CMAKE_INTDIR=\"Debug\"" -D_MBCS -D"CMAKE_INTDIR=\"Debug\"" -Xcompiler "/EHsc /W1 /nologo /Od /FdcmTC_0362e.dir\Debug\vc141.pdb /FS /Zi /RTC1 /MDd " -o cmTC_0362e.dir\Debug\main.obj "D:\mmdet\amirstan_plugin\build\CMakeFiles\CMakeTmp\main.cu" 

cl : 命令行 error D8021: 无效的数值参数“/Wno-deprecated-declarations” [D:\mmdet\amirstan_plugin\build\CMakeFiles\CMakeTmp\cmTC_0362e.vcxproj]

  main.cu

D:\SoftWare\VS2017\Common7\IDE\VC\VCTargets\BuildCustomizations\CUDA 11.1.targets(785,9): error MSB3721: 命令“"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\bin\nvcc.exe" -gencode=arch=compute_52,code=\"sm_52,compute_52\" --use-local-env -ccbin "D:\SoftWare\VS2017\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64" -x cu    -I"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1\include"     --keep-dir x64\Debug -maxrregcount=0  --machine 64 --compile -cudart static -Xcompiler="-Wno-deprecated-declarations -Zi -Ob0" -g   -D"CMAKE_INTDIR=\"Debug\"" -D_MBCS -D"CMAKE_INTDIR=\"Debug\"" -Xcompiler "/EHsc /W1 /nologo /Od /FdcmTC_0362e.dir\Debug\vc141.pdb /FS /Zi /RTC1 /MDd " -o cmTC_0362e.dir\Debug\main.obj "D:\mmdet\amirstan_plugin\build\CMakeFiles\CMakeTmp\main.cu"”已退出，返回代码为 2。 [D:\mmdet\amirstan_plugin\build\CMakeFiles\CMakeTmp\cmTC_0362e.vcxproj]