Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Failing to compile torchvision C++ API from source (>C++14 compiler required, which is installed) #4175

Open
mattpopovich opened this issue Jul 13, 2021 · 6 comments
Open

Failing to compile torchvision C++ API from source (>C++14 compiler required, which is installed) #4175

mattpopovich opened this issue Jul 13, 2021 · 6 comments
Labels
topic: build

Comments

@mattpopovich
Copy link

mattpopovich commented Jul 13, 2021 •
edited
Loading

🐛 Bug

I have installed PyTorch version 1.8.0 from source and am now attempting to install the C++ API to torchvision 0.9.0 from source.

When making torchvision's C++ API I am getting an error that I need to use a C++14 or later compatible compiler (I am using gcc 9.3.0 which supports C++14):

root@pc:/resources/vision/build# make VERBOSE=1 -j4
...
...
...
[ 71%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp
[ 73%] Building CUDA object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu.o
/usr/local/cuda/bin/nvcc  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem=/usr/local/lib/python3.8/dist-packages/torch/include -isystem=/usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem=/usr/include/python3.8  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11 -x cu -c /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu.o
In file included from /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu:69:
/usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:4:2: error: #error C++14 or later compatible compiler is required to use ATen.
    4 | #error C++14 or later compatible compiler is required to use ATen.
      |  ^~~~~
In file included from /usr/local/lib/python3.8/dist-packages/torch/include/c10/util/ArrayRef.h:19,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/DispatchKey.h:7,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Backend.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Layout.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/core/TensorBody.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Tensor.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Context.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:9,
                 from /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu:69:
/usr/local/lib/python3.8/dist-packages/torch/include/c10/util/C++17.h:24:2: error: #error You need C++14 to compile PyTorch
   24 | #error You need C++14 to compile PyTorch
      |  ^~~~~
[ 75%] Building CUDA object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o
/usr/local/cuda/bin/nvcc  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem=/usr/local/lib/python3.8/dist-packages/torch/include -isystem=/usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem=/usr/include/python3.8  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11 -x cu -c /resources/vision/torchvision/csrc/ops/cuda/nms_kernel.cu -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o
In file included from /resources/vision/torchvision/csrc/ops/cuda/nms_kernel.cu:1:
/usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:4:2: error: #error C++14 or later compatible compiler is required to use ATen.
    4 | #error C++14 or later compatible compiler is required to use ATen.
      |  ^~~~~

I think the places of interest are the following:

  • [ 71%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp.o
    • Uses the flag -std=gnu++14
  • [ 75%] Building CUDA object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o
    • Uses the flag -std=c++11

It seems like all of the CPU builds use the C++14 flag and the CUDA builds (incorrectly) use the C++11 flag.
Can I point this to CMAKE_CUDA_FLAGS being set in vision/CMakeLists.txt? It looks like CUDA_FLAGS contains the flag -std=c++11 as shown in vision/build/CMakeFiles/torchvision.dir/flags.make:

# CMAKE generated file: DO NOT EDIT!
# Generated by "Unix Makefiles" Generator, CMake Version 3.16

# compile CUDA with /usr/local/cuda/bin/nvcc
# compile CXX with /usr/bin/c++
CUDA_FLAGS =  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11

Any idea where this is coming from? How it's getting set? And how to change it so that it is set to -std=gnu++14 or similar? Is it coming from my PyTorch install?

To Reproduce

Steps to reproduce the behavior:

git clone https://github.com/pytorch/vision.git
cd vision
git checkout v0.9.0
mkdir build && cd build

cmake -DTorch_DIR=/usr/local/lib/python3.8/dist-packages/torch/share/cmake/Torch/ -DWITH_CUDA=on .. 

root@pc:/resources/vision/build# cmake -DTorch_DIR=/usr/local/lib/python3.8/dist-packages/torch/share/cmake/Torch/ -DWITH_CUDA=on ..
-- The C compiler identification is GNU 9.3.0
-- The CXX compiler identification is GNU 9.3.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- 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: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- The CUDA compiler identification is NVIDIA 11.2.67
-- Check for working CUDA compiler: /usr/local/cuda/bin/nvcc
-- Check for working CUDA compiler: /usr/local/cuda/bin/nvcc -- works
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Found Python3: /usr/lib/x86_64-linux-gnu/libpython3.8.so (found version "3.8") found components: Development 
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE  
-- Caffe2: Found protobuf with new-style protobuf targets.
-- Caffe2: Protobuf version 3.6.1
-- Found CUDA: /usr/local/cuda (found version "11.2") 
-- Caffe2: CUDA detected: 11.2
-- Caffe2: CUDA nvcc is: /usr/local/cuda/bin/nvcc
-- Caffe2: CUDA toolkit directory: /usr/local/cuda
-- Caffe2: Header version is: 11.2
-- Found CUDNN: /usr/lib/x86_64-linux-gnu/libcudnn.so  
-- Found cuDNN: v8.1.0  (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libcudnn.so)
CMake Warning at /usr/local/lib/python3.8/dist-packages/torch/share/cmake/Caffe2/public/cuda.cmake:198 (message):
  Failed to compute shorthash for libnvrtc.so
Call Stack (most recent call first):
  /usr/local/lib/python3.8/dist-packages/torch/share/cmake/Caffe2/Caffe2Config.cmake:88 (include)
  /usr/local/lib/python3.8/dist-packages/torch/share/cmake/Torch/TorchConfig.cmake:68 (find_package)
  CMakeLists.txt:17 (find_package)


-- Autodetected CUDA architecture(s):  6.1 6.1 6.1
-- Added CUDA NVCC flags for: -gencode;arch=compute_61,code=sm_61
-- Found Torch: /usr/local/lib/python3.8/dist-packages/torch/lib/libtorch.so  
-- Found ZLIB: /usr/lib/x86_64-linux-gnu/libz.so (found version "1.2.11") 
-- Found PNG: /usr/lib/x86_64-linux-gnu/libpng.so (found version "1.6.37") 
-- Found JPEG: /usr/lib/x86_64-linux-gnu/libjpeg.so (found version "80") 
-- Configuring done
-- Generating done
-- Build files have been written to: /resources/vision/build

make -j4 

root@pc:/resources/vision/build# make VERBOSE=1 -j4 
/usr/bin/cmake -S/resources/vision -B/resources/vision/build --check-build-system CMakeFiles/Makefile.cmake 0
/usr/bin/cmake -E cmake_progress_start /resources/vision/build/CMakeFiles /resources/vision/build/CMakeFiles/progress.marks
make -f CMakeFiles/Makefile2 all
make[1]: Entering directory '/resources/vision/build'
make -f CMakeFiles/torchvision.dir/build.make CMakeFiles/torchvision.dir/depend
make[2]: Entering directory '/resources/vision/build'
cd /resources/vision/build && /usr/bin/cmake -E cmake_depends "Unix Makefiles" /resources/vision /resources/vision /resources/vision/build /resources/vision/build /resources/vision/build/CMakeFiles/torchvision.dir/DependInfo.cmake --color=
make[2]: Leaving directory '/resources/vision/build'
make -f CMakeFiles/torchvision.dir/build.make CMakeFiles/torchvision.dir/build
make[2]: Entering directory '/resources/vision/build'
[  2%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/common_jpeg.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/common_jpeg.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/common_jpeg.cpp
[  6%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_png.cpp.o
[  6%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_image.cpp.o
[  8%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_jpeg.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_png.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/decode_png.cpp
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_image.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/decode_image.cpp
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/decode_jpeg.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/decode_jpeg.cpp
[ 10%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/encode_jpeg.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/encode_jpeg.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/encode_jpeg.cpp
[ 12%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/encode_png.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/encode_png.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/encode_png.cpp
[ 14%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/read_write_file.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/cpu/read_write_file.cpp.o -c /resources/vision/torchvision/csrc/io/image/cpu/read_write_file.cpp
[ 16%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/io/image/image.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/io/image/image.cpp.o -c /resources/vision/torchvision/csrc/io/image/image.cpp
[ 18%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/alexnet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/alexnet.cpp.o -c /resources/vision/torchvision/csrc/models/alexnet.cpp
[ 20%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/densenet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/densenet.cpp.o -c /resources/vision/torchvision/csrc/models/densenet.cpp
[ 22%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/googlenet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/googlenet.cpp.o -c /resources/vision/torchvision/csrc/models/googlenet.cpp
[ 24%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/inception.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/inception.cpp.o -c /resources/vision/torchvision/csrc/models/inception.cpp
[ 26%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/mnasnet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/mnasnet.cpp.o -c /resources/vision/torchvision/csrc/models/mnasnet.cpp
[ 28%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/mobilenet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/mobilenet.cpp.o -c /resources/vision/torchvision/csrc/models/mobilenet.cpp
[ 30%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/resnet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/resnet.cpp.o -c /resources/vision/torchvision/csrc/models/resnet.cpp
[ 32%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/shufflenetv2.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/shufflenetv2.cpp.o -c /resources/vision/torchvision/csrc/models/shufflenetv2.cpp
[ 34%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/squeezenet.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/squeezenet.cpp.o -c /resources/vision/torchvision/csrc/models/squeezenet.cpp
[ 36%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/models/vgg.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/models/vgg.cpp.o -c /resources/vision/torchvision/csrc/models/vgg.cpp
[ 38%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/deform_conv2d_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/deform_conv2d_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/deform_conv2d_kernel.cpp
[ 40%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/nms_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/nms_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/nms_kernel.cpp
[ 42%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/ps_roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/ps_roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/ps_roi_align_kernel.cpp
[ 44%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/ps_roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/ps_roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/ps_roi_pool_kernel.cpp
[ 46%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/roi_align_kernel.cpp
[ 48%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autocast/roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autocast/roi_pool_kernel.cpp
[ 51%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/deform_conv2d_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/deform_conv2d_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autograd/deform_conv2d_kernel.cpp
[ 53%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/ps_roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/ps_roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autograd/ps_roi_align_kernel.cpp
[ 55%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/ps_roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/ps_roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autograd/ps_roi_pool_kernel.cpp
[ 57%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autograd/roi_align_kernel.cpp
[ 59%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/autograd/roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/autograd/roi_pool_kernel.cpp
[ 61%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/deform_conv2d_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/deform_conv2d_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/deform_conv2d_kernel.cpp
[ 63%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/nms_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/nms_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/nms_kernel.cpp
[ 65%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/ps_roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/ps_roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/ps_roi_align_kernel.cpp
[ 67%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/ps_roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/ps_roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/ps_roi_pool_kernel.cpp
[ 69%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_align_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_align_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/roi_align_kernel.cpp
[ 71%] Building CXX object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp.o
/usr/bin/c++  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem /usr/local/lib/python3.8/dist-packages/torch/include -isystem /usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem /usr/local/cuda/include -isystem /usr/include/python3.8  -fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=gnu++14 -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp.o -c /resources/vision/torchvision/csrc/ops/cpu/roi_pool_kernel.cpp
[ 73%] Building CUDA object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu.o
/usr/local/cuda/bin/nvcc  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem=/usr/local/lib/python3.8/dist-packages/torch/include -isystem=/usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem=/usr/include/python3.8  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11 -x cu -c /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu.o
In file included from /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu:69:
/usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:4:2: error: #error C++14 or later compatible compiler is required to use ATen.
    4 | #error C++14 or later compatible compiler is required to use ATen.
      |  ^~~~~
In file included from /usr/local/lib/python3.8/dist-packages/torch/include/c10/util/ArrayRef.h:19,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/DispatchKey.h:7,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Backend.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Layout.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/core/TensorBody.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Tensor.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Context.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:9,
                 from /resources/vision/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu:69:
/usr/local/lib/python3.8/dist-packages/torch/include/c10/util/C++17.h:24:2: error: #error You need C++14 to compile PyTorch
   24 | #error You need C++14 to compile PyTorch
      |  ^~~~~
[ 75%] Building CUDA object CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o
/usr/local/cuda/bin/nvcc  -DWITH_CUDA -D__CUDA_NO_HALF_OPERATORS__ -Dtorchvision_EXPORTS -I/resources/vision/torchvision/csrc -isystem=/usr/local/lib/python3.8/dist-packages/torch/include -isystem=/usr/local/lib/python3.8/dist-packages/torch/include/torch/csrc/api/include -isystem=/usr/include/python3.8  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11 -x cu -c /resources/vision/torchvision/csrc/ops/cuda/nms_kernel.cu -o CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o
In file included from /resources/vision/torchvision/csrc/ops/cuda/nms_kernel.cu:1:
/usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:4:2: error: #error C++14 or later compatible compiler is required to use ATen.
    4 | #error C++14 or later compatible compiler is required to use ATen.
      |  ^~~~~
make[2]: *** [CMakeFiles/torchvision.dir/build.make:518: CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/deform_conv2d_kernel.cu.o] Error 1
make[2]: *** Waiting for unfinished jobs....
In file included from /usr/local/lib/python3.8/dist-packages/torch/include/c10/util/ArrayRef.h:19,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/DispatchKey.h:7,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Backend.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/c10/core/Layout.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/core/TensorBody.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Tensor.h:3,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/Context.h:4,
                 from /usr/local/lib/python3.8/dist-packages/torch/include/ATen/ATen.h:9,
                 from /resources/vision/torchvision/csrc/ops/cuda/nms_kernel.cu:1:
/usr/local/lib/python3.8/dist-packages/torch/include/c10/util/C++17.h:24:2: error: #error You need C++14 to compile PyTorch
   24 | #error You need C++14 to compile PyTorch
      |  ^~~~~
make[2]: *** [CMakeFiles/torchvision.dir/build.make:531: CMakeFiles/torchvision.dir/torchvision/csrc/ops/cuda/nms_kernel.cu.o] Error 1
make[2]: Leaving directory '/resources/vision/build'
make[1]: *** [CMakeFiles/Makefile2:76: CMakeFiles/torchvision.dir/all] Error 2
make[1]: Leaving directory '/resources/vision/build'
make: *** [Makefile:130: all] Error 2

Expected behavior

The make (build) completes without error.

Environment

Click to display environment 

root@pc:~# python3 -m torch.utils.collect_env
Collecting environment information...
PyTorch version: 1.8.0a0+56b43f4
Is debug build: False
CUDA used to build PyTorch: 11.2
ROCM used to build PyTorch: N/A

OS: Ubuntu 20.04.1 LTS (x86_64)
GCC version: (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Clang version: Could not collect
CMake version: version 3.16.3

Python version: 3.8 (64-bit runtime)
Is CUDA available: True
CUDA runtime version: Could not collect
GPU models and configuration: 
GPU 0: GeForce GTX 1080
GPU 1: GeForce GTX 1080
GPU 2: GeForce GTX 1080

Nvidia driver version: 460.84
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.1.0
/usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.1.0
HIP runtime version: N/A
MIOpen runtime version: N/A

Versions of relevant libraries:
[pip3] numpy==1.21.0
[pip3] pytorch-lightning==1.3.8
[pip3] torch==1.8.0a0+56b43f4
[pip3] torchmetrics==0.4.1
[pip3] torchvision==0.9.0a0+01dfa8e
[conda] Could not collect

  • PyTorch / torchvision Version (e.g., 1.0 / 0.4.0): 1.8.0 / 0.9.0
  • OS (e.g., Linux): Ubuntu 20.04
  • How you installed PyTorch / torchvision (conda, pip, source): source / source
  • Build command you used (if compiling from source): Listed above
  • Python version: 3.8
  • CUDA/cuDNN version: 11.2
  • GPU models and configuration: 3x GeForce GTX 1080

Additional context

The commands I used to build PyTorch:
git clone https://github.com/pytorch/pytorch.git
cd pytorch
git checkout v1.8.0
git submodule sync
git submodule update --init --recursive
python3 setup.py build -j4 --cmake-only
cd build
cmake -DBUILD_CUSTOM_PROTOBUF=0 ..
cmake --build . -j4 --target install
cd ..
python3 setup.py install

Additionally,
python maps to Python 2.7.18
python3 maps to Python 3.8.10

Lastly, I do have a Dockerfile that I am using... but it builds cuDNN, PyTorch, openCV, and torchvision from source... it's about 20GB and takes a few hours to build.
@mattpopovich mattpopovich changed the title Failing to compile torchvision C++ API from source (C++14 or later compatible compiler required) Failing to compile torchvision C++ API from source (>C++14 compiler required, which is installed) Jul 13, 2021
@mattpopovich mattpopovich mentioned this issue Jul 13, 2021
Closed
@mattpopovich
Copy link
Author

Looks to be similar to #2470 as they both fail at the same step in the build process.

@mattpopovich
Copy link
Author

mattpopovich commented Jul 14, 2021 •
edited
Loading

Update: I was able to successfully compile by manually changing the CUDA_FLAGS variable (removing the -std=c++11) in vision/build/CMakeFiles/torchvision.dir/flags.make:

flags.make before:

# CMAKE generated file: DO NOT EDIT!
# Generated by "Unix Makefiles" Generator, CMake Version 3.16

# compile CUDA with /usr/local/cuda/bin/nvcc
# compile CXX with /usr/bin/c++
CUDA_FLAGS =  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 -std=c++11
...

flags.make after:

# CMAKE generated file: DO NOT EDIT!
# Generated by "Unix Makefiles" Generator, CMake Version 3.16

# compile CUDA with /usr/local/cuda/bin/nvcc
# compile CXX with /usr/bin/c++
CUDA_FLAGS =  --expt-relaxed-constexpr -Xcompiler=-fPIC   -D_GLIBCXX_USE_CXX11_ABI=1 
...

Now to figure out how to fix that in CMake!

This was referenced Jul 14, 2021
Open
Open
@malfet
Copy link
Contributor

malfet commented Jul 15, 2021

Hmm, can you share output of which c++ and which g++ commands as well as g++ version?

@mattpopovich
Copy link
Author

@malfet

root@pc:/~# which c++ 
/usr/bin/c++

root@pc:/~# /usr/bin/c++ --version 
c++ (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Copyright (C) 2019 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.

root@pc:/~# which g++ 
/usr/bin/g++

root@pc:/~# /usr/bin/g++ --version 
g++ (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
Copyright (C) 2019 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.

@beniz
Copy link

beniz commented Aug 17, 2021

Thanks for the report and fix @mattpopovich I had the exact same issue on my side, and this did help fixing it, simply add -DCMAKE_CUDA_STANDARD=14 to the pytorch vision build call. Somehow nvcc defaults to C++11 on my system (Ubuntu 20.04).

I tried adding set(CMAKE_CUDA_STANDARD 14) to CMakeLists.txt next to set(CMAKE_CXX_STANDARD 14), but this did not fix the issue, and this I can't explain though.

@beniz beniz mentioned this issue Aug 17, 2021
Merged
@husnejahan
Copy link

husnejahan commented Oct 19, 2023 •
edited
Loading

How can use C++17 in setup.py files? @mattpopovich

#error C++17 or later compatible compiler is required to use ATen.
My version:
gcc --version
gcc (Ubuntu 10.5.0-1ubuntu1~22.04) 10.5.0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
topic: build
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@malfet @beniz @mattpopovich @husnejahan