search

pytorch / vision

Datasets, Transforms and Models specific to Computer Vision
https://pytorch.org/vision
BSD 3-Clause "New" or "Revised" License
15.99k stars 6.92k forks source link

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

Open mattpopovich opened 3 years ago

mattpopovich commented 3 years ago

🐛 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:

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 ```

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 commented 3 years ago

Looks to be similar to https://github.com/pytorch/vision/issues/2470 as they both fail at the same step in the build process.

mattpopovich commented 3 years ago

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!

malfet commented 3 years ago

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

mattpopovich commented 3 years ago

@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 commented 3 years ago

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.

husnejahan commented 11 months ago

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