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.
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
Failed to build torch-quiver #110
Comments
I tried
|
I suspect this is because your cpp compiler has not supported |
oh, thanks. |
I also tried installing gcc, gxx and cudatoolkit-dev inside conda
and the new error with the above changes is that the newer GCC makes it pass the previous /home/user/miniconda3/envs/phoenix/bin/nvcc -I/tmp/pip-req-build-vrn7wfzm/./srcs/cpp/include -I/usr/local/cuda/include -I/home/user/miniconda3/envs/phoenix/lib/python3.8/site-packages/torch/include -I/home/user/miniconda3/envs/phoenix/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/user/miniconda3/envs/phoenix/lib/python3.8/site-packages/torch/include/TH -I/home/user/miniconda3/envs/phoenix/lib/python3.8/site-packages/torch/include/THC -I/home/sai/miniconda3/envs/phoenix/include/python3.8 -c srcs/cpp/src/quiver/cuda/quiver_sample.cu -o build/temp.linux-x86_64-3.8/srcs/cpp/src/quiver/cuda/quiver_sample.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options '-fPIC' -O3 --expt-extended-lambda -lnuma -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="_gcc" -DPYBIND11_STDLIB="_libstdcpp" -DPYBIND11_BUILD_ABI="_cxxabi1011" -DTORCH_EXTENSION_NAME=torch_quiver -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_61,code=compute_61 -gencode=arch=compute_61,code=sm_61 -std=c++14 |
so, It looks like the nvcc on my system is 9.1 which has gcc 6 as supported version, and the conda cuda version isn't used while building. Looks like there isn't easy way to update cuda except via root permissions, the conda package cudatoolkit-dev has just created nvcc without any other actual cuda headers and else. |
You can try nvidia docker, but I'm not sure if it helps without root privilege. |
ok thanks, since this was now a specific setup issue, we can close it. |
Hello, I tried both pip install and source install but the build fails. error log:
The text was updated successfully, but these errors were encountered: