Sven-Hendrik Haase pushed to branch main at Arch Linux / Packaging / Packages / python-pytorch
Commits: da68adac by Sven-Hendrik Haase at 2025-01-31T23:24:06+01:00 upgpkg: 2.6.0-2: Clean up package quite a bit - - - - - 2 changed files: - .SRCINFO - PKGBUILD Changes: ===================================== .SRCINFO ===================================== @@ -1,7 +1,7 @@ pkgbase = python-pytorch pkgdesc = Tensors and Dynamic neural networks in Python with strong GPU acceleration pkgver = 2.6.0 - pkgrel = 1 + pkgrel = 2 url = https://pytorch.org arch = x86_64 license = BSD ===================================== PKGBUILD ===================================== @@ -8,7 +8,7 @@ pkgname=("${pkgbase}" "${pkgbase}-opt" "${pkgbase}-cuda" "${pkgbase}-opt-cuda" " # When updating pytorch, also check the compatibility table for torchvision # https://github.com/pytorch/vision?tab=readme-ov-file#installation pkgver=2.6.0 -pkgrel=1 +pkgrel=2 _pkgdesc='Tensors and Dynamic neural networks in Python with strong GPU acceleration' pkgdesc="${_pkgdesc}" arch=('x86_64') @@ -220,26 +220,20 @@ _prepare() { export ATEN_NO_TEST=ON # do not build ATen tests export USE_MKLDNN=ON export BUILD_CUSTOM_PROTOBUF=OFF - # Caffe2 support was removed from pytorch with version 2.2.0 - export BUILD_CAFFE2=OFF - export BUILD_CAFFE2_OPS=OFF - # export BUILD_SHARED_LIBS=OFF export USE_FFMPEG=ON export USE_GFLAGS=ON export USE_GLOG=ON export USE_VULKAN=ON - # Currently broken https://github.com/pytorch/pytorch/issues/146042 - export BUILD_BINARY=OFF export USE_OBSERVERS=ON export USE_OPENCV=ON # export USE_SYSTEM_LIBS=ON # experimental, not all libs present in repos export USE_SYSTEM_NCCL=ON export USE_SYSTEM_PYBIND11=ON export USE_SYSTEM_EIGEN_INSTALL=ON + export USE_GOLD_LINKER=ON export NCCL_VERSION=$(pkg-config nccl --modversion) export NCCL_VER_CODE=$(sed -n 's/^#define NCCL_VERSION_CODE\s*\(.*\).*/\1/p' /usr/include/nccl.h) # export BUILD_SPLIT_CUDA=ON # modern preferred build, but splits libs and symbols, ABI break - # export USE_FAST_NVCC=ON # parallel build with nvcc, spawns too many processes export USE_CUPTI_SO=ON # make sure cupti.so is used as shared lib export CC=/usr/bin/gcc-13 export CXX=/usr/bin/g++-13 @@ -251,17 +245,17 @@ _prepare() { export CUDNN_LIB_DIR=/usr/lib export CUDNN_INCLUDE_DIR=/usr/include export TORCH_NVCC_FLAGS="-Xfatbin -compress-all" - # from ./Dockerfile + + # This list is from ./Dockerfile export TORCH_CUDA_ARCH_LIST="7.0 7.2 7.5 8.0 8.6 8.7 8.9 9.0 9.0a" - export OVERRIDE_TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST}" + export ROCM_PATH=/opt/rocm export HIP_ROOT_DIR=/opt/rocm - # from .ci/docker/libtorch/build.sh - export PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100;gfx1101;gfx942" - # copied from rocBLAS + + # Looking into enabling more architectures, see e.g. rocBLAS # https://github.com/ROCm/rocBLAS/blob/9c8a7dfeb3d0a808321541567447b5c1d17cd070/CMakeLists.txt#L114 - # export PYTORCH_ROCM_ARCH="gfx900;gfx906:xnack-;gfx908:xnack-;gfx90a:xnack+;gfx90a:xnack-;gfx940;gfx941;gfx942;gfx1010;gfx1012;gfx1030;gfx1100;gfx1101;gfx1102" - export PYTORCH_ROCM_ARCH="gfx900" + # This list is from .ci/docker/libtorch/build.sh + export PYTORCH_ROCM_ARCH="gfx900;gfx906;gfx908;gfx90a;gfx1030;gfx1100;gfx1101;gfx942" # 1. Compile source code for supported GPU archs in parallel # 2. Use gcc 13 toolchain as ROCm is not compatible with gcc 14. View it on GitLab: https://gitlab.archlinux.org/archlinux/packaging/packages/python-pytorch/-/commit/da68adac63be1aed1de67a964a40bb04a295fd41 -- View it on GitLab: https://gitlab.archlinux.org/archlinux/packaging/packages/python-pytorch/-/commit/da68adac63be1aed1de67a964a40bb04a295fd41 You're receiving this email because of your account on gitlab.archlinux.org.