| Summary: | media-libs/opensubdiv compile fails with nvidia-cuda-toolkit-12.8 | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Mark Purtill <gentoo> |
| Component: | Current packages | Assignee: | Paul Zander <negril.nx+gentoo> |
| Status: | RESOLVED FIXED | ||
| Severity: | minor | CC: | proxy-maint |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | AMD64 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Bug Depends on: | |||
| Bug Blocks: | 949198 | ||
| Attachments: |
emerge --info
opensubdiv build log |
||
Created attachment 919809 [details]
emerge --info
Created attachment 919810 [details]
opensubdiv build log
With nvidia-cuda-toolkit-12.6.1-r1, opensubdiv compiles fine. The bug has been closed via the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=a8ab18e9b1a9957b8cf39d72d808825752ea0e3f commit a8ab18e9b1a9957b8cf39d72d808825752ea0e3f Author: Paul Zander <negril.nx+gentoo@gmail.com> AuthorDate: 2025-02-24 19:14:30 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2025-02-24 19:38:04 +0000 dev-util/nvidia-cuda-toolkit: 12.8.0-r2 use glibc version in patch Checking __GLIBC_USE (IEC_60559_FUNCS_EXT_C23) like the glibc-2.41 header does doesn't work under <glibc-2.41. So use the GLIBC version. Closes: https://bugs.gentoo.org/950235 Signed-off-by: Paul Zander <negril.nx+gentoo@gmail.com> Signed-off-by: Sam James <sam@gentoo.org> ...t-glibc-2.41.patch => nvidia-cuda-toolkit-glibc-2.41-r1.patch} | 8 ++++---- ...lkit-12.8.0-r1.ebuild => nvidia-cuda-toolkit-12.8.0-r2.ebuild} | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) Confirmed works for me, thanks. |
Not sure which version of nvidia-cuda-toolkit I should be using (all are marked ~amd64) but upgrading to 12.8 caused several errors like this when compiling opensubdiv: #$ "/usr/x86_64-pc-linux-gnu/gcc-bin/14"/gcc -D__CUDA_ARCH__=520 -D__CUDA_ARCH_LIST__=520 -D__NV_LEGACY_LAUNCH -c -x c++ -DCUDA_DOUBLE_MATH_FUNCTIONS "-march=alderlake" "-mabm" "-mno-cldemote" "-mno-kl" "-mno-sgx" "-mno-widekl" "-mshstk" "--param=l1-cache-line-size=64" "--param=l1-cache-size=32" "--param=l2-cache-size=30720" "-O2" "-pipe" "-fno-diagnostics-color" "-ggdb" -march=alderlake -mabm -mno-cldemote -mno-kl -mno-sgx -mno-widekl -mshstk --param=l1-cache-line-size=64 --param=l1-cache-size=32 --param=l2-cache-size=30720 -O2 -pipe -fno-diagnostics-color -ggdb -O2 -Wno-psabi -I"/opt/cuda/include" -I"/var/tmp/portage/media-libs/opensubdiv-3.6.0-r2/work/OpenSubdiv-3_6_0/opensubdiv" -I"/var/tmp/portage/media-libs/opensubdiv-3.6.0-r2/work/OpenSubdiv-3_6_0/glLoader" "-I/opt/cuda/bin/../targets/x86_64-linux/include" -m64 "/var/tmp/portage/media-libs/opensubdiv-3.6.0-r2/temp/tmpxft_00005c27_00000000-6_cudaKernel.cudafe1.cpp" -o "/var/tmp/portage/media-libs/opensubdiv-3.6.0-r2/work/OpenSubdiv-3_6_0_build/opensubdiv/CMakeFiles/osd_static_gpu.dir/osd/./osd_static_gpu_generated_cudaKernel.cu.o" In file included from /opt/cuda/include/crt/math_functions.h:5994, from /opt/cuda/include/crt/common_functions.h:303, from /opt/cuda/include/crt/host_runtime.h:285, from /var/tmp/portage/media-libs/opensubdiv-3.6.0-r2/temp/tmpxft_00005c27_00000000-6_cudaKernel.cudafe1.stub.c:8, from tmpxft_00005c27_00000000-6_cudaKernel.cudafe1.stub.c:1: /opt/cuda/include/crt/math_functions.h:6006:1: error: declaration of ‘double sinpi(double)’ has a different exception specifier 6006 | __func__(double sinpi(double a)); | ^~~~~~~~ /opt/cuda/include/crt/math_functions.h:2557:15: note: from previous declaration ‘double sinpi(double) noexcept’ Reproducible: Always Steps to Reproduce: 1. Add dev-util/nvidia-cuda-toolkit ~amd64 2. Make sure cuda use flag is set for opensubdiv 3. emerge opensubdiv