https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122281
--- Comment #2 from Benjamin Schulz <schulz.benjamin at googlemail dot com> --- What I want to note is that there was an abi change in Cuda recently that threw out many old gpu's (including the one that gcc compiles for per default) https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html and made it appears nvidia made cuda incompatible with previous versions. Clang had to include fixes for this ABI change: https://github.com/llvm/llvm-project/pull/159354 Maybe this problem related to that? I tried to downgrade to cuda 12, or use an old nvidia driver, but the problem stayed.
