Source: nvidia-cuda-toolkit Version: 9.2.148-7 Severity: important The contents of /etc/nvcc.profile on ppc64le seem to be incorrect, where the following default is defined:
NVVMIR_LIBRARY_DIR = $(TOP)/nvvm/libdevice This location resolves to a location that doesn't exist (/usr/lib/nvidia-cuda-toolkit/bin/../nvvm/libdevice/) and causes compilation to fail (and I assume other task as well): > nvcc $1 -o ~/bin/$1.lx_ppc -O3 -std=c++11 -lgsl -lm -lgslcblas error in open: /usr/lib/nvidia-cuda-toolkit/bin/../nvvm/libdevice/libdevice.10.bc The correct location is /usr/lib/nvidia-cuda-toolkit/libdevice Additional paths also seem to be incorrect, including LD_LIBRARY_PATH, INCLUDES, and PATH. At first glance, /usr/lib/cuda/ seems to be the correct target for TOP, as the folder hierarchy looks similar. However, it doesn't actually match what nvcc.profile defines and most folders are empty. The following fixes my current blockers (compilation), but I assume the complete fix will require fixing all values in /nvcc.profile: NVVMIR_LIBRARY_DIR = /usr/lib/nvidia-cuda-toolkit/libdevice ---Alex -- System Information: Debian Release: 10.7 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 4.19.0 Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) (ignored: LC_ALL set to en_US.UTF-8), LANGUAGE=en_US.UTF-8 (charmap=UTF-8) (ignored: LC_ALL set to en_US.UTF-8) Shell: /bin/sh linked to /usr/bin/dash Init: systemd (via /run/systemd/system)
