Given that bug #1130057 (nvidia-kernel-dkms) seems to have been resolved, if the already applied fix will solve the nvidia-open-kernel-dkms package it might be worth tagging it in this bug as well. If not, but applying the same approach to address the issue would work, I am not sure if the fix could be applied to nvidia-open-kernel-dkms as well.
Thanks, Jonas. On Mon, 16 Mar 2026 02:21:38 +0100 Nicolas Le Cam <[email protected]> wrote: > Confirming on 6.19.6-1 and 6.19.6-2 which have landed on testing: > > Error! Bad return status for module build on kernel: 6.19.6+deb14-amd64 (x86_64) > Consult /var/lib/dkms/nvidia-current-open/550.163.01/build/make.log for more inf > ormation. > > The log has many errors duplicated, here is the deduplicated list: > error: ‘const struct dma_map_ops’ has no member named ‘map_resource’ > error: implicit declaration of function ‘in_irq’ > [-Wimplicit-function-declaration] > error: ‘struct vm_area_struct’ has no member named ‘__vm_flags’; did > you mean ‘vm_flags’? > error: too few arguments to function ‘pci_resize_resource’; expected 4, have 3 > > relates to: > - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1130539 > - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1130540 > - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1130541 > > patch available in CachyOS: > https://github.com/CachyOS/kernel-patches/blob/master/6.19/misc/nvidia/0003-Fix-compile-for-6.19.patch > it seems to require more adaptations on debian based distros, see > https://github.com/NVIDIA/open-gpu-kernel-modules/pull/ 1015#issuecomment-3826920700 > > -- > Nicolas Le Cam > >
