I just installed the latest Tumbleweed update, including a new kernel. I tried to reinstall nvidia driver 570.153.02 but I got the error message that “nvidia.ko” couldn’t be loaded.
Is there a patch for this driver and kernel-default 6.14.6?
@pilotgi Hi, switch to a VT (tty) with ctrl+alt+F1, login as root user, run systemctl isolate multi-user.target also make sure no nvidia services are running, then force the re-install of the 570.153.02 rpm. Assuming your running the 6.14.6-2 kernel…
Thank you. Is this something I have to do from now on to install/build nvidia drivers?
@pilotgi so that worked?
Yes. Then I used systemctl isolate graphical.target.
@pilotgi I would wait and see what happens with the next snapshot release which should contain the 6.15.n kernel.
This appears to be another issue. The G04 driver does not build according to the ML thread.
The TO uses G06 and it installed fine after forced reinstallation of the package. On all local machines here which use the G06 driver, the update to the latest kernel went through without any issues.
The main difference to the ML thread is, that with the G06 driver, the updates directory (/usr/lib/modules/6.14.6-2-default/updates) is properly populated with the Nvidia modules (at least at all local machines).
The responses in this thread seem to be referring to the G06 driver. The issue I had was with the 570.153.02 .run file. I see now that malcolmlewis responded with " force the re-install of the 570.153.02 rpm."
I was using the .run file when I had my issue and malcolmlewis’s suggestion solved it.
@pilotgi So if as @arvidjaar indicated a CC version issue, then in the dkms config can you add run file command options? If so add --no-cc-version-check that’s what I use with the run file here.