Rainer Klier wrote:

hi,

Am 14.10.22 um 13:25 schrieb Pit Suetterlin:

...

It did work flawlessly before, with kernel 5.19.8 and version 470.141.03-54.1 of nvidia.

Anyone else seeing this? Hints what to look for?

nvidia driver version 470.141.03 does not work out of the box with kernel 6.x

see https://forums.developer.nvidia.com/t/470xx-drivers-and-linux-6-0-kernel/229...

Thanks for the link. It seems to be about the compilation of the drivers though? As I understood, the neccessary patches have been applied by Stefan Dirsch. So they compile without errors. But they don't initialize the card...

Just found some other messages (repeatedly - coming from tried X starts?) kernel: NVRM: GPU 0000:02:00.0: Failed to copy vbios to system memory. kernel: NVRM: GPU 0000:02:00.0: RmInitAdapter failed! (0x30:0xffff:874) kernel: NVRM: GPU 0000:02:00.0: rm_init_adapter failed, device minor number 0 Pit Suetterlin wrote:

Rainer Klier wrote:

...

hi,

Am 14.10.22 um 13:25 schrieb Pit Suetterlin:

...

It did work flawlessly before, with kernel 5.19.8 and version 470.141.03-54.1 of nvidia.

Anyone else seeing this? Hints what to look for?

nvidia driver version 470.141.03 does not work out of the box with kernel 6.x

see https://forums.developer.nvidia.com/t/470xx-drivers-and-linux-6-0-kernel/229...

Thanks for the link. It seems to be about the compilation of the drivers though? As I understood, the neccessary patches have been applied by Stefan Dirsch. So they compile without errors. But they don't initialize the card...

Stefan Dirsch wrote:

On Fri, Oct 14, 2022 at 12:25:40PM +0100, Pit Suetterlin wrote:

...

I updated my TW laptop yesterday to TW20221012. After that, X does not start anymore. The laptop has an NVIDIA optimus card, GM108M [GeForce 940MX], I'm using the G05 series of drivers from the nvidia repo. The modules built w/o issues, and are loaded. But journalctl tells me

kernel: [drm:nv_drm_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to allocate NvKmsKapiDevice kernel: [drm:nv_drm_probe_devices [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000200] Failed to register device

and as a consequence X won't start (from Xorg.0.log) (EE) NVIDIA(GPU-0): Failed to initialize the NVIDIA GPU at PCI:2:0:0.

Hmm. Works for me with G05/Kernel 6.0 on a Quadro Kepler card without any issues. I suggest to try without

Then I might try to upgrade my home machine, too. That has a GeForce GTX 980. I've held that back so far...

options nvidia-drm modeset=1

in /usr/lib/modprobe.d/50-nvidia-default.conf.

Tried that (it's enough to just unload and reload the modules after changing it, yes?) But still the same error :( Can it be related to suse-prime somehow?

Pit Suetterlin wrote:

Just wanted to boot to the previous kernel, but I'm not at the machine, and not sure about the proper syntax for grub2-once :-o

Found it. grub2-once 1\>1 Booted fine, TW20221013 with kernel 5.19.8-1-default. nvidia module recognizes and initializes the card, X is running. Will try the home machine now

Patrick Shanahan wrote:

* Pit Suetterlin [10-14-22 10:50]:

...

Pit Suetterlin wrote:

...

Just wanted to boot to the previous kernel, but I'm not at the machine, and not sure about the proper syntax for grub2-once :-o

Found it. grub2-once 1\>1 Booted fine, TW20221013 with kernel 5.19.8-1-default. nvidia module recognizes and initializes the card, X is running.

Will try the home machine now

looks like grup2-reboot 1;reboot will also do the deed :) with "1" being the menu-entry

Nah, it's in the advanced submenu. So you first have to open that, and then there the second entry. That's the 1\>1, following grub2-once --help ;^> As for the difference between grub2-once and grub2-reboot I'm not sure. Seems indeed it does the same Will be easier with the home desktop in front of me :D