nvidia-gfxG04 broke again?

Can’t be my kernel because the current one I’m running now with the nvidia driver is locked; I generally lock the kernel unless there’s a issue as I don’t need a new kernel every update or two. I also do this so I know the driver will work as opposed to a new kernel without any patches for the driver. I didn’t see any new kernels in the dup output unless this pulls one in despite a lock.

The following product is going to be upgraded:
openSUSE Tumbleweed 20220428-0 -> 20220602-0

I have a two part question…assuming it’s just a broken driver can I just lock nvidia with the kernel and how has this worked out for you guys? I assume just lock nvidia*?

Second, I’m really getting tired of this nvidia **** so on the new laptop or computer I get what’s a good card to look for that I won’t have this bs to deal with. I’m not a gamer but I do like to watch and stream 4k video and youtube? current kernel is “[FONT=monospace]5.17.4-1-default”

With TW you have to use

zypper dup

For TW use AMD or Intel graphics. For simple workload any modern builtin GPU will be OK (AMD GCN5 or RDNA2, Intel Xe).
Nvidia drivers have troubles with too new kernel.


zypper se -si nvid

My locking has nothing to do with NVidia. I’ve been locking since before TW existed. Currently on this PC there are 47 locks, most of which include wildcard characters. I lock kernels with zypper. I also lock initrds with chattr, so that rebuilds for content updates can’t break an initrd known to work.

I think you misunderstand …I do run dup. The driver was broke…I locked it and all is well now.

I usually just lock the kernel…this time the new nvidia driver was broke despite the old kernel still being there that ran fine with a previous nvidia driver. I just locked nvidia* and ran dup again and all is well.