My environment:
Lenovo Thinkpad P53
NVidia Quadro RTX 3000 Mobile GPU
nouveau driver
two external 1920x1200 monitors
Fedora 40 with latest updates.
I have had five complete hangs of my Wayland sessions in recent weeks, since I reinstalled my machine with Fedora 40. To show the frequency of the problem, timestamps are:
Fri Aug 9 17:36:00
Fri Sep 6 19:27:02
Mon Sep 9 19:28:03
Tue Oct 8 11:30:19
Mon Nov 4 13:15:43
The screens just freeze. I don’t believe these hangs were related to any particular activity, but I can’t be sure.
I can log on to the system from another machine, and I did a “dmesg -T”. The output is very similar in each case. It starts
[Fri Aug 9 17:36:00 2024] general protection fault, probably for non-canonical address 0xdead0000000000f0: 0000 [#1] PREEMPT SMP NOPTI
[Fri Aug 9 17:36:00 2024] CPU: 9 PID: 3585 Comm: kwin_wayland Tainted: G W OE 6.9.12-200.fc40.x86_64 #1
[Fri Aug 9 17:36:00 2024] Hardware name: LENOVO 20QN0005UK/20QN0005UK, BIOS N2NET59W (1.44 ) 02/21/2024
[Fri Aug 9 17:36:00 2024] RIP: 0010:nouveau_vma_find+0x4d/0x60 [nouveau]
I appreciate this isn’t likely to be enough information, but as I’m not an expert at the deeper levels of the software, could I get some advice on what information to collect to report this formally?
Thanks in advance.
– Peter