Created attachment 292297 [details] Gnome log file under important tab HI All, Please find my important error messages under Gnome log manager. If you need additional information don't hesitate to ask. Best regards, Laci.
Which of those 6 lines do you mean exactly? And even with those messages, your machine works ok?
Hi, I mean the following lines in Gnome important tab. 13:25:01 kernel: nouveau 0000:01:00.0: bsp: init failed, -2 13:25:01 kernel: nouveau 0000:01:00.0: bsp: init failed, -2 13:25:01 kernel: nouveau 0000:01:00.0: vp: init failed, -2 The PC is really frozen for 1-3 minutes, not usable for web browsing, LibreOffice document creation/read, opening a terminal, starting new app, . . .
That's nouveau saying that some accelerators? cannot be initialized or so. Does your PC recover after those 1-3 minutes? Can you upload full dmesg? Thx.
Created attachment 292305 [details] lower level noveau log records
Created attachment 292307 [details] complete dmesg
That GpuWatchdog process, where does it come from? I cannot find it in my linux distros and people are reporting similar issues with it and mentioning the proprietary nvidia drivers. Do you have them installed, per chance? If so, remove them - you're using nouveau - and try to reproduce then. HTH.
I didn't install any additional driver for nvidia / GpuWatchdog. I think it came from automatic Ubuntu 18.04 install or a 3rd party. You mean I should remove GpuWatchdog and let nouveau manage the Gpu ?
Let's first see where it comes from. Do this as root: # dpkg -S $(which GpuWatchdog)
Ok, I am going to run the requested command. the 1rst GpuWatchdog record appeared when I started chrome browser. (signed page segmentation fault) this could be true since I am not familiar with SUID sandbox & the error message: "Most likely you need to configure your SUID sandbox correctly" 11:12:46 Main: Most likely you need to configure your SUID sandbox correctly 11:12:46 Main: Most likely you need to configure your SUID sandbox correctly 11:12:46 Main: [0903/111246.159445:ERROR:nacl_helper_linux.cc(308)] NaCl helper process running without a sandbox! 11:12:46 Main: Opening in existing browser session.
laci@sanyika:~$ sudo dpkg -S $(which GpuWatchdog) [sudo] password for laci: dpkg-query: error: --search needs at least one file name pattern argument
Hmm, so it looks like that GpuWatchdog thing is chrome thread. And the error in your dmesg before it: [ 266.208179] [TTM] Buffer eviction failed [ 266.346883] GpuWatchdog[2984]: segfault at 0 ip says that a buffer eviction in the TTM fails which probably causes the watchdog to segfault. In any case, this is a DRM issue, not a platform one. Reassigning... What you could do for starters is try the latest kernel 5.8 to check whether this has been fixed in the meantime. HTH.
Ok, then what shall I do?
> Ok, then what shall I do? Did you not read this? "What you could do for starters is try the latest kernel 5.8 to check whether this has been fixed in the meantime."
Then please assign a person at the mentioned new thread who can examine the reason of this bug, verify the root cause, assess whether they have fixed it or not. Then he can suggest the appropriate kernel and probably driver . . . I am user who needs some support during the correcting procedure . . . .
I've opened a bug report at Launchpad for this but cannot say for certain there is a kernel bug here. https://bugs.launchpad.net/ubuntu/+source/linux-signed-hwe-6.2/+bug/2045951