I thought I'd better give another update given it has been a while and
I've not returned with the information requested. I've not found the
switch point to provide the bisect output asked for, but I can add some
additional info that may be of help.

Any 5.8 series kernel I have tried from rc1 through rc7 and the
5.8.0-53.60 package from the install prior to upgrade is pretty stable -
although not completely as it will crash in the same manner after a few
days use sometimes.

Any 5.11 series kernel from the rc2 onwards is decidedly unstable and
can usually crash within a few minutes of booting, but sometimes manages
a day or two.

A manual suspend using the menu generally causes no issues, but leaving
the computer to automatically suspend frequently causes problems,
frustratingly not always.

I have switched to Wayland from X.org in order to speed up the process
of tracking down the issue because it is orders of magnitude less
stable. I was running X.org when I logged the bug, but even the more
recent kernels would be fine for a few days around half the time, then
cause problems over a few boots and settle again. Switching to Wayland
has meant that on a problematic kernel I will either see the issue on
boot or within a day or two at the most. The details above in terms of
time scales are based on Wayland not X.org.

I will now fill in the gap in those kernel versions and look at the 5.9
and 5.10 series which I seem to have neglected!

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1928600

Title:
  [amdgpu] Gnome locks up with artifacts on screen and
  [drm:amdgpu_dm_commit_planes [amdgpu]] *ERROR* Waiting for fences
  timed out!

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1928600/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to