Bug 402221 - Plasma Wayland session locks up immediately on sign in with Nouveau drivers
Summary: Plasma Wayland session locks up immediately on sign in with Nouveau drivers
Status: RESOLVED UPSTREAM
Alias: None
Product: plasmashell
Classification: Plasma
Component: generic-wayland (show other bugs)
Version: 5.12.5
Platform: openSUSE Linux
: NOR crash
Target Milestone: 1.0
Assignee: Plasma Bugs List
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-12-17 01:18 UTC by etna
Modified: 2021-10-11 16:06 UTC (History)
3 users (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments
kernel demsg (4.23 KB, text/plain)
2018-12-17 15:48 UTC, etna
Details
wayland-session.log entries (8.03 KB, text/plain)
2018-12-17 15:48 UTC, etna
Details
kwin_wayland backtrace (31.40 KB, text/plain)
2018-12-17 15:49 UTC, etna
Details

Note You need to log in before you can comment on or make changes to this bug.
Description etna 2018-12-17 01:18:25 UTC
SUMMARY:
As per topic. Machine boots up to SDDM, Plasma(Wayland) is selected, enter sign-in password, and machine completely locks up.


STEPS TO REPRODUCE
1. Boot computer
2. Sign in to Plasma (Wayland) at SDDM

OBSERVED RESULT
Machine instantly locks up. Sometimes, I might be able to SSH in, but dmesg only prints a bunch of Nouveau FIFO engine fault errors 

EXPECTED RESULT
PLasma Wayland session starts normally.

SOFTWARE/OS VERSIONS
Linux/KDE Plasma: OpenSUSE Leap 15.0
(available in About System)
KDE Plasma Version: 5.12

ADDITIONAL INFORMATION
It's almost definitely a Plasma / KWin fault as i have done the following:

- Upgrade the default OpenSUSE kernel to a vanilla 4.18 kernel from kernel.org. Lockup still occurs upon sign in
- Tested a Wayland session with Gnome successfully
- Upgraded Plasma to 5.14 with the repo recommended by OpenSUSE here and lockup still occurs
- Booted with nouveau.noaccel=1 and Plasma Wayland successfully loads but with terrible performance
Comment 1 David Edmundson 2018-12-17 12:28:36 UTC
>dmesg only prints a bunch of Nouveau FIFO engine fault errors 


Sounds pretty critical.

Please attach those messages.


Is kwin_wayland running after? Can you get a backtrace of it's current state.

Please also attach ~/.local/share/sddm/wayland-session.log
Comment 2 etna 2018-12-17 15:48:11 UTC
Created attachment 116966 [details]
kernel demsg

The last and most important part of the kernel dmesg log containing Nouveau information. Everything else above it was not relevant.
Comment 3 etna 2018-12-17 15:48:58 UTC
Created attachment 116967 [details]
wayland-session.log entries

SDDM/wayland-session.log entries
Comment 4 etna 2018-12-17 15:49:59 UTC
Created attachment 116968 [details]
kwin_wayland backtrace

Backtrace for kwin_wayland. I think this one is useless since it complains about too much missing information.
Comment 5 Bug Janitor Service 2019-01-01 03:44:22 UTC
Dear Bug Submitter,

This bug has been in NEEDSINFO status with no change for at least
15 days. Please provide the requested information as soon as
possible and set the bug status as REPORTED. Due to regular bug
tracker maintenance, if the bug is still in NEEDSINFO status with
no change in 30 days the bug will be closed as RESOLVED > WORKSFORME
due to lack of needed information.

For more information about our bug triaging procedures please read the
wiki located here:
https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging

If you have already provided the requested information, please
mark the bug as REPORTED so that the KDE team knows that the bug is
ready to be confirmed.

Thank you for helping us make KDE software even better for everyone!
Comment 6 etna 2019-01-01 13:33:03 UTC
logs attached.
Comment 7 etna 2019-01-01 13:34:08 UTC
Just to add on: this behavior is observed in a nested kwin_wayland session as well under Weston. Even though plasmashell is launched in a windowed kwin_wayland process, it locks up the entire system.
Comment 8 David Edmundson 2019-01-01 14:52:45 UTC
Anything that locks up the entire system is generally an issue at a lower level than kwin/plasma. 

Sorry, but this seems more like a mesa/nouveau problem. I don't think there's much we can do.