Bug 482964 - KDE neon 6.0 X11 session does not recover properly from a suspend state
Summary: KDE neon 6.0 X11 session does not recover properly from a suspend state
Status: RESOLVED DUPLICATE of bug 481308
Alias: None
Product: plasmashell
Classification: Plasma
Component: general (show other bugs)
Version: 6.0.0
Platform: Neon Linux
: NOR major
Target Milestone: 1.0
Assignee: Plasma Bugs List
URL:
Keywords: qt6
Depends on:
Blocks:
 
Reported: 2024-03-09 11:21 UTC by Jonathan Richards
Modified: 2024-03-14 10:55 UTC (History)
2 users (show)

See Also:
Latest Commit:
Version Fixed In: 6.0.2
Sentry Crash Report:


Attachments
Extract from a pkcon update to KDE Neon 6.0 (4.82 KB, text/plain)
2024-03-09 11:21 UTC, Jonathan Richards
Details
Extract from a pkcon update including plasma 6.0.2 (13.81 KB, text/plain)
2024-03-13 20:27 UTC, Jonathan Richards
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Jonathan Richards 2024-03-09 11:21:00 UTC
Created attachment 166785 [details]
Extract from a pkcon update to KDE Neon 6.0

PROBLEM: It is my practice to suspend the PC overnight rather than shutdown and restart.  Recently the Plasma session has not resumed properly from the suspend state.
SUMMARY:
Following a recent update, KDE neon 6.0 does not recover properly from a suspend state.  The problem occurred first after updating on Mar 3, affecting several KDE packages (extract from pkcon log is attached)
After this update, resuming an X11 session from a suspend state results in a black screen, i.e. the session is not rendered, although the session is functionally present (see steps to reproduce, below).
The following error from /var/log/syslog seems to be relevant (this is the oldest)
Mar  4 09:16:09 Odin sddm-greeter-qt6[4609]: QRhiGles2: Context is lost.
Mar  4 09:16:14 Odin sddm-greeter-qt6[4609]: QRhiGles2: Context is lost.
STEPS TO REPRODUCE:
1. Enter suspend state from X11 session command line in a Konsole window:
[code]$ sudo systemctl suspend[/code]
Observed: System suspends as expected, with no signal to monitor, standby indication on motherboard.
2. Resume with momentary power switch activation
Observed: Signal to monitor is restored, screen is deep black.
3. Move mouse
Observed: mouse cursor, which changes according to the UI elements not being displayed: the edges of the Konsole window are "visible" to the mouse cursor, and are displayed on the monitor as changed mouse cursor appearance.
4. Open a new shell on tty3 [Alt-Ctl-F3] and log in
Observed: new session is successfully started.
5. [code]$top[/code]
Observed: plasmashell is running, as are the backends of the desktop widgets.
6. Switch back to the X11 session on tty2 [Alt-Ctl-F2]
Observed: black screen as at steps 2-3.
7. Switch again to tty3, examine /var/log/syslog
$ grep -E "^Mar  9 10:1.*Gles2" /var/log/syslog
Mar  9 10:10:19 Odin plasmashell[3285]: QRhiGles2: Context is lost.
Mar  9 10:15:53 Odin plasmashell[3285]: QRhiGles2: Context is lost.
Mar  9 10:18:06 Odin plasmashell[3285]: QRhiGles2: Context is lost.
Observed: every time the X11 session is asked to write to tty2, the QRhiGles2 message is issued.
Also Observed: comments at https://codebrowser.dev/qt6/qtbase/src/gui/rhi/qrhigles2.cpp.html are incomprehensible to ordinary humans :)
8. From tty3, restart the sddm greeter
$ sudo service sddm restart
Observed: an X11 session, properly rendered on the screen is started, but on tty1.  tty2 session is ended.
Also observed: the same effect can be achieved by issuing the restart command "blind" in the non-displaying X11 session Konsole on tty2: it's there, and functional, just not displaying anything.
Info Centre report:
Operating System: KDE neon 6.0
KDE Plasma Version: 6.0.0
KDE Frameworks Version: 6.0.0
Qt Version: 6.6.2
Kernel Version: 6.5.0-25-generic (64-bit)
Graphics Platform: X11
Processors: 6 × AMD Phenom(tm) II X6 1090T Processor
Memory: 15.6 GiB of RAM
Graphics Processor: NVIDIA GeForce GTX 1080/PCIe/SSE2
Comment 1 Jonathan Richards 2024-03-10 09:38:19 UTC
From original reporter:
Possibly the installed GLES packages are of interest (dpkg listing)
ii  libgles-dev:amd64                    1.4.0-1                                              amd64    Vendor neutral GL dispatch library -- GLES development files
ii  libgles1:amd64                          1.4.0-1                                             amd64     Vendor neutral GL dispatch library -- GLESv1 support
ii  libgles2:amd64                          1.4.0-1                                             amd64     Vendor neutral GL dispatch library -- GLESv2 support
ii  libnvidia-gl-535:amd64             535.161.07-0ubuntu0.22.04.1    amd64    NVIDIA OpenGL/GLX/EGL/GLES GLVND libraries and Vulkan ICD
ii  libnvidia-gl-535:i386                  535.161.07-0ubuntu0.22.04.1    i386         NVIDIA OpenGL/GLX/EGL/GLES GLVND libraries and Vulkan ICD
Comment 2 Nate Graham 2024-03-13 15:55:42 UTC
Any chance you can try again with Plasma 6.0.2 which was released yesterday? There are some changes in there what may have fixed this.
Comment 3 Jonathan Richards 2024-03-13 20:27:18 UTC
Created attachment 167121 [details]
Extract from a pkcon update including plasma 6.0.2

Selected from the pkcon log with 
$ grep -E "^ .*6\.0\.2" 202403131931-update.log
Comment 4 Jonathan Richards 2024-03-13 20:30:36 UTC
(In reply to Nate Graham from comment #2)
> Any chance you can try again with Plasma 6.0.2 which was released yesterday?
> There are some changes in there what may have fixed this.

Thanks, Nate.  I have repeated steps 1 and 2 of my original report (i.e. suspending and resuming) and can confirm that the X11 session on tty2 is resumed exactly as I would expect.  I've attached a list of the 6.0.2 packages that were updated; some change in there somewhere has eliminated this bug for me.
Comment 5 Nate Graham 2024-03-13 20:40:09 UTC
Excellent, I had a feeling it would. Most likely it was Bug 481308 all along, or one of the other ones fixed alongside that for 6.0.2.

*** This bug has been marked as a duplicate of bug 481308 ***
Comment 6 Jonathan Richards 2024-03-14 10:55:36 UTC
(In reply to Nate Graham from comment #5)
> Excellent, I had a feeling it would. Most likely it was Bug 481308 all
> along, or one of the other ones fixed alongside that for 6.0.2.
> 
> *** This bug has been marked as a duplicate of bug 481308 ***

Many thanks - although I searched before reporting, I didn't find 481308 - must try harder!
Also I note that my hypothesis about the QRhiGles2 message is wrong: that message occurs when switching between tty2 and tty3, with a Kwin popup about a graphics reset.  Not a bug, not even a minor irritation.
Thanks again
Jonathan
(In reply to Nate Graham from comment #5)
> Excellent, I had a feeling it would. Most likely it was Bug 481308 all
> along, or one of the other ones fixed alongside that for 6.0.2.
> 
> *** This bug has been marked as a duplicate of bug 481308 ***