kwin-wayland using all my CPU cycles.
I'm running Debian/Trixie on an AMD64 system. For the past day my desktop is slowed to a crawl. It started yesterday and continues today even after the massive updating of over 300 packages. Top reports that kwin-wayland is using over 2,000% of my CPU time while plasmashell is using another 200%. If I don't do anything, eventually the CPU usage drops to nothing but it starts up again when I start working again.
Tackling remaining Wayland related issues (was: Re: Remembering Konsole sessions)
Hi Gary, hi. Gary Dale - 14.03.25, 14:35:00 Mitteleuropäische Normalzeit: > >> Was the behavior changed, or am I missing something? > > > > It changed: https://bugs.kde.org/show_bug.cgi?id=461780 > > I find KDE-Wayland to be a real pain. In addition to this problem, it > restores all open programs to the centre of virtual desktop 1 no matter > where they were on shutdown. The sole exception to this seems to be > GKrellM which keeps its position but still on desktop 1 instead of the > one it used to be on. I wonder how about to tackle all those remaining issues. I look forward to test Plasma on Wayland again. But from reading about all of the remaining Wayland issues on this list I feel reluctant to do so. However without testing and reporting remaining issues that are not yet reported the situation might be unlikely to improve. While upstream as far as I know intends to keep Plasma on X11 running until the release of Plasma 7, at one point X11 support is likely to be dropped. Also a certain Debian Qt/KDE developer likes to nag about X11 support breaking things :). And I basically agree: Wayland is the more modern approach with way less legacy issues and baggage. I believe there is a reason for upstream splitting out X11 parts of kwin into its own compositor, likely to appear with Plasma 6.4. However I really get your pain, Gary. Whenever I switched in the past to see whether things work on Wayland it usually took less than a day and I was back on X11 again. For my main systems. A ThinkPad X1 Gen 1 tablet and my music laptop, a ThinkPad X260, are switched over to Wayland since quite a while. But on my main systems for private and freelance stuff as well as work I use a lot of different applications including some games and so far always something did not work on Wayland. I mostly use Debian packages apps and games, but also some stuff installed in a containerized approach as Flatpaks which needs further integration to make things work. On X11 all of this works. On Wayland so far not so. Most does… but anyway, my last test has been a while, so I hope to test again after completing some other currently more important tasks. I remember similar cycles with Pulseaudio for example. Or with Network Manager. It took years over years for those to finally work okay for me. And it was painful. The switch to Pipewire was an exception. It mostly works. But currently it does not seem to detect whether I plugged in head phones or not. Sound still comes out of laptop speakers when head phones are plugged in. I remember this worked after switching but got broke by some update. Off topic part: I complicate all of this for myself as I run Devuan¹ with Runit. And thus things like this may also be related to some service that should be run but is not running. While I took over Alpine Linux approach to start Pipewire via XDG auto start, it could be I missed something. I doubt it, but it could still be. But on the other hand I just love the predictability of a Runit based system. All of these strange behaviors with policy based decisions and bugs within Systemd that I encountered on various systems are gone. But on the other hand, I sometimes need to take care to make things work on my own. And Debian on Systemd is much better tested than Debian on another init. Especially for desktop environments. It is a mixed bag. I do not intend to go deeper into this here as it is off topic. And of course no offense meant to all of you who run Debian with Systemd :). For me it is about choice, so choose whatever works best for you! [1] As most of you probably know: Devuan is based on Debian and most packages are identical. But due to forking some packages it becomes much easier to run another init system like Runit. But it is also still possible to run other init systems in Debian as well. Switching is more painful though. Best, -- Martin
Re: Configure autologin for user in SystemSettings
On Wed, 12 Mar 2025 15:12:06 +
Roland Hamel wrote:
Hello Roland,
>LoginScreen (sddm) under
>Colours & Themes. Maybe a missing kde-config module.
Yes, kde-config-sddm is currently not in testing.
Further info starts here;
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1090041
--
Regards _ "Valid sig separator is {dash}{dash}{space}"
/ ) "The blindingly obvious is never immediately apparent"
/ _)rad "Is it only me that has a working delete key?"
He signed up for just three years, it seemed a small amount
Tin Soldiers - Stiff Little Fingers
pgps2CqbF2bca.pgp
Description: OpenPGP digital signature
kwin_wayland crash today on Trixie after upgrade (kwin package)
On Debian Trixie we have a huge package upgrade today (~325 packages).
I do "apt full-upgrade" every day, so this pack has arrived over last day.
One of them is "kwin" package which migrated to testing today:
https://tracker.debian.org/pkg/kwin
After "apt full-upgrade" and "reboot" system starts properly, everything
in green and login screen appears as usual.
But then, when I try to login, I'v got the following coredump report in
journal and login screen is back (instead of desktop session):
+++
Mar 14 11:58:54 desktop kwin_wayland_wrapper[2381]: No backend
specified, automatically choosing drm
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_core: Applying
output config failed!
Mar 14 11:58:55 desktop rtkit-daemon[1709]: Supervising 12 threads of 6
processes of 2 users.
Mar 14 11:58:55 desktop rtkit-daemon[1709]: Supervising 12 threads of 6
processes of 2 users.
Mar 14 11:58:55 desktop rtkit-daemon[1709]: Supervising 12 threads of 6
processes of 2 users.
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
0x502: GL_INVALID_OPERATION error generated. and are
incompatible
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
Invalid framebuffer status: "GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT"
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
0x502: GL_INVALID_OPERATION error generated. and are
incompatible
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
Invalid framebuffer status: "GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT"
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
0x502: GL_INVALID_OPERATION error generated. and are
incompatible
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
Invalid framebuffer status: "GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT"
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
0x502: GL_INVALID_OPERATION error generated. and are
incompatible
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: kwin_scene_opengl:
Invalid framebuffer status: "GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT"
Mar 14 11:58:55 desktop kwin_wayland_wrapper[2381]: KCrash: Application
'kwin_wayland' crashing... crashRecursionCounter = 2
Mar 14 11:58:55 desktop systemd-coredump[2433]: Process 2381
(kwin_wayland) of user 1000 terminated abnormally with signal 11/SEGV,
processing...
Mar 14 11:58:55 desktop systemd[1]: Created slice
system-drkonqi\x2dcoredump\x2dprocessor.slice - Slice
/system/drkonqi-coredump-processor.
Mar 14 11:58:55 desktop systemd[1]: Created slice
system-systemd\x2dcoredump.slice - Slice /system/systemd-coredump.
Mar 14 11:58:55 desktop systemd[1]: Started
systemd-coredump@0-2433-0.service - Process Core Dump (PID 2433/UID 0).
Mar 14 11:58:55 desktop systemd[1]: Started
drkonqi-coredump-processor@0-2433-0.service - Pass systemd-coredump
journal entries to relevant user for potential DrKonqi handling.
Mar 14 11:58:55 desktop systemd-coredump[2434]: Process 2381
(kwin_wayland) of user 1000 dumped core.
Stack trace of thread 2381:
#0 0x7f84c229e8ac
n/a (n/a + 0x0)
ELF object binary
architecture: AMD x86-64
Mar 14 11:58:56 desktop systemd[1]: systemd-coredump@0-2433-0.service:
Deactivated successfully.
Mar 14 11:58:56 desktop systemd[1]: systemd-coredump@0-2433-0.service:
Consumed 270ms CPU time, 151M memory peak.
Mar 14 11:58:56 desktop kcminit_startup[2385]: qt.qpa.wayland:
"wl-shell" is a deprecated shell extension, prefer using "xdg-shell" if
supported by the compositor by setting the environment
variable QT_WAYLAND_SHELL_INTEGRATION
Mar 14 11:58:56 desktop kcminit_startup[2385]: qt.qpa.wayland: Loading
shell integration failed.
Mar 14 11:58:56 desktop kcminit_startup[2385]: qt.qpa.wayland: Attempted
to load the following shells QList("xdg-shell", "wl-shell", "ivi-shell",
"qt-shell")
Mar 14 11:58:56 desktop kcminit_startup[2385]: qt.qpa.plugin: Could not
load the Qt platform plugin "wayland" in "" even though it was found.
Mar 14 11:58:56 desktop kwin_wayland_wrapper[2449]: No backend
specified, automatically choosing drm
Mar 14 11:58:56 desktop kwin_wayland_wrapper[2449]: kwin_core: Applying
output config failed!
Mar 14 11:58:56 desktop drkonqi-coredump-processor[2435]:
"/usr/bin/kwin_wayland" 2381
"/var/lib/systemd/coredump/core.kwin_wayland.1000.586376cbe46b4adebe7e9a8451cb7d11.2381.1741949935
00.zst"
Mar 14 11:58:56 desktop systemd[1805]: Started
drkonqi-coredump-launcher@0-2435-0.service - Launch DrKonqi for a
systemd-coredump crash (PID 2435/UID 0).
Mar 14 11:58:56 desktop systemd[1]:
drkonqi-coredump-processor@0-2433-0.service: Deactivated successfully.
Mar 14 11:58:56 desktop rtkit-daemon[1709]: Supervising 12 threads of 6
processes of 2 users.
Mar 14 11:58:56 desktop rtkit-daemon[1709]: Supervising 12 threa
journalctl kalandarac
Hey, I am looking at the output of a journalctl command and I am looking at the output of thousands of messages titled : kalandarac [x] . x always being the same number is this normal? How can we prevent these outputs? Thanks. Philippe Merlin signature.asc Description: This is a digitally signed message part.
Reply time format changed
Hi, I just realized that kmail2 6.3.2 uses a long format for inserting the time including the name of the time zone when replying to an e-mail. Just have a look at some of the recent answers in this mailing list, e.g. the one from Martin 2 days ago: > MERLIN Philippe - 10.03.25, 10:51:05 Mitteleuropäische Normalzeit: Another e-mail from Martin from 2025-02-12 shows > Erwan David - 12.02.25, 10:17:00 MEZ: And of course I see this in my own e-mails as well. To me it's a little bit annoying that the time zone is spelled out. I use a workaround and changed the corresponding templates to use '%OTIME' instead of '%OTIMELONG' but maybe there is a better fix. Frank
Re: kwin_wayland crash today on Trixie after upgrade (kwin package)
On 3/14/25 10:53 PM, Shmerl wrote: You didn't specify what GPU and driver you use. But from the looks of it, it could be an issue with Nvidia's proprietary OpenGL, which it very often is. Try may be installing the latest or beta driver from Nvidia? # lspci | grep VGA 01:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1070] (rev a1) # apt list nvidia-driver nvidia-driver/testing,now 535.216.03-2 amd64 [installed] Everything worked fine before that upgrade of this "kwin-wayland" package today... Best regards, Rafal
