Bug 261832 - kwin crash on return from screensaver
Summary: kwin crash on return from screensaver
Status: RESOLVED DUPLICATE of bug 241163
Alias: None
Product: kwin
Classification: Plasma
Component: general (show other bugs)
Version: unspecified
Platform: openSUSE Linux
: NOR crash
Target Milestone: ---
Assignee: KWin default assignee
URL:
Keywords:
: 263191 263984 (view as bug list)
Depends on:
Blocks:
 
Reported: 2011-01-02 01:00 UTC by Nathan Bradshaw
Modified: 2011-03-19 17:19 UTC (History)
3 users (show)

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


Attachments
New crash information added by DrKonqi (3.20 KB, text/plain)
2011-01-25 11:47 UTC, mickey mouse
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Nathan Bradshaw 2011-01-02 01:00:48 UTC
Application: kwin (4.5.4 (KDE 4.5.4) "release 9")
KDE Platform Version: 4.5.4 (KDE 4.5.4) "release 9"
Qt Version: 4.6.3
Operating System: Linux 2.6.34.7-0.5-desktop x86_64
Distribution: "openSUSE 11.3 (x86_64)"

-- Information about the crash:
Wiggled mouse to disable blank screen screensaver and kwin crashed and was autorestarted with effects turned off. Re-enable effects works but then actions like selecting a window from expose also trigger crash

-- Backtrace:
Application: KWin (kwin), signal: Aborted
[KCrash Handler]
#6  0x00007f5a4ceb29e5 in raise () from /lib64/libc.so.6
#7  0x00007f5a4ceb3ee6 in abort () from /lib64/libc.so.6
#8  0x00007f5a485bdcdd in __gnu_cxx::__verbose_terminate_handler() () from /usr/lib64/libstdc++.so.6
#9  0x00007f5a485bbef6 in ?? () from /usr/lib64/libstdc++.so.6
#10 0x00007f5a485bbf23 in std::terminate() () from /usr/lib64/libstdc++.so.6
#11 0x00007f5a485bc076 in __cxa_rethrow () from /usr/lib64/libstdc++.so.6
#12 0x00007f5a497f3563 in QEventLoop::exec (this=0x7ffff30b1290, flags=...) at kernel/qeventloop.cpp:214
#13 0x00007f5a497f788b in QCoreApplication::exec () at kernel/qcoreapplication.cpp:1003
#14 0x00007f5a4d2a7474 in kdemain (argc=<value optimized out>, argv=0x7ffff30b1350) at /usr/src/debug/kdebase-workspace-4.5.4/kwin/main.cpp:531
#15 0x00007f5a4ce9eb7d in __libc_start_main () from /lib64/libc.so.6
#16 0x0000000000400699 in _start () at ../sysdeps/x86_64/elf/start.S:113

Possible duplicates by query: bug 261616, bug 261585, bug 261482, bug 261400, bug 261278.

Reported using DrKonqi
Comment 1 Thomas Lübking 2011-01-02 16:33:30 UTC
_could_ be a dupe of #260575 and/or #261400 (all share the "window activation" aspect, but this is not through alt+tab)

@Nathan:
I assume the backtrace from the screensaver looks different from this (we've some similar, iirc crash some _dri funcs of the driver/mesa)

This trace is from an unhandled exception and unless you can provide a reproducable way to this issue and/or are able/willing to compile Qt w/o weird (*cough* broken *cough*) exception handling, there's nothing we can do about it, since this kind of "trace" says nothing, but "something is wrong somewhere, bye" :-(
Comment 2 Nathan Bradshaw 2011-01-04 17:34:58 UTC
Hi Thomas, mine doesn't seem to be an exact dupe to those guys as my crashes seem to be randomly spaced and annoyingly unrepeatable rather than for example on every alt-tab invocation. If the backtrace isn't any use then may as well just close this report. Of the handful of  machines at home and work  I run kde4 on, this one's video set up has always been the most flakey so it's probably not worth trying to track down issues with it... thanks for your time and all the effort you and the team put into kwin :)

cheers
N,
Comment 3 Thomas Lübking 2011-01-04 17:59:39 UTC
I assume the core issue is the crash after the screensaver (trace?) and afterwards (the "select on exposé" issues are just follow-ups?) sth. in the X11 driver is just borked pot. leading to weird memory bugs.

Therefore the original crash is the one to solve (but i fear in the driver)
Comment 4 Martin Flöser 2011-01-13 19:41:53 UTC
setting to backtrace as there is not much we can do till we get a backtrace showing where the exception is really thrown :-(
Comment 5 Martin Flöser 2011-01-15 10:01:19 UTC
*** Bug 263191 has been marked as a duplicate of this bug. ***
Comment 6 Martin Flöser 2011-01-22 17:32:21 UTC
*** Bug 263984 has been marked as a duplicate of this bug. ***
Comment 7 mickey mouse 2011-01-25 11:47:59 UTC
Created attachment 56413 [details]
New crash information added by DrKonqi

kwin (4.5.5 (KDE 4.5.5)) on KDE Platform 4.5.5 (KDE 4.5.5) using Qt 4.7.1

- What I was doing when the application crashed:
Yesterday before shutdown the system goes well, nvidia driver correctly loaded and running desktop effects, also boinc client was able to see and use nvidia gpu compute capability. There may be some troubles with them and nvidia proprietary driver, because cpu usage when desktop effects enables is always quite high (about 7-15% of dual core cpu). Two days ago I have installed new nvidia driver version (260.19.36), but yesterday always was fine.
Today, just after system boot, a window saying that kwin crashed is shown to me. Desktop effects were disabled. I made the usual system update via yum and then reboot to see if the trouble appears again. After reboot desktop effect were disabled, and when I have tried to activate them kwin crashed again. Another reboot, same situation: desktop effects still enabled for about half of a second, next kwin crash an they becomes disabled again.

- Custom settings of the application:
* Fedora 14 x86_64 with nvidia proprietary drivers installed
* nomodeset option added at kernel line for blocking the nouveau driver load into initrd
* boinc client disabled, I prefer to launch it by hand at every boot, so it don't had access to gpu because it was not loaded in memory
* the window opened at last shutdown yesterday were firefox with 4-5 tabs on desktop 2 and boinc gui manager with a konsole on desktop 3. All windows were placed on the first desktop the first time the crash occours
* widgets on the desktop: calendar, analog clock, two system monitors, a sticky note, a comic strip and a weather applet, all installed by default nothing custom

Ask for any other missing detail ;-)

-- Backtrace (Reduced):
#8  0x00000037d40bc08d in __gnu_cxx::__verbose_terminate_handler () at ../../../../libstdc++-v3/libsupc++/vterminate.cc:93
#9  0x00000037d40ba2a6 in __cxxabiv1::__terminate (handler=<value optimized out>) at ../../../../libstdc++-v3/libsupc++/eh_terminate.cc:39
#10 0x00000037d40ba2d3 in std::terminate () at ../../../../libstdc++-v3/libsupc++/eh_terminate.cc:49
#11 0x00000037d40ba426 in __cxxabiv1::__cxa_rethrow () at ../../../../libstdc++-v3/libsupc++/eh_throw.cc:116
#12 0x000000351275ae82 in QEventLoop::exec (this=0x7fff8a7b2850, flags=...) at kernel/qeventloop.cpp:214
Comment 8 Thomas Lübking 2011-03-19 17:19:43 UTC
could be caused by the lanczos filter ("acurate" scale method - try "smooth" instead) - see bug #241163

i'll mark it as dupe because all we have is the rethrown exception and w/o further information it could be about everything...

the dupe also contains information on how to provide a useful backtrace in this particular case (comment #6)

*** This bug has been marked as a duplicate of bug 241163 ***