Bug 252910 - kwin crash when using a lot pictures with gimp and lyx
Summary: kwin crash when using a lot pictures with gimp and lyx
Status: RESOLVED UPSTREAM
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:
Depends on:
Blocks:
 
Reported: 2010-09-30 23:19 UTC by ingarp
Modified: 2018-08-28 20:20 UTC (History)
1 user (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description ingarp 2010-09-30 23:19:59 UTC
Application: kwin (4.5.1 (KDE 4.5.1))
KDE Platform Version: 4.5.1 (KDE 4.5.1)
Qt Version: 4.7.0
Operating System: Linux 2.6.34.7-0.2-desktop x86_64
Distribution: "openSUSE 11.3 (x86_64)"

-- Information about the crash:
- What I was doing when the application crashed:

I am making a presentation with lyx and used digikam and gimp.  Kwin crashes after a while with this use, however I have not been able to narrow it down to a single event. Sometimes the load sspikes, but kwin does not crash as the kernel killes apps top get more memory., but in the end kwin crashes.

- Unusual behavior I noticed:
Xorg seems to take more and more space >500Mb. Might not be as unusuall, however, just before and after kwin crashes Xorg spiked in CPU as well, giving load of 12-15.
I use the nvidia-driver-blob version 256.53 

Another observation is that kwin does not appear to use a lot of ram before it crashes.

I guess this isn't the best of crash reports, I'll add more info if I find any more infromation...

-- Backtrace:
Application: KWin (kwin), signal: Aborted
[KCrash Handler]
#6  0x000000300d0329e5 in raise () from /lib64/libc.so.6
#7  0x000000300d033ee6 in abort () from /lib64/libc.so.6
#8  0x00000030034706f4 in qt_message_output (msgType=<value optimized out>, buf=<value optimized out>) at global/qglobal.cpp:2259
#9  0x0000003002f17d8e in QDebug::~QDebug (this=0x7fff85470790, __in_chrg=<value optimized out>) at /usr/include/QtCore/qdebug.h:85
#10 0x0000003002eb90ea in KApplicationPrivate::init (this=<value optimized out>, GUIenabled=<value optimized out>) at /usr/src/debug/kdelibs-4.5.1/kdeui/kernel/kapplication.cpp:513
#11 0x0000003002eba1f9 in KApplication::KApplication (this=0x7fff85471200, GUIenabled=true) at /usr/src/debug/kdelibs-4.5.1/kdeui/kernel/kapplication.cpp:351
#12 0x00007f93c75b7d65 in ?? () from /usr/lib64/libkdeinit4_kwin.so
#13 0x00007f93c75b98ae in kdemain () from /usr/lib64/libkdeinit4_kwin.so
#14 0x000000300d01eb7d in __libc_start_main () from /lib64/libc.so.6
#15 0x0000000000400699 in _start ()

Possible duplicates by query: bug 246404.

Reported using DrKonqi
Comment 1 Thomas Lübking 2010-09-30 23:59:53 UTC
arises from a badalloc catch in qdebug - you're indeed out of memory. the reason in the Xorg leak. Though this is likely a bug in the nvidia driver, you could check whether it derives from high resource (pixmap) usage (as gimp+digikam+composited(?!) kwin could induce)
start "xrestop" and have a look at the (rising) pixmap allocations and where they source.

If you've got some sort of shared memory GPU it could also be nvidias pixmap cache which is known to flood (caused by sth. in kdelibs...)
in this case calling
    nvidia-settings -a PixmapCache=0; nvidia-settings -a PixmapCache=1
from time to time resets (and frees) the cache (also providing speed improvements)
Comment 2 Martin Flöser 2010-10-21 21:24:19 UTC
not much we can do about out of memory situations.