kcmshell4 crashed with SIGSEGV in QDataStream::operator>>()

Application that crashed: kwrite
Version of the application: 4.3.2 (KDE 4.3.2)
KDE Version: 4.3.2 (KDE 4.3.2)
Qt Version: 4.5.2
Operating System: Linux 2.6.31.1-56.fc12.x86_64 x86_64

What I was doing when the application crashed:
Application crashed, when just run it.
following packaged installed.
system is fresh installed
kdebase-4.3.2-3.fc12.x86_64
qt-4.5.2-25.fc12.x86_64
pango-1.26.0-1.fc12.x86_64

 -- Backtrace:
Application: के-राइट (kwrite), signal: Segmentation fault
[KCrash Handler]
#5 0x00007f62b8e8d77c in QDataStream::operator>>(int&) () from /usr/lib64/libQtCore.so.4
#6 0x00007f62ba168f1f in KServiceFactory::KServiceFactory() () from /usr/lib64/libkdecore.so.5
#7 0x00007f62ba1691fa in KServiceFactory::self() () from /usr/lib64/libkdecore.so.5
#8 0x00007f62ba16162d in KService::serviceByDesktopName(QString const&) () from /usr/lib64/libkdecore.so.5
#9 0x00007f62bb9c69b2 in KTextEditor::EditorChooser::editor(QString const&, bool) () from /usr/lib64/libktexteditor.so.4
#10 0x00007f62bbbe93af in KDialog::setButtons(QFlags<KDialog::ButtonCode>) () from /usr/lib64/libkdeinit4_kwrite.so
#11 0x00007f62bbbebedf in kdemain () from /usr/lib64/libkdeinit4_kwrite.so
#12 0x00007f62b809ab4d in __libc_start_main () from /lib64/libc.so.6
#13 0x00000000004006c9 in _start ()

Reported using DrKonqi

- Do you have user permissions in /var/tmp/kdecache-$USER/ ?
- Does running "kbuildsycoca4 --noincremental" as user fixes this ?
- Does reinstalling "kdelibs", "kdebase" ("kdebase-apps" and "kdebase-runtime") fixes this ?
I guess you are not using SSH (like bug 211323), but you are using Fedora too.. that may be the connection.
Thanks

*** Bug 215420 has been marked as a duplicate of this bug. ***

Happens with many applications, probably started from a different user account, reassigning to kdelibs.

*** Bug 215460 has been marked as a duplicate of this bug. ***

*** Bug 215742 has been marked as a duplicate of this bug. ***

*** Bug 216028 has been marked as a duplicate of this bug. ***

*** Bug 216704 has been marked as a duplicate of this bug. ***

*** Bug 216724 has been marked as a duplicate of this bug. ***

A backtrace with debugging symbols enabled would be nice. Please install them, and try to reproduce the issue.

See also http://techbase.kde.org/Development/Tutorials/Debugging/How_to_create_useful_crash_reports

From bug 217781:
---
#5 QDataStream::operator>> (this=0x0, i=@0x7fffeab00f4c) at
io/qdatastream.cpp:715
#6 0x00007f4cff57578f in KServiceFactory::KServiceFactory (this=0x7530d0) at
/usr/src/debug/kdelibs-4.3.4/kdecore/services/kservicefactory.cpp:44
#7 0x00007f4cff575a52 in self (this=<value optimized out>) at
/usr/src/debug/kdelibs-4.3.4/kdecore/sycoca/ksycocafactory.h:200
#8 KServiceFactory::self (this=<value optimized out>) at
/usr/src/debug/kdelibs-4.3.4/kdecore/services/kservicefactory.cpp:75
#9 0x00007f4cff56da1d in KService::serviceByDesktopName (_name=<value
optimized out>) at
/usr/src/debug/kdelibs-4.3.4/kdecore/services/kservice.cpp:616
#10 0x00007f4d0127b192 in KTextEditor::EditorChooser::editor (postfix=<value
optimized out>, fallBackToKatePart=true) at
/usr/src/debug/kdelibs-4.3.4/interfaces/ktexteditor/editorchooser.cpp:129
#11 0x00007f4d017f9cff in KWrite::KWrite (this=0x752180, doc=0x0,
__in_chrg=<value optimized out>, __vtt_parm=<value optimized out>) at
/usr/src/debug/kdebase-4.3.4/apps/kwrite/kwritemain.cpp:67
#12 0x00007f4d017fc80f in kdemain (argc=<value optimized out>, argv=<value
optimized out>) at /usr/src/debug/kdebase-4.3.4/apps/kwrite/kwritemain.cpp:715

*** Bug 217781 has been marked as a duplicate of this bug. ***

*** Bug 206225 has been marked as a duplicate of this bug. ***

*** Bug 214934 has been marked as a duplicate of this bug. ***

Is this really a KWrite issue? Looks like a crash in kdecore... and this=0x0 in QDataStream...

The bug report is already on kdelibs/kdecore... as it affected other apps too...

*** Bug 219251 has been marked as a duplicate of this bug. ***

I see the following problem:

In KSycocaFactory constructor, the m_stream is only initialized when the sycoca is not building. In KServiceFactory constructor, this stream is used after the same check, but I am not sure if between those two checks the value of isBuilding() could change.

*** Bug 211323 has been marked as a duplicate of this bug. ***

Christoph: this can't be an issue. isBuilding() only true in kbuildsycoca4 and always false in every other KDE application that uses ksycoca.

Are there actual steps to reproduce this bug?

*** Bug 219910 has been marked as a duplicate of this bug. ***

*** Bug 218305 has been marked as a duplicate of this bug. ***

*** Bug 220019 has been marked as a duplicate of this bug. ***

*** Bug 220784 has been marked as a duplicate of this bug. ***

*** Bug 220924 has been marked as a duplicate of this bug. ***

*** Bug 220954 has been marked as a duplicate of this bug. ***

*** Bug 221670 has been marked as a duplicate of this bug. ***

*** Bug 222908 has been marked as a duplicate of this bug. ***

SVN commit 1076301 by sebsauer:

backport r1048522 by kuemmel from trunk to 4.3 branch;
don't crash
BUG:211401

 M +2 -0 kservicefactory.cpp

WebSVN link: http://websvn.kde.org/?view=rev&revision=1076301

Fixed in 4.3.5 and 4.4.0.

If the assert isn't compiled to a no-op, this still "crashes".

*** Bug 224402 has been marked as a duplicate of this bug. ***

*** Bug 224657 has been marked as a duplicate of this bug. ***

*** Bug 226101 has been marked as a duplicate of this bug. ***

*** Bug 226225 has been marked as a duplicate of this bug. ***

*** Bug 230375 has been marked as a duplicate of this bug. ***

*** Bug 232156 has been marked as a duplicate of this bug. ***

*** Bug 232309 has been marked as a duplicate of this bug. ***

*** Bug 236451 has been marked as a duplicate of this bug. ***

Git commit 5205232fe23d0a2997a308eee85524a3caad3763 by David Faure.
Committed on 07/06/2011 at 13:26.
Pushed by dfaure into branch 'KDE/4.6'.

Fix DBus timeout when kbuilsycoca takes too much time to run

... or when a kuniqueapplication takes time to start.

This should fix a large number of bugs where ksycoca code crashes
due to not waiting enough for kbuildsycoca.

I'll also get a patch into future versions of Qt to make the timeout
configurable in QDBusAbstractInterface, so that we don't have to port
all the possibly-long dbus calls to QDBusConnection like in this patch.

CCBUG: 211401

M +1 -1 kdecore/kernel/ktoolinvocation.cpp
M +7 -4 kdeui/kernel/kuniqueapplication.cpp

http://commits.kde.org/kdelibs/5205232fe23d0a2997a308eee85524a3caad3763

And now in KF 5.15, ksycoca rebuilds the cache itself, no DBus involved anymore.