Red Hat Bugzilla – Bug 1275844
[abrt] kwrite: __kernel_vsyscall(): kwrite killed by SIGSEGV
Last modified: 2015-10-28 21:20:56 EDT
Description of problem:
Testing crash reporting while suffering from #1256456. This is an artificially induced crash, it is not a valid crash report.
Version-Release number of selected component:
runlevel: N 5
Thread no. 2 (6 frames)
#1 poll at ../sysdeps/unix/syscall-template.S:84
#3 _xcb_conn_wait at xcb_conn.c:459
#4 xcb_wait_for_event at xcb_in.c:648
#5 QXcbEventReader::run at qxcbconnection.cpp:1214
#6 QThreadPrivate::start at thread/qthread_unix.cpp:331
Created attachment 1087075 [details]
Created attachment 1087076 [details]
Created attachment 1087077 [details]
Created attachment 1087078 [details]
Created attachment 1087079 [details]
Created attachment 1087080 [details]
Created attachment 1087081 [details]
Created attachment 1087082 [details]
Created attachment 1087083 [details]
Created attachment 1087084 [details]
Created attachment 1087085 [details]
Another user experienced a similar problem:
Testing crash report with updated libreport/abrt, not a genuine report.
reason: kwrite killed by SIGSEGV
runlevel: N 5
Now, artificial or not, why does this kwrite crash reach ABRT to begin with? All crashes (and that includes "crashes" from something like "killall -SIGSEGV kwrite") in KDE GUI apps should be caught by DrKonqi and never get to ABRT!
I don't think I've seen a KDE crash handler in all my 23 Final testing, all crashes (including the ones we fixed as blockers) went to abrt.
neither kde-runtime-drkonqi nor plasma-workspace-drkonqi is in 23 Final RC6 KDE 32-bit live.
'drkonq' doesn't appear in comps-f23.xml.in or spin-kickstarts.
Which sounds like a bad bug in the spin kickstarts which is going to get us flooded with bug reports that should go upstream instead. :-(
This regression was introduced when Rex Dieter removed the hard dependency on DrKonqi to allow users to work around bugs in DrKonqi that way.