[Bug backtrace/31122] gdb crashed when drkonqi ran it on plasmashell crashes

["matt.fagnani at bell dot net" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=31122

--- Comment #2 from Matt Fagnani <matt.fagnani at bell dot net> ---
Created attachment 15245
  --> https://sourceware.org/bugzilla/attachment.cgi?id=15245&action=edit
Journal for the drkonqi, kioslave, gdb crashes when creating a plasmashell
crash trace

Thanks. The core dumps of the first 3 gdb crashes I mentioned were removed from
my system automatically due to crashes which happened after them and journal
rotation. I have the full traces saved. I did look at the string1 values for
the crash in skip_ws.

(gdb) frame 5
#5  0x0000559d5be3c93a in skip_ws (
    string1=@0x7ffed94630f8: 0x559d65f1c4f0 "ConversionCheck::supported", 
    string2=@0x7ffed94630f0: 0x7ffed94632e8 "QThread",
end_str2=end_str2@entry=0x7ffed94632ef "")
    at ../../gdb/utils.c:2049
2049      while (ISSPACE (*string1))
(gdb) p string1
$3 = (const char *&) @0x7ffed94630f8: 0x559d65f1c4f0
"ConversionCheck::supported"
(gdb) p *string1
$4 = 67 'C'

The journal at the time of the last crash with gdb 14.1 showed that
plasma-plasmashell.service sent SIGABRT to plasmashell, drkonqi, kioslave5, and
gdb. The gdb traces might just have been the state gdb was in
plasma-plasmashell.service aborted them which might be related to them varying
so much. I'm attaching the journal around the time of these crashes.

Dec 08 05:42:28 systemd[1444]: plasma-plasmashell.service: State 'stop-sigterm'
timed out. Aborting.
Dec 08 05:42:28 audit[3571]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3
subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=3571
comm="kioslave5" exe="/usr/libexec/kf5/kioslave5" sig=6 res=1
Dec 08 05:42:28 plasmashell[3581]: Fatal signal: Aborted
Dec 08 05:42:28 plasmashell[3581]: ----- Backtrace -----
Dec 08 05:42:28 systemd[1444]: plasma-plasmashell.service: Killing process 3472
(plasmashell) with signal SIGABRT.
Dec 08 05:42:28 plasmashell[3556]: KCrash: Application 'drkonqi' crashing...
Dec 08 05:42:28 plasmashell[3556]: KCrash: Attempting to start
/usr/libexec/drkonqi
Dec 08 05:42:28 systemd[1444]: plasma-plasmashell.service: Killing process 3556
(drkonqi) with signal SIGABRT.
Dec 08 05:42:28 systemd[1444]: plasma-plasmashell.service: Killing process 3571
(kioslave5) with signal SIGABRT.
Dec 08 05:42:28 systemd[1444]: plasma-plasmashell.service: Killing process 3581
(gdb) with signal SIGABRT.
Dec 08 05:42:28 systemd[1]: Created slice system-systemd\x2dcoredump.slice -
Slice /system/systemd-coredump.

There was a trace in the journal involving the drkonqi Python program
/usr/share/drkonqi/gdb/preamble.py though I'm not sure if it was for the first
or second drkonqi crash on plasmashell or drkonqi. plasmashell might've crashed
a second time when drkonqi ran /usr/share/drkonqi/gdb/preamble.py in gdb.

Dec 08 05:42:53 plasmashell[3633]: Traceback (most recent call last):
Dec 08 05:42:53 plasmashell[3633]:   File "<string>", line 1, in <module>
Dec 08 05:42:53 plasmashell[3633]:   File "/usr/share/drkonqi/gdb/preamble.py",
line 620, in print_preamble
Dec 08 05:42:53 plasmashell[3633]:     print_qml_trace()
Dec 08 05:42:53 plasmashell[3633]:   File "/usr/share/drkonqi/gdb/preamble.py",
line 578, in print_qml_trace
Dec 08 05:42:53 plasmashell[3633]:     ret = qml_trace_frame(frame)
Dec 08 05:42:53 plasmashell[3633]:           ^^^^^^^^^^^^^^^^^^^^^^
Dec 08 05:42:53 plasmashell[3633]:   File "/usr/share/drkonqi/gdb/preamble.py",
line 518, in qml_trace_frame
Dec 08 05:42:53 plasmashell[3633]:     dereferenced_type =
typeobj.target().unqualified()
Dec 08 05:42:53 plasmashell[3633]:                         ^^^^^^^^^^^^^^^^
Dec 08 05:42:53 plasmashell[3633]: KeyboardInterrupt
Dec 08 05:42:53 plasmashell[3633]: /tmp/drkonqi.wHnfVC:3: Error in sourced
command file:
Dec 08 05:42:53 plasmashell[3633]: Error while executing Python code.


I reported this problem at https://bugs.kde.org/show_bug.cgi?id=478287

-- 
You are receiving this mail because:
You are on the CC list for the bug.