Red Hat Bugzilla – Bug 990679
[abrt] pinpoint-0.1.4-13.fc19: cogl_matrix_entry_ref: Process /usr/bin/pinpoint was killed by signal 11 (SIGSEGV)
Last modified: 2013-08-16 12:29:09 EDT
Description of problem:
I was working in a terminal with an unset $DISPLAY, and apparently pinpoint doesn't deal well with that. :) Exiting is fine, but segfaulting is probably not the desired behavior.
To reproduce: run pinpoint from a text console, or unset DISPLAY.
Version-Release number of selected component:
runlevel: N 5
Thread no. 1 (10 frames)
#0 cogl_matrix_entry_ref at ./cogl-matrix-stack.c:332
#1 cogl_matrix_stack_new at ./cogl-matrix-stack.c:662
#2 _cogl_framebuffer_init at ./cogl-framebuffer.c:119
#3 cogl_onscreen_new at ./cogl-onscreen.c:101
#4 clutter_stage_x11_realize at x11/clutter-stage-x11.c:589
#5 clutter_stage_realize at ./clutter-stage.c:767
#6 _g_closure_invoke_va at gclosure.c:840
#9 clutter_actor_realize at ./clutter-actor.c:1930
#10 _clutter_stage_manager_set_default_stage at ./clutter-stage-manager.c:226
#11 clutter_stage_constructed at ./clutter-stage.c:1682
Created attachment 781283 [details]
Created attachment 781284 [details]
Created attachment 781285 [details]
Created attachment 781286 [details]
Created attachment 781287 [details]
Created attachment 781288 [details]
Created attachment 781289 [details]
Created attachment 781290 [details]
Created attachment 781291 [details]
Created attachment 781292 [details]
Why is $DISPLAY unset?
But the issue remains: I doubt the desired behavior is dumping core when DISPLAY is unset. :)
OK, I wouldn't expect that to happen in a GNOME session where pinpoint is expected to run. That may be why upstream hasn't worried about it. Nevertheless I filed https://bugzilla.gnome.org/show_bug.cgi?id=706150 so upstream can decide whether to handle. This isn't something I'll handle downstream separately in Fedora, though, since it's truly a corner case.