GNOME Bugzilla – Bug 300421
gnometris crash at startup (seems pretty much immediate)
Last modified: 2012-01-31 23:23:25 UTC
Distribution/Version: garnome Dont spend any great amount of time on this, there may just be something wrong with my garnome setup but I decided it was best to report the crash and provide a backtrace and let you decide if it was worth looking at any closer GNU gdb 6.0-2mdk (Mandrake Linux) Copyright 2003 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i586-mandrake-linux-gnu"...Using host libthread_db l ibrary "/lib/tls/libthread_db.so.1". (gdb) run Starting program: /tmp/garnome/bin/gnometris [Thread debugging using libthread_db enabled] [New Thread 1090022624 (LWP 18284)] Detaching after fork from child process 18287. Program received signal SIGSEGV, Segmentation fault.
+ Trace 58103
Thread 1090022624 (LWP 18284)
Alan: 3 things, and I'm really just doing off the wall guesses here... :) 1) Is this ubuntu? 2) Run from console. Does it quit with a warning "(pango_color_parse): assertion `spec != NULL' failed"? 3) Does this happen every time you start the game?
1) it is garnome (i wrote it above and again in the distribution details box) garnome on top of Mandarake 10.0 not that it should make a difference. 3) Yes, crashes every time. 2) will try and check this later
Considering all the goofy bugs I've seen lately (mostly outside of gnome-games) stemming from minor misbehavior from something deep inside Ubuntu/Gentoo/FCtest(GCC4) lately, #1 has been worth asking and makes a difference. :)
Well, gnometris can crash if gdk_color_parse (in Tetris::initoptions()) is given NULL for the bgcolour... Doesn't seem like that's possible though. When looking up the color, gconf will only return NULL on an error. Gnometris looks for that and would replace NULL with "Black". Alan: 4) "gconftool-2 -g /apps/gnometris/options/bgcolour"
No worries, I do something similar to anytime I think I'm dealing with a Debian users, especially if they dare have the hypocrisy to complain about something not being easy enough (it is like complaining that the vegetarian special isn't meaty enough). 2) Segemenation Fault That is all I get, everytime, no additional information.
Mahjongg is hosed as well, immediate segfault. In case it might be interacting badly removed ~/.gconf/apps/mahjongg but it made no difference. I'll probably just have a big clear out and rebuild next time there is a new release of Garnome.
Alan: Go ahead and report your Mahjongg bugs and any others as new ones. If fixable, they will be fixed.
Not sure what I changed but the Mahjongg sefault no longer occurs and instead it fails more gracefully with an error message from GConf.
The gnometris problem is the same gconf bug as discussed in bug 300423 (and I'm willing to bet that mahjongg problem is similar). Basically pango_color_parse () can only crash if it is passed a pointer that is neither valid nor NULL. Something deep down in gconf is going really, really, bad. (It probably isn't gnometris since there isn't any gap for anything to stomp over the value between it being set and used.) I am marking this a dup of 300423 (forward duping since I discussed the problem more there). *** This bug has been marked as a duplicate of 300423 ***
This bug is being reassigned to the "quadrapassel" component so we can close the gnometris bugzilla component. Apologies for the mass email!