On Sun, 16 May 2010 23:05:50 +0200 Andreas Volz <li...@brachttal.net> said:
good. pebkac :) good to know. > Am Sun, 2 May 2010 17:44:57 +0200 schrieb Andreas Volz: > > Hello, > > I only like to inform that I was able to fix that bug. It was for sure > an bug in my application. But from this backtrace I wasn't able to find > it. While hinting another bug I got a backtrace in my dispatcher and > found out that I do some Evas object stuff in the wrong thread. I > simply moved this code after dispatching in my ecore thread and it > works now great without any crashes. :-) > > regards > Andreas > > > Hello, > > > > I've a strange problem in a multi-threaded Edje application. And yes I > > know Edje isn't thread save. I dispatch all events to the main loop. > > It seems to work well in 9 of 10 cases. But sometimes I start my > > application and loading the first Edje file/group leads direct into an > > X assert: > > > > Filename: > > '/home/andreas/src/svn/oiui/trunk/data/themes/default/screens.edj', > > Groupname: navigation_main lt-oiui: ../../src/xcb_lock.c:77: > > _XGetXCBBuffer: Assertion `((int) ((xcb_req) - (dpy->request)) >= 0)' > > failed. > > > > Now I was able to hit this assert in a gdb session and got a stack > > trace: > > > > http://img638.imageshack.us/i/xassert.jpg/ > > > > Sorry, I had to make a screenshot, because Anjuta doesn't allow to > > save stack traces as text (bug is reported). :-) > > > > If I search this bug in google I see it's nothing unknown: > > > > http://tinyurl.com/33dutf3 > > > > Sounds like that it's not a X bug, but a usage bug. The only question > > is if it's a ecore/edje bug or a bug in my application. > > > > Do you've any ideas about this bug? > > > > regards > > Andreas > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > > enlightenment-devel mailing list > > enlightenment-devel@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > > > > ------------------------------------------------------------------------------ > > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) ras...@rasterman.com ------------------------------------------------------------------------------ _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
