(Thomas, sorry for the dup mail, I'm getting sick of GMail) 2011/11/25 Thomas Lübking <thomas.luebk...@gmail.com>: > Am Fri, 25 Nov 2011 23:11:33 +0100 > schrieb Dario Freddi <drf54...@gmail.com>: > >> 2011/11/25 Dario Freddi <drf54...@gmail.com>: >> > might as well be KCrash misuse. >> >> So it was, commit 3a9906d2b5d1c504670442277906ed53cd9686e8, game over. > While it's great that you spotted and fixed it (looking into the kcrash > sources confirms there's an arg access) - do you have any idea why it > doesn't segfault here at all? > (tested debug & release builds of polkit-kde, added usleep(100000) after > kcrash flag setting. No problem.
Now that's a very good question, since it behaved correctly on my 4.7 installation as well. I don't have enough knowledge about KCrash and KCmdLine internals to give out an informed opinion, and I solved the issue exactly through spotting that very accessor, so I guess I'm not the best person to answer that. Would be cool to have this investigated though, I will try to see if the behavior matches on KDE/4.7 latest and if I can go somewhere from there. > > kdelibs is compiled w/o mode (neither debug nor release) and i've > restarted that polkit-kde version at least a dozen times. > > @Zé: did you compile kdelibs in release mode or maybe even -O3? > > Cheers, > Thomas > >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
