From kde-devel Thu Aug 23 03:34:41 2001 From: Waldo Bastian Date: Thu, 23 Aug 2001 03:34:41 +0000 To: kde-devel Subject: Re: Backtrace analysis question - follow up X-MARC-Message: https://marc.info/?l=kde-devel&m=99853796204352 Make sure to compile the kdelibs with --enable-debug and you will have line numbers in the libs as well. If you get a crash in KURL::reset() it probably means that you either pass corrupt data or have caused memory corruption earlier already. On Wednesday 22 August 2001 05:58 pm, Julian Rockey wrote: > Of course, I just realised the below looks like a very stupid question.. > the source line numbers are right there for kfssite.cpp (my program). What > I really meants was for library functions where the line number isn't > reported! So for below, how do I find out at exactly which call in > KURL::reset() that the QString::operator()= was called that ended up with > the crash. Er.. if that makes sense. > > ---------- Forwarded Message ---------- > > Subject: Backtrace analysis question > Date: Wed, 22 Aug 2001 18:52:29 -0600 > From: Julian Rockey > To: > > Can someone tell me how a backtrace address can be traced to a point in the > source? > > To clarify, in this section of a backtrace: > > #4 0x40d73b68 in killpg () from /lib/libc.so.6 > #5 0x407e4c7b in QString::operator= () from /usr/lib/qt2/lib/libqt.so.2 > #6 0x405d4603 in KURL::reset () from > /usr/dist/kde/kde2.2/lib/libkdecore.so.3 #7 0x405d60f4 in KURL::operator= > () > from /usr/dist/kde/kde2.2/lib/libkdecore.so.3 > #8 0x0804f74f in KFSSite::KFSSite (this=0x818c7c8) at kfssite.cpp:25 > > I can work out, for example, that the crash occured when the KFSSite > constructor called the KURL operator=() method, and so on. But how can I > find out exactly where? How can the address x405d60f4 be mapped back to the > source code? > > thanks! > Julian > > ------------------------------------------------------- > > >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to > >> unsubscribe << -- KDE 2.2: We deliver. >> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<