Need external opinion about backtrace

[CP Hennessy]

In KNode there are 8 crash reports  associated with the QtGDict and
QtRegex classes. 
For any I've looked at the backtrace did not even seem to be correct
as it seemed to indicate that methods were being called which were
not used by the caller. e.g.
http://bugs.kde.org/buglist.cgi?product=knode&long_desc_type=anywordssubstr&long_desc=QRegex+QGDict&bug_status=UNCONFIRMED&bug_status=NEW&bug_severity=crash
I asked on a qt mailing list about 6-8 weeks ago and was told that 
it was probably a problem with the compiler, but I was not able to go any 
further :(
However for any bug where the compiler was indicated it was always 2.95.3.

CPH

On Wednesday 06 November 2002 13:35, Michael Goffioul wrote:
> I received a bug report about kprinter crashing with Alarm clock
> signal. The backtrace is below. I doubt the bug lies in kdeprint's
> code as the crash occurs somewhere else. I'm wondering about the
> compiler (2.95.3). Any opinion?
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x40840a1c in free () from /usr/kde/3.1/lib/libkdecore.so.4
> (gdb) bt
> #0  0x40840a1c in free () from /usr/kde/3.1/lib/libkdecore.so.4
> #1  0x410ea390 in bad_alloc virtual table ()
>    from /usr/lib/gcc-lib/i686-pc-linux-gnu/2.95.3/libstdc++-libc6.2-2.so.3
> #2  0x40d42bc2 in QGDict::hashKeyString () from /usr/qt/3/lib/libqt-mt.so.3
> #3  0x40d57549 in QStringData::deleteSelf () from
> /usr/qt/3/lib/libqt-mt.so.3 #4  0x4076055a in KStandardDirs::resourceDirs
> ()
>    from /usr/kde/3.1/lib/libkdecore.so.4
> #5  0x4075c3d0 in KStandardDirs::findResourceDir ()
>    from /usr/kde/3.1/lib/libkdecore.so.4
> #6  0x4075b6c8 in KStandardDirs::findResource ()
>    from /usr/kde/3.1/lib/libkdecore.so.4
> #7  0x08049bf8 in ?? ()
> #8  0xbfffef10 in ?? ()
> #9  0x080f6e18 in ?? ()
> Cannot access memory at address 0x3
>
> Michael.