dynamic cast bug.

[Karl Nelson]

> > Okay, I didn't see it on the list of well known bugs and it is pretty
> > devistating to my application so I thought I should send it in.
> > Hmm, how could it be Linux specific if the bug appears on Sun, HPUX and
> > IRIX systems running the native OS?
> 
> If it does indeed, then it is not the bug I'm talking about; I'll
> analyse it in more detail.

Okay then I would assume either it is a different bug or the bug spread.
The original report was from a Sun box and I am working on an HPUX 
machine.

elm:/tmp> uname -a
HP-UX elm B.10.20 A 9000/785 2007604323 two-user license
elm:/tmp> c++ -v
Reading specs from /opt/pkg/egcs-1.1/hp_ux102/lib/gcc-lib/hppa1.1-hp-hpux10.20/egcs-2.91.60/specs
gcc version egcs-2.91.60 19981201 (egcs-1.1.1 release)
elm:/tmp> c++ -O0 -g3 -Wall bug.cc
elm:/tmp> ./a.out 
0x400047b0 0x400047b0 0x400047b8
0x400047b0 0x400047b0 0x400047b0
0x400047b0 0x400047b0 0x400047b0

Definitely not a linux nor even an elf binary system.
Same result on my home PC.

eve:~> c++ -v
Reading specs from /usr/lib/gcc-lib/i386-linux/2.95.2/specs
gcc version 2.95.2 20000220 (Debian GNU/Linux)


Also it seems unlikely that this is a thunk bug because it was also
found on a 2.8.1 gcc which used a different ABI than the current gcc
if I understand correctly.  But then since I have never altered the C++ 
front end of gcc I can't be sure.  Have you been about to replicate it
on a non-linux machine?

Hope it helps.

--Karl