21.4.19 with gcc 4.1

[Ville Skyttä]

On Mon, 2006-04-10 at 11:32 -0600, Jerry James wrote:
> Ville Skyttä <scop at xemacs.org> wrote:
> > A somewhat minimal reproducer for one of the reported crashes (also with
> > vanilla 21.4.19) is this in init.el:
> >
> >   (require 'un-define)
> >   (set-default-coding-systems 'utf-8)
> >   (set-coding-priority-list '(utf-8))
> >   (set-coding-category-system 'utf-8 'utf-8)
> >
> > ...LANG=en_US.UTF-8 and XEmacs built with:
> >
> >   export CFLAGS="-O2 -g -fstack-protector"
> >   ./configure --pdump --with-mule --without-wnn
> >
> > And to trigger the crash, M-x manual-entry RET netstat RET
> 
> So I lied.  I'm just getting around to looking at this today.  Note
> that, when compiled with gcc 4.1 but without -fstack-protector, the
> above recipe sends XEmacs 21.4.19 into an apparently infinite loop.  It
> gets stuck in re_search_2(), in the loop starting at line 4177.

Yep, there have been some similar comments/reports over at
bugzilla.redhat.com, even against builds done with gcc 3.2.

Let me know if you come up with something, I'd be happy to help with
testing stuff.