fix for PR 4447: is this really correct?

[Joe Buck]

This issue is moot, Mark.  Kriang already pointed out that the odd
behavior is evidently mandated by the ABI standard.

> --On Friday, November 30, 2001 03:21:50 PM -0800 mike stump 
> <mrs@windriver.com> wrote:
> 
> >> From: Joe Buck <jbuck@synopsys.COM>
> >> To: gcc@gcc.gnu.org, lerdsuwa@users.sourceforge.net
> >> Date: Fri, 30 Nov 2001 11:20:51 -0800 (PST)
> >
> >> Mark approved, assuming the usual testing requirements are met.
> >
> > :-(
> 
> I assume that if the patch went in on the mainline it was
> already reviewed for correctness.
> 
> >> I've now verified that this fix doesn't break any C++ or libstdc++
> >> tests (other tests aren't relevant since this only affects cc1plus).
> >
> > If I understand the fix, it is worse than not having it, as it hides a
> > real bug?
> >
> >> But I am now not sure that this fix is quite correct, though it does
> >> improve things.
> >
> > I think the ICE is preferable, as otherwise you have to explain that
> > you have to break the ABI, which is worse.
> 
> Did this change affect the mangling of functions that we were already
> able to compile successfully?
> 
> --
> Mark Mitchell                   mark@codesourcery.com
> CodeSourcery, LLC               http://www.codesourcery.com
>