This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Virtual Base Bug
- To: law at cygnus dot com
- Subject: Re: Virtual Base Bug
- From: hjl at lucon dot org (H.J. Lu)
- Date: Mon, 8 Nov 1999 13:48:22 -0800 (PST)
- Cc: martin at mira dot isdn dot cs dot tu-berlin dot de (Martin v. Loewis),hjl at lucon dot org, oliva at lsd dot ic dot unicamp dot br, mgb at yosemite dot net,gcc-bugs at gcc dot gnu dot org
>
>
> In message <199911081741.SAA00646@mira.isdn.cs.tu-berlin.de>you write:
> > > I am not sure Linux can afford the glibc ABI change again unless we can
> > > make it invisible to user.
> >
> > I fully agree.
> Right. And that is one of the reasons why we have not addressed this problem.
>
> We had originally planned to include Martin's fixes for dynamic thunks in
> the next major gcc release, along with all the other changes which are going
> to break compatibility.
>
I believe the reason you did
Sat Jul 17 23:49:59 1999 Jeffrey A Law (law@cygnus.com)
* Makefile.in (INTERFACE): Bump to 3.
was the C++ ABI/API was changed slightly. I don't see any reason not
to include Martin's fix unless we have to make an compatible change
to the glibc ABI.
> However, the impression I've got is that thunks are on their way out as
> part of the ia64 C++ ABI standardization. If that is indeed the case then
> we need to think very seriously about whether or not we want to use similar
> standards for ia32 and other ports. Which in turn means we may not want to
> invest into fixing dynamic thunks at this time. Particularly since they are
> an ABI change.
>
May I ask a question, are we willing to fix it for Linux in gcc 2.95.3?
Pleae don't tell me it is a major change. I don't want people to think
C++ is broken on Linux.
--
H.J. Lu (hjl@gnu.org)