Re: GCC vs GLIBC: why this stance, Drepper ?!?

["H . J . Lu" <hjl at lucon dot org>]

On Sun, Jul 08, 2001 at 11:09:13AM -0700, mike stump wrote:
> > Date: Sat, 30 Jun 2001 17:23:44 -0700
> > From: "Zack Weinberg" <zackw@Stanford.EDU>
> > To: gcc@gcc.gnu.org, libc-alpha@sources.redhat.com.com
> 
> > If you post a patch for GCC to gcc-patches, I hereby promise that I
> > will not ignore it.  (Be aware that I may have only sporadic net
> > access for the next month or so, however.)
> 
> There are some problems patches cannot fix, ever.  If an
> incompatibility is allowed to go in, and is then shipped, but caught
> after the release, one is screwed.  If you accept a patch to fix it,
> you restore binary compatibility with the older version, but you break
> compatibility with the newer version.  If you don't accept the patch,
> you maintain binary compatibility with the new version, but break with
> the old.
> 
> The game is rigged so one cannot win.  The only way to win, is to
> never, ever, do this.  That cannot be done, without extreme caution
> and control and knowledge and skill and attention.  To me, it sounds
> like this is one of HJ concerns.  HJ, if it isn't, apologies.

It is something like that. I think there is a way to do it. See

http://gcc.gnu.org/ml/gcc/2001-07/msg00524.html

My next goal is to provide a patch to gcc 3.0 so that glibc and gcc can
share the same mixed g++ v2 and v3 frame unwind code. It should be able
to be conditionally compiled for

1. Turn on/off of the mixed g++ v2 and v3 support.
2. In gcc or glibc.


H.J.