Date: Wed, 3 Mar 1999 19:34:18 +0000
From: Mark Brown <broonie AT tardis DOT ed DOT ac DOT uk>
To: "'pgcc AT delorie DOT com'" <pgcc AT delorie DOT com>
Subject: Re: Intel/Cygnus
Message-ID: <19990303193418.B19013@tardis.ed.ac.uk>
Mail-Followup-To: "'pgcc AT delorie DOT com'" <pgcc AT delorie DOT com>
References: <36DD6D94 DOT 79AFEC8F AT mitre DOT org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mutt 0.95.3i
In-Reply-To: <36DD6D94.79AFEC8F@mitre.org>; from Philip Long on Wed, Mar 03, 1999 at 12:12:52PM -0500
X-Cookie: All true wisdom is found on T-shirts.
X-WWW-Homepage: http://www.tardis.ed.ac.uk/~broonie/
Reply-To: pgcc AT delorie DOT com

On Wed, Mar 03, 1999 at 12:12:52PM -0500, Philip Long wrote:
>     I read a news story a while back stating that Intel was going to
> help cygnus with pentium MMX/KNI etc. optimizing compilers.

>     Is that for egcs?  Does it have any relation to the pgcc patch.  For

I don't know, but I imagine it would be based on egcs.  Marc?

> that matter, why isn't pgcc merged into the egcs tree anyway?

To avoid hurting non Pentium machines.  PGCC optimizatons are often
Pentium-specific, and may produce worse code on other architectures.
Before they can be introduced into EGCS they need to be rewritten more
generally.

In addition, the PGCC source needn't even build on other architectures.
Current snapshots probably don't build on any Intel boxes - they
certainly don't build on Alpha.  It's intended to be portable, but it
may not always be portable.

-- 
Mark Brown  mailto:broonie AT tardis DOT ed DOT ac DOT uk   (Trying to avoid grumpiness)
            http://www.tardis.ed.ac.uk/~broonie/
EUFS        http://www.eusa.ed.ac.uk/societies/filmsoc/