Date: Sun, 8 Aug 1999 14:02:47 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Alain Magloire <alainm AT rcsm DOT ece DOT mcgill DOT ca>
cc: djgpp-workers AT delorie DOT com
Subject: Re: CPU identification (Was: Re: uname -m ?)
In-Reply-To: <199908060116.VAA02276@mccoy2.ECE.McGill.CA>
Message-ID: <Pine.SUN.3.91.990808140229.23423M-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Thu, 5 Aug 1999, Alain Magloire wrote:

> I actually think it is a good idea to add this extension via
> a new switch for example '-x' or -f etc .. and a new member name
> to utsname.

Switches for `uname' program is something that should be suggested to
Jim Meyering, the maintainer of Sh-utils.  Unlike Sun, we don't have
our own version of `uname'.

As for a new member of `struct utsname', I'm not particularly against
it, although we should at least go out and see if some other platform
adds that, and if so, make the additional member have the same name.
Also, if detecting x87 could be dangerous (i.e. risk crashing the
program), I would vote against it.

> config.sub, config.guess are merely aids to help a maintainer
> to figure out the type of the system.

Yes, but given the trail of trouble we had (and still have) with
making config.guess work with DJGPP, I'd hate to break it.  We still
have packages where the only way to reliably configure them is to pass
an explicit --host parameter to configure.