Subject: Re: Non-ANSI ism in system headers
To: Perry E. Metzger <>
From: Bill Studenmund <>
List: current-users
Date: 04/25/2005 23:54:27
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Tue, Apr 26, 2005 at 01:53:29AM -0400, Perry E. Metzger wrote:
> Bill Studenmund <> writes:
> >
> > Not really. We used to, but there are too many places in the code that=
> > require gcc extensions.
> That's certainly true of things like the kernel (where we absolutely
> need __asm etc.) but is that true of our headers? (Of course, the
> __rename hack is kind of critical...)

I think that your () indicates we do really need gcc, or at least some of=
the common extensions.

However it would be interesting to see how we do when the header defines=20
for ANSI C are turned on. I suspect we will do much better in that case;=20
the headers assume gcc, but will work in other environments. We will=20
probably still need __rename() though...

Take care,


Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.2.3 (NetBSD)