From owner-freebsd-arch  Mon Jul  9  5:59: 3 2001
Delivered-To: freebsd-arch@freebsd.org
Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16])
	by hub.freebsd.org (Postfix) with ESMTP id B637A37B403
	for <arch@FreeBSD.ORG>; Mon,  9 Jul 2001 05:58:57 -0700 (PDT)
	(envelope-from bde@zeta.org.au)
Received: from bde.zeta.org.au (bde.zeta.org.au [203.2.228.102])
	by mailman.zeta.org.au (8.9.3/8.8.7) with ESMTP id WAA19888;
	Mon, 9 Jul 2001 22:58:50 +1000
Date: Mon, 9 Jul 2001 22:56:45 +1000 (EST)
From: Bruce Evans <bde@zeta.org.au>
X-Sender: bde@besplex.bde.org
To: Harti Brandt <brandt@fokus.gmd.de>
Cc: arch@FreeBSD.ORG
Subject: Re: -fno-builtin
In-Reply-To: <Pine.BSF.4.33.0107091420430.17758-100000@beagle.fokus.gmd.de>
Message-ID: <Pine.BSF.4.21.0107092242340.87868-100000@besplex.bde.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: owner-freebsd-arch@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-arch.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo?subject=subscribe%20freebsd-arch>
List-Unsubscribe: <mailto:majordomo?subject=unsubscribe%20freebsd-arch>
X-Loop: FreeBSD.ORG

On Mon, 9 Jul 2001, Harti Brandt wrote:

> On Mon, 9 Jul 2001, Bruce Evans wrote:
> 
> BE>I think it would work right except in broken code which "knows" that
> BE>strlen isn't a macro, e.g.:
> BE>
> BE>	size_t (*p)(const char *s) = strlen;
> 
> Why is that broken? My last copy of the ISO-C draft specifically states,
> that strlen is a function. (Not that I intend to use that construct. Just
> beeing curious).

Does it really?  Almost any library function can be implemented as a
(safe) macro (see section 7.1.4).  I can't find any exception for
strlen.  This possibility is not completely transparent to applications.
Code like the above should be aware of it.  Workarounds are easy in
the above (just used "&strlen" or "(strlen)".  Section 7.1.4 documents
these.  Howver, I now remember a nasty example which shows why
implementations should not do this:

	len = strlen(
#ifdef FOO
	    "foo"
#else
	    "default"
#endif
	    );

This causes errors like:

z.c:11: warning: preprocessing directive not recognized within macro arg
z.c:11: warning: preprocessing directive not recognized within macro arg
z.c:11: warning: preprocessing directive not recognized within macro arg
z.c: In function `main':
z.c:6: undefined or invalid # directive
z.c:8: undefined or invalid # directive
z.c:10: undefined or invalid # directive

I only learned of this example a weeks or two ago.  gcc-3.0 was reported
to make a very fundamental function (printf?) a macro.

Bruce


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-arch" in the body of the message