Path: utzoo!attcan!uunet!lll-winken!ncis.llnl.gov!helios.ee.lbl.gov!pasteur!agate!bionet!csd4.milw.wisc.edu!nic.MR.NET!umn-d-ub!rutgers!cmcl2!adm!smoke!gwyn
From: gwyn@smoke.BRL.MIL (Doug Gwyn )
Newsgroups: comp.lang.c
Subject: Re: `va_dcl' macro question
Message-ID: <9507@smoke.BRL.MIL>
Date: 26 Jan 89 22:41:11 GMT
References: <1964@kappl.cs.vu.nl>
Reply-To: gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 19

In article <1964@kappl.cs.vu.nl> maart@cs.vu.nl (Maarten Litmaath) writes:
>Anybody got an example (of some weird architecture) to show why it has been
>decided that no semicolon shall follow `va_dcl'?

No, but it's always been that way and nobody is going to change it now.

>	int	printf(fmt, va_alist)
>	char	*fmt;
>	va_dcl;
>	{
>		...
>	}

Apart from the question of the semicolon, this is wrong anyway.
It's not permissible to mix fixed parameter declarations with the
va_alist.  va_alist must stand alone as the only parameter, with
known parameters then being fetched from it via va_arg.  This of
course applies only to varargs; in ANSI C, one has at least one
fixed parameter and uses the ,... notation.