From owner-freebsd-current Fri Dec 8 01:17:22 1995 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id BAA13662 for current-outgoing; Fri, 8 Dec 1995 01:17:22 -0800 (PST) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id BAA13647 for ; Fri, 8 Dec 1995 01:17:09 -0800 (PST) Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id UAA19607; Fri, 8 Dec 1995 20:12:45 +1100 Date: Fri, 8 Dec 1995 20:12:45 +1100 From: Bruce Evans Message-Id: <199512080912.UAA19607@godzilla.zeta.org.au> To: current@freebsd.org, scrappy@hub.org Subject: Re: /usr/src/lib/libc/gen/err.c patch Sender: owner-current@freebsd.org Precedence: bulk > Here's another one that failed with conflicting prototypes. >Does anyone read this? Or do these just go to /dev/null? *raised eyebrow* >*** lib/libc/gen/err.c.orig Thu Dec 7 21:32:37 1995 >--- lib/libc/gen/err.c Thu Dec 7 21:51:35 1995 >*************** >*** 91,97 **** > verr(eval, fmt, ap) > int eval; > const char *fmt; >! va_list ap; > { > int sverrno; > >--- 91,97 ---- > verr(eval, fmt, ap) > int eval; > const char *fmt; >! _BSD_VA_LIST_ ap; > { > int sverrno; > gcc's has had broken support for Net/2 and BSD4.4-derived systems for many years. It declares its own va_list type and doesn't attempt to match it with the version defined in (_VA_LIST_ in Net/2 and _BSD_VA_LIST_ in BSD4.4). It bogusly #undef's _BSD_VA_LIST, but this is harmless because the macro is actually spelled _BSD_VA_LIST_. It goes to a lot of trouble to avoid #undef'ing _VA_LIST_ for obsolete Net/2 systems. Bruce