From owner-freebsd-hackers  Mon Feb 27 14:17:46 1995
Return-Path: hackers-owner
Received: (from majordom@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id OAA09354 for hackers-outgoing; Mon, 27 Feb 1995 14:17:46 -0800
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.9/8.6.6) with ESMTP id OAA09328 for <hackers@FreeBSD.org>; Mon, 27 Feb 1995 14:17:20 -0800
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id JAA20133; Tue, 28 Feb 1995 09:11:23 +1100
Date: Tue, 28 Feb 1995 09:11:23 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199502272211.JAA20133@godzilla.zeta.org.au>
To: alm@netcom.com, hasty@netcom.com
Subject: Re: patches for gcc-2.6.3?
Cc: hackers@FreeBSD.org, phk@ref.tfs.com
Sender: hackers-owner@FreeBSD.org
Precedence: bulk

>[objc build failure]

>I solved this by conditionally including <stddef.h> instead of
><machine/ansi.h> in <stdio.h>, <time.h>, <string.h> and <sys/types.h>

gcc-2.6's includes have broken support for 4.4lite stuff, mostly in
<stddef.h> and <stdarg.h>.  The best fix is not to use and gcc includes.
This requires some changes to the Makefiles because the standard
configuration wants to use them.

>...
>I don't claim this to be a fix so much as a hack.  But the problem
>it addresses (permanently undef'ing SIZE_T) does need fixing one way or
>another....

The brokenness involves gcc's includes not really understanding when
to undef things.

Bruce