From owner-freebsd-stable  Tue Jul  2 04:14:59 1996
Return-Path: owner-stable
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id EAA29168
          for stable-outgoing; Tue, 2 Jul 1996 04:14:59 -0700 (PDT)
Received: from robin.camelot.de (root@robin.camelot.de [194.97.87.3])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id EAA29163
          for <stable@FreeBSD.org>; Tue, 2 Jul 1996 04:14:55 -0700 (PDT)
Received: from lancelot.camelot.de(really [194.97.87.4]) by camelot.de
	via sendmail with smtp
	id <m0ua1yJ-00037cC@robin.camelot.de>
	for <stable@FreeBSD.org>; Sat, 29 Jun 96 17:30:07 +0200 (MET DST)
	(/\##/\ Smail3.1.30.13 #30.3 built 25-mar-96)
Received: from jelal.hb.north.de by camelot.de
	via rmail with uucp
	id <m0ua1yF-002eMjC@lancelot.camelot.de>
	for FreeBSD.org!stable; Sat, 29 Jun 96 17:30:03 +0200 (MET DST)
	(/\##/\ Smail3.1.30.13 #30.3 built 25-mar-96)
Received: by jelal.hb.north.de (SMail-ST 0.95gcc/2.5+)
	id AA00131; Sat, 29 Jun 1996 17:19:28 +0100 (CET)
Received: (from nox@localhost) by saturn.hb.north.de (8.7.5/8.7.3) id RAA00434; Sat, 29 Jun 1996 17:13:45 +0200 (MET DST)
From: Juergen Lock <nox@jelal.hb.north.de>
Message-Id: <199606291513.RAA00434@saturn.hb.north.de>
Subject: Re: ipfw (was: Re: lockups.)
To: alex@fa.tdktca.com (Alex Nash)
Date: Sat, 29 Jun 1996 17:13:44 +0200 (MET DST)
Cc: davidg@Root.COM, jhay@mikom.csir.co.za,
        nasim!FreeBSD.org!stable@jelal.hb.north.de
In-Reply-To: <31D446BA.9BDE134@fa.tdktca.com> from Alex Nash at "Jun 28, 96 03:55:22 pm"
X-Mailer: ELM [version 2.4ME+ PL19 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-stable@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Alex Nash writes:

> > > Nifty trick, I like it.  I just wanted to point out one thing: this
> > > is only necessary if you plan to boot kernels supped before Feb 23,
> > > not 0696.
> > 
> > Umm you are thinking of -current maybe? :)  or my -stable must have
> > had an, errm, interesting problem for quite a while now...
> >
> >  actually no (thanks cvs :), looks like current is also affected:
> 
> You mean -stable.  Yes, as of Monday this week, -current and -stable
> are in sync.

Actually i was talking about the (date of the) struct ip_fw change,
but nevermind.

> As far as accepting both versions: getting the new kernel to accept
> the old struct would be fairly easy, the new ip_fw struct has some
> extensions and two new flags, all easily defaulted.

 like... below? :)
> 
> Alex
 ATB,
	Juergen

Index: /www/usr/home/cvs/src/src/sys/netinet/ip_fw.c
===================================================================
RCS file: /home/cvs/cvs/src/sys/netinet/ip_fw.c,v
retrieving revision 1.14.4.9
diff -u -r1.14.4.9 ip_fw.c
--- ip_fw.c	1996/06/25 03:16:41	1.14.4.9
+++ ip_fw.c	1996/06/29 14:58:29
@@ -600,14 +600,31 @@
 {
 	struct ip_fw *frwl;
 
-	if (m->m_len != sizeof(struct ip_fw)) {
-		dprintf(("ip_fw_ctl: len=%d, want %d\n", m->m_len,
-		    sizeof(struct ip_fw)));
-		return (NULL);
+	static struct ip_fw ftmp64;
+	static int warned;
+	int f_mask = ~IP_FW_F_MASK;
+
+	/* backward compatibility hack for older ipfw(8)... */
+	if (m->m_len == 64 && sizeof(struct ip_fw) == 100) {
+		frwl = mtod(m, struct ip_fw *);
+		bcopy(frwl, &ftmp64, (size_t)64);
+		frwl = &ftmp64;
+		/* old IP_FW_F_MASK */
+		f_mask = ~0xfff;
+		if (!warned) {
+			dprintf(("ip_fw_ctl: old format ip_fw struct (converted), need to update ipfw(8)?\n"));
+			++warned;
+		}
+	} else {
+		if (m->m_len != sizeof(struct ip_fw)) {
+			dprintf(("ip_fw_ctl: len=%d, want %d\n", m->m_len,
+			    sizeof(struct ip_fw)));
+			return (NULL);
+		}
+		frwl = mtod(m, struct ip_fw *);
 	}
-	frwl = mtod(m, struct ip_fw *);
 
-	if ((frwl->fw_flg & ~IP_FW_F_MASK) != 0) {
+	if ((frwl->fw_flg & f_mask) != 0) {
 		dprintf(("ip_fw_ctl: undefined flag bits set (flags=%x)\n",
 		    frwl->fw_flg));
 		return (NULL);