From owner-freebsd-current  Thu Dec  4 17:20:54 1997
Return-Path: <owner-freebsd-current>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.7/8.8.7) id RAA22958
          for current-outgoing; Thu, 4 Dec 1997 17:20:54 -0800 (PST)
          (envelope-from owner-freebsd-current)
Received: from ns1.yes.no (ns1.yes.no [195.119.24.10])
          by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id RAA22941
          for <current@FreeBSD.ORG>; Thu, 4 Dec 1997 17:20:47 -0800 (PST)
          (envelope-from eivind@bitbox.follo.net)
Received: from bitbox.follo.net (bitbox.follo.net [194.198.43.36])
	by ns1.yes.no (8.8.7/8.8.7) with ESMTP id BAA04028;
	Fri, 5 Dec 1997 01:20:36 GMT
Received: (from eivind@localhost)
	by bitbox.follo.net (8.8.6/8.8.6) id CAA27250;
	Fri, 5 Dec 1997 02:20:26 +0100 (MET)
To: Ari Suutari <ari@suutari.iki.fi>
Cc: John Kelly <jak@cetlink.net>, current@FreeBSD.ORG
Subject: Re: 3.0 -release ?
References: <Pine.BSF.3.96.971204115814.335A-100000@kn6-045.ktvlpr.inet.fi>
From: Eivind Eklund <perhaps@yes.no>
Date: 05 Dec 1997 02:20:24 +0100
In-Reply-To: Ari Suutari's message of Thu, 4 Dec 1997 12:00:13 +0200 (EET)
Message-ID: <86k9dkegbr.fsf@bitbox.follo.net>
Lines: 37
X-Mailer: Gnus v5.4.52/XEmacs 20.2
Sender: owner-freebsd-current@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Ari Suutari <ari@suutari.iki.fi> writes:
> On Wed, 3 Dec 1997, Eivind Eklund wrote:
> > 
> > Problem both places, methinks.  See if the following fixes it (from
> > aage@aage.priv.no):
> > 
> > --- /tmp/natd.c Tue Dec  2 22:24:53 1997
> > +++ /usr/src/usr.sbin/natd/natd.c       Wed Oct 29 22:11:32 1997
> > @@ -1332,7 +1332,7 @@
> >  
> >         port = strtol (str, &end, 10);
> >         if (end != str)
> > -               return htonl (port);
> > +               return htons (port);
> >  
> >         sp = getservbyname (str, proto);
> >         if (!sp)
> > 
> > If it does, I'll commit it - it looks correct, and I've got one other
> > favourable report :-)
> > 
> 
> 	Hi,
> 
> 	After making this change, it no longer worked 
> 	on 2.2, I guess that it won't work on 3.0 either
> 	(I didn't have time to check it any futher, I just
> 	tried it).

Weird.  I've had reports from two users (one that is using natd and
had problems, and Aage Roebekk that actually fixed it) that this
solves a genuine problem.  

Using htonl() on a port number at least seems to be wrong - port
numbers are 16 bits.  Signed/unsigned trouble too, perhaps?

Eivind.