From owner-freebsd-hackers Wed Jan 29 06:58:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA03599 for hackers-outgoing; Wed, 29 Jan 1997 06:58:34 -0800 (PST) Received: from nic.follonett.no (nic.follonett.no [194.198.43.10]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA03594 for ; Wed, 29 Jan 1997 06:58:24 -0800 (PST) Received: (from uucp@localhost) by nic.follonett.no (8.8.5/8.8.3) with UUCP id PAA11776; Wed, 29 Jan 1997 15:56:59 +0100 (MET) Received: from oo7 (oo7.dimaga.com [192.0.0.65]) by dimaga.com (8.7.5/8.7.2) with SMTP id PAA02810; Wed, 29 Jan 1997 15:53:06 +0100 (MET) Message-Id: <3.0.32.19970129155305.00ab11a0@dimaga.com> X-Sender: eivind@dimaga.com X-Mailer: Windows Eudora Pro Version 3.0 (32) Date: Wed, 29 Jan 1997 15:53:06 +0100 To: Brian Somers From: Eivind Eklund Subject: Re: ipdivert & masqd Cc: Archie Cobbs , hackers@freebsd.org, ari.suutari@ps.carel.fi, cmott@srv.net Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk At 08:13 AM 1/29/97 +0000, Brian Somers wrote: >[.....] >> >> Under which version(s) of FreeBSD are you guys having this problem ? >> I'm trying to track it down... >> > >I'm running 3.0-current and will test it on 2.2-961014-SNAP today. I believe >Ari is running 2.2(-BETA?) and suspect that Charles is running much the same >or maybe 2.1.6.... Charles is running 2.1.0 (or at least he was running 2.1.0 a week ago.), and is (in his own words) "too backwards to test [natd]" BTW: Both masqd and natd need to allocate a large enough buffer to handle PORT-commands being extended. This is done without a buffer check in alias_ftp.c; the maximum size it can be extended to is "PORT 123,123,123,123,123,123\r\n" - 32 characters (including termination) - an extension of 8 characters. The packet payload area _has_ to be large enough handle this. For the IRC DCC case (which I'll hopefully bring to testing point tomorrow - any volunteers?) the expansion can be by 11 characters for each DCC hook in a PRIVMSG, totalling max (payload size)*(4/3), ie expanding to 7/3 the original size for a constructed nasty case. This has bounds-checking, though, and will not do overwrites. Still, extra buffer-space do make it work more reliably. Eivind Eklund / perhaps@yes.no / http://maybe.yes.no/perhaps/