From owner-freebsd-net@FreeBSD.ORG  Wed Aug 31 19:59:47 2005
Return-Path: <owner-freebsd-net@FreeBSD.ORG>
X-Original-To: freebsd-net@freebsd.org
Delivered-To: freebsd-net@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 7E80C16A41F
	for <freebsd-net@freebsd.org>; Wed, 31 Aug 2005 19:59:47 +0000 (GMT)
	(envelope-from emaste@phaedrus.sandvine.ca)
Received: from mailserver.sandvine.com (sandvine.com [199.243.201.138])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 119B543D46
	for <freebsd-net@freebsd.org>; Wed, 31 Aug 2005 19:59:46 +0000 (GMT)
	(envelope-from emaste@phaedrus.sandvine.ca)
Received: from labgw2.phaedrus.sandvine.com ([192.168.3.11]) by
	mailserver.sandvine.com with Microsoft SMTPSVC(5.0.2195.6713); 
	Wed, 31 Aug 2005 15:59:45 -0400
Received: by labgw2.phaedrus.sandvine.com (Postfix, from userid 12627)
	id 5F9F213646; Wed, 31 Aug 2005 15:59:45 -0400 (EDT)
Date: Wed, 31 Aug 2005 15:59:45 -0400
From: Ed Maste <emaste@phaedrus.sandvine.ca>
To: freebsd-net@freebsd.org
Message-ID: <20050831195945.GA22805@sandvine.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.4.2.1i
X-OriginalArrivalTime: 31 Aug 2005 19:59:45.0609 (UTC)
	FILETIME=[88DE3B90:01C5AE66]
Subject: BIOCSSEESENT ioctl not honoured for single-mbuf packets
X-BeenThere: freebsd-net@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Networking and TCP/IP with FreeBSD <freebsd-net.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-net>,
	<mailto:freebsd-net-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-net>
List-Post: <mailto:freebsd-net@freebsd.org>
List-Help: <mailto:freebsd-net-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-net>,
	<mailto:freebsd-net-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Wed, 31 Aug 2005 19:59:47 -0000

A coworker of mine discovered a bug in bpf with BIOCSEESENT.
In sys/net/bpf.c, bpf_mtap() does

        if (pktlen == m->m_len) {
                bpf_tap(bp, mtod(m, u_char *), pktlen);
                return;
        }

        BPFIF_LOCK(bp);
        LIST_FOREACH(d, &bp->bif_dlist, bd_next) {
                if (!d->bd_seesent && (m->m_pkthdr.rcvif == NULL))
                        continue;
        [...]

The pktlen == m->m_len is an optimization for the case where
the entire packet is in a single mbuf, added in version 1.95.
However, bd_seesent then isn't checked so all packets will be
seen.

In order to make bpf_tap work correctly with BIOCSEESENT for
both this case and the few drivers that use it, I think it needs
another argument to indicate if the packet is being sent or not.
Is it possible to change the API for bpf_tap?  Or add a bpf_tap2
that includes the flag, and make bpf_tap call it (for any third
party drivers using bpf_tap)?

--
Ed Maste
Sandvine Incorporated