From owner-freebsd-arch@FreeBSD.ORG  Tue Oct 12 15:44:10 2004
Return-Path: <owner-freebsd-arch@FreeBSD.ORG>
Delivered-To: freebsd-arch@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP
	id A3B9716A4CE; Tue, 12 Oct 2004 15:44:10 +0000 (GMT)
Received: from athena.softcardsystems.com (mail.softcardsystems.com
	[12.34.136.114])	by mx1.FreeBSD.org (Postfix) with ESMTP
	id 4993C43D41; Tue, 12 Oct 2004 15:44:10 +0000 (GMT)
	(envelope-from sah@softcardsystems.com)
Received: from athena (athena [12.34.136.114])i9CGgWlq031724;
	Tue, 12 Oct 2004 11:42:32 -0500
Date: Tue, 12 Oct 2004 11:42:32 -0500 (EST)
From: Sam <sah@softcardsystems.com>
X-X-Sender: sah@athena
To: =?iso-8859-1?q?Dag-Erling_Sm=F8rgrav?= <des@des.no>
In-Reply-To: <xzpy8ict0cy.fsf@dwp.des.no>
Message-ID: <Pine.LNX.4.60.0410121136540.30953@athena>
References: <Pine.LNX.4.60.0410121115430.30953@athena>
	<xzpy8ict0cy.fsf@dwp.des.no>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
cc: freebsd-current@freebsd.org
cc: freebsd-arch@freebsd.org
Subject: Re: mbuf w/o pkthdr?
X-BeenThere: freebsd-arch@freebsd.org
X-Mailman-Version: 2.1.1
Precedence: list
List-Id: Discussion related to FreeBSD architecture
	<freebsd-arch.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-arch>,
	<mailto:freebsd-arch-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-arch>
List-Post: <mailto:freebsd-arch@freebsd.org>
List-Help: <mailto:freebsd-arch-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-arch>,
	<mailto:freebsd-arch-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Tue, 12 Oct 2004 15:44:10 -0000

>> Are all packets supposed to have the M_PKTHDR flag?  Why?
>
> IIRC, M_PKTHDR indicates the first mbuf in a chain when a packet is
> split across multiple mbufs.  This usually only happens for outgoing
> packets, where protocol headers are constructed in separate mbufs
> which are prepended to the chain as the packet moves down the stack.

That's kind of my understanding (with some PACKET_TAG* stuff going on).

I don't have split headers, though.  Neither would an arp frame, but
he too gets a packet header and fills it out.

I don't mind following suit, I'm just wondering what the convention is
for.  Perhaps we use mbufs without packet headers for something special?

Sam