From owner-freebsd-net  Wed Jan  8 14: 3:12 2003
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 C561A37B401; Wed,  8 Jan 2003 14:03:10 -0800 (PST)
Received: from mel-rto6.wanadoo.fr (smtp-out-6.wanadoo.fr [193.252.19.25])
	by mx1.FreeBSD.org (Postfix) with ESMTP
	id F30A843ED8; Wed,  8 Jan 2003 14:03:09 -0800 (PST)
	(envelope-from vjardin@wanadoo.fr)
Received: from mel-rta7.wanadoo.fr (193.252.19.61) by mel-rto6.wanadoo.fr (6.7.015)
        id 3E0C343F006924B8; Wed, 8 Jan 2003 23:03:09 +0100
Received: from there (217.128.206.18) by mel-rta7.wanadoo.fr (6.7.015)
        id 3E075B1B006D4334; Wed, 8 Jan 2003 23:03:09 +0100
Message-ID: <3E075B1B006D4334@mel-rta7.wanadoo.fr> (added by postmaster@wanadoo.fr)
Content-Type: text/plain;
  charset="iso-8859-1"
From: Vincent Jardin <vjardin@wanadoo.fr>
To: Jeffrey Hsu <hsu@FreeBSD.org>
Subject: Re: rt_refcnt and rtentry protection ??
Date: Wed, 8 Jan 2003 23:22:22 +0100
X-Mailer: KMail [version 1.3.2]
Cc: net@freebsd.org
References: <0H8E00ASUXB1HW@mta5.snfc21.pbi.net>
In-Reply-To: <0H8E00ASUXB1HW@mta5.snfc21.pbi.net>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: owner-freebsd-net@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-net.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-net>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-net>
X-Loop: FreeBSD.org


>
> The roadmap for locking up the IP stack looks like roughly like ifaddr
> ref counts, ifnet list, routing radix tree, ARP, ifaddr uses, and rtentry.

However, it seems that the rtentries' rt_refcnt are not used like ifaddr, 
ifnet, ... It looks like something dirty has been done a long time ago.
For example, the following lines:

        rt = rtalloc1((struct sockaddr *)&sin, create, 0UL);
        if (rt == 0)
                return (0);
        rt->rt_refcnt--;

        if (rt->rt_flags & RTF_GATEWAY)
                why = "host is not on local network";
        else if ((rt->rt_flags & RTF_LLINFO) == 0)
                why = "could not allocate llinfo";
        else if (rt->rt_gateway->sa_family != AF_LINK)
                why = "gateway route is not ours";
   [...]
        return ((struct llinfo_arp *)rt->rt_llinfo);

Why is rt_refnt decreased so early and not later ?

I would think that it would be more correct to have:
        rt = rtalloc1((struct sockaddr *)&sin, create, 0UL);
        if (rt == 0)
                return (0);

        if (rt->rt_flags & RTF_GATEWAY)
                why = "host is not on local network";
        else if ((rt->rt_flags & RTF_LLINFO) == 0)
                why = "could not allocate llinfo";
        else if (rt->rt_gateway->sa_family != AF_LINK)
                why = "gateway route is not ours";
   [...]
        rt->rt_refcnt--;

        return ((struct llinfo_arp *)rt->rt_llinfo);

It looks like rt_refcnt is decreased just after many rtalloc() in order just 
to be sure that it won't be forgotten later, doesn'it ?

> I'm a little over 3/5 of the way done along this path.

Good luck,
  Vincent

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message