From owner-freebsd-net@FreeBSD.ORG  Tue Jan 10 10:24:06 2012
Return-Path: <owner-freebsd-net@FreeBSD.ORG>
Delivered-To: freebsd-net@freebsd.org
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id E4A05106566B
	for <freebsd-net@freebsd.org>; Tue, 10 Jan 2012 10:24:06 +0000 (UTC)
	(envelope-from dkmail@mail.neveragain.de)
Received: from mail.neveragain.de (mail.neveragain.de [IPv6:2001:aa8:fffc::25])
	by mx1.freebsd.org (Postfix) with ESMTP id B23A38FC17
	for <freebsd-net@freebsd.org>; Tue, 10 Jan 2012 10:24:06 +0000 (UTC)
Received: by mail.neveragain.de (Postfix, from userid 1029)
	id 3C80817024; Tue, 10 Jan 2012 11:24:05 +0100 (CET)
Date: Tue, 10 Jan 2012 11:24:05 +0100
From: Dennis Koegel <dk@neveragain.de>
To: freebsd-net@freebsd.org
Message-ID: <20120110102405.GA82356@neveragain.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
User-Agent: Mutt/1.4.2.3i
Subject: Unnecessary sleep in network.subr: ipv6_up()
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: Tue, 10 Jan 2012 10:24:07 -0000

Cheers,

problem: Having a *lot* of IPv6 interfaces (Vlan interfaces in this case)
causes a huge and annoying delay time at system boot in 9.0R.

ipv6_up() in network.subr does this:

+	# wait for DAD
+	sleep `${SYSCTL_N} net.inet6.ip6.dad_count`
+	sleep 1

This happens for each and every interface, at a minimum (and default) of
two seconds per interface.

It seems the behaviour was introduced with r197139. Before this merge,
/etc/rc.d/network_ipv6 did the same sleeps, but only once for the whole
network startup.

I don't see why this should happen per interface, so I suggest the extra
sleeps are limited to "once per network startup" once again (or maybe
removed?).

- D.