From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Denis V. Lunev" Subject: Re: IPv6 unique local address flushes on up/down Date: Tue, 25 Mar 2008 18:11:38 +0300 Message-ID: <1206457898.14357.16.camel@iris.sw.ru> References: <200803251557.21563.chunkeey@web.de> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: hostap@lists.shmoo.com, netdev@vger.kernel.org To: Chr Return-path: Received: from mailhub.sw.ru ([195.214.232.25]:23695 "EHLO relay.sw.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754174AbYCYPLQ (ORCPT ); Tue, 25 Mar 2008 11:11:16 -0400 In-Reply-To: <200803251557.21563.chunkeey@web.de> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, 2008-03-25 at 15:57 +0100, Chr wrote: > Hi, > > I have a problem with hostapd & IPv6... > > as you might know, or not... hostapd needs to initialize > the interface (e.g it changes it to master/accesspoint mode, > which can only be done if the interface is down)... > > However if the interface is brought down & up again (e.g whenever > hostap starts or restarts)... the IPv6 logic flushes the unique-local ipv6 > address that I gave it (by hand) with ip addr add fddb:xyz...::1//64 dev ath0 > in /etc/network/interfaces ifup script (-probably debian specific-). > > And that's bad, because bind9,ISC dhcp-server v4.1 and racoon(IPsec/VPN) > needs a static address for their records/lease database and configuration... > > This wasn't problem with the old IPv4... because there the > 10.x.y.z or 192.168.a.b addresses aren't deleted when the interface > was _reinitialized_ by the hostap... > > Now my question, which _one_ should we fix the applications or the stack?! > I think the network stack, but I don't know if there are RFC about this > issue... or is there already some flag/setting which I've missed so far? On interface down IPv6 stack unconditionally deletes all addresses from the interface in the addrconf_ifdown routine.