From mboxrd@z Thu Jan 1 00:00:00 1970 From: YOSHIFUJI Hideaki Subject: Re: [PATCH] ipv6: do not add link-local address if one already exists Date: Mon, 11 Aug 2014 09:27:50 +0900 Message-ID: <53E80E06.3070704@miraclelinux.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Cc: hideaki.yoshifuji@miraclelinux.com, "David S. Miller\"" , Alexey Kuznetsov , Hideaki YOSHIFUJI , Patrick McHardy To: Roy Marples , netdev@vger.kernel.org Return-path: Received: from exprod7og105.obsmtp.com ([64.18.2.163]:47042 "HELO exprod7og105.obsmtp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1751407AbaHKA1z (ORCPT ); Sun, 10 Aug 2014 20:27:55 -0400 Received: by mail-pd0-f175.google.com with SMTP id r10so9709778pdi.6 for ; Sun, 10 Aug 2014 17:27:55 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Hello. Roy Marples wrote: > Hi > > Currently the kernel will always add an IPv6 link-local address > based on the hardware address when the interface is brought up. > This is probably based on the assumption that userland would > never add one before the interface is brought up. > > However, one at least one userland application (dhcpcd) does this so > it can implement RFC7217 which can be used for link-local addresses > as well. > > Attached is a patch which checks to see if a link-local address exists > before indiscriminately adding one. Well, I think this is incomplete, for example, we will continue generating global address from eui64. I think we should implement such functionality through "if token" bits. --yoshfuji