From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?iso-8859-1?Q?Fran=E7ois-Xavier_Le_Bail?= Subject: Re: [PATCH net-next] IPv6: add option to use anycast addresses as source addresses in icmp error messages Date: Wed, 15 Jan 2014 01:46:03 -0800 (PST) Message-ID: <1389779163.69122.YahooMailBasic@web125504.mail.ne1.yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE To: netdev@vger.kernel.org, Bill Fink , "David S. Miller" , Alexey Kuznetsov , James Morris , Hideaki Yoshifuji , Patrick McHardy , Hannes Frederic Sowa Return-path: Received: from nm49.bullet.mail.ne1.yahoo.com ([98.138.120.56]:41726 "HELO nm49.bullet.mail.ne1.yahoo.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1750754AbaAOJtG convert rfc822-to-8bit (ORCPT ); Wed, 15 Jan 2014 04:49:06 -0500 Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Jan 14, 2014 at 02:13:44PM +0100, Hannes Frederic Sowa wrote: > > On Mon, Jan 13, 2014 at 06:22:44PM +0100, Francois-Xavier Le Bail w= rote: > > > - Add "anycast_src_icmp_error" sysctl to control the use of anyca= st addresses > > >=A0=A0=A0as source addresses for ICMPv6 error messages. This sysct= l is false by > > >=A0=A0=A0default to preserve existing behavior. > > > - Use it in icmp6_send(). > > >=20 > > > Suggested-by: Bill Fink > > > Signed-off-by: Francois-Xavier Le Bail > >=20 > > Regarding the anycast patches, I contacted someone from IETF. > >=20 > > The number of sysctls needed to get introduced to have all the flex= ibility > > regarding source address selection and don't break backward compati= bility > > concerns me a bit. > >=20 > > Especially on end hosts, where those switches will be important, I = think we > > really have to think about sensible defaults without breaking curre= nt > > software. > >=20 > > I currently consider a per-address flag, if those anycast addresses > > should be available in source address selection (also with an enhan= cement to > > current IPV6_JOIN_ANYCAST logic). >=20 > Francois, we should really think about this. Also if we should just > make the pre-defined subnet address just a normal anycast address in = the > long-term (which just happens to get automatically added to an interf= ace > if forwarding is enabled) and bundle all the source address selection > logic on the per-address state. Hi Hannes, Please submit patches with your solution, so that we can have a basis for discussion. > If that would be the case, we could revert > 509aba3b0d366b7f16a9a2eebac1156b25f5f622 ("IPv6: add the option to us= e > anycast addresses as source addresses in echo reply") and thus would > eliminate one sysctl. If your solution achieve the same goal without this sysctl, I agree wit= h you. > It would be fine if we can make this decision before David merges wit= h > Linus.=A0 I guess we can still do this decision while in -rc phase. B= ut > as soon as the knob is in a released version of linux we can never ta= ke > it back (I really don't like sysctls). Sure. Cheers, =46ran=E7ois-Xavier