From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nicolas Dichtel Date: Mon, 10 Sep 2012 12:47:54 +0000 Subject: Re: [PATCH 1/2] dst: take into account policy update on check() Message-Id: <504DE17A.1000908@6wind.com> List-Id: References: <504A08EA.3080506@6wind.com> <1347030598.2484.777.camel@edumazet-glaptop> <504A0F1F.5060504@6wind.com> <20120907.144828.97793990734588625.davem@davemloft.net> In-Reply-To: <20120907.144828.97793990734588625.davem@davemloft.net> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: David Miller Cc: eric.dumazet@gmail.com, vyasevich@gmail.com, sri@us.ibm.com, linux-sctp@vger.kernel.org, netdev@vger.kernel.org Le 07/09/2012 20:48, David Miller a =E9crit : > From: Nicolas Dichtel > Date: Fri, 07 Sep 2012 17:13:35 +0200 > >> Le 07/09/2012 17:09, Eric Dumazet a =E9crit : >>> On Fri, 2012-09-07 at 16:47 +0200, Nicolas Dichtel wrote: >>>> Le 07/09/2012 16:35, Eric Dumazet a =E9crit : >>>>> >>>>> Hmm... cant we reuse rt_genid ? >>>>> >>>>> (When changing flow_cache_genid, change &net->ipv4.rt_genid) >>>> >>>> And so adding a new field in net->ipv6? >>> >>> or move net->ipv4.rt_genid to net->rt_genid >>> >>> Having separate field for IPv4/IPv6 is of little interest IMHO >>> >> Ok, I will wait feedback from other people and repost a patch after. > > A global net->rt_genid is definitely the way to go. > So it means that IPv6 dst entries will be invalidated by IPv4 route managem= ent.=20 For example, calling rt_cache_flush() will flush IPv6 dst too. Is this acce= ptable? I will send a new version. From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nicolas Dichtel Subject: Re: [PATCH 1/2] dst: take into account policy update on check() Date: Mon, 10 Sep 2012 14:47:54 +0200 Message-ID: <504DE17A.1000908@6wind.com> References: <504A08EA.3080506@6wind.com> <1347030598.2484.777.camel@edumazet-glaptop> <504A0F1F.5060504@6wind.com> <20120907.144828.97793990734588625.davem@davemloft.net> Reply-To: nicolas.dichtel@6wind.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: eric.dumazet@gmail.com, vyasevich@gmail.com, sri@us.ibm.com, linux-sctp@vger.kernel.org, netdev@vger.kernel.org To: David Miller Return-path: Received: from mail-ee0-f46.google.com ([74.125.83.46]:57844 "EHLO mail-ee0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757838Ab2IJMr6 (ORCPT ); Mon, 10 Sep 2012 08:47:58 -0400 Received: by eekc1 with SMTP id c1so1097309eek.19 for ; Mon, 10 Sep 2012 05:47:57 -0700 (PDT) In-Reply-To: <20120907.144828.97793990734588625.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: Le 07/09/2012 20:48, David Miller a =E9crit : > From: Nicolas Dichtel > Date: Fri, 07 Sep 2012 17:13:35 +0200 > >> Le 07/09/2012 17:09, Eric Dumazet a =E9crit : >>> On Fri, 2012-09-07 at 16:47 +0200, Nicolas Dichtel wrote: >>>> Le 07/09/2012 16:35, Eric Dumazet a =E9crit : >>>>> >>>>> Hmm... cant we reuse rt_genid ? >>>>> >>>>> (When changing flow_cache_genid, change &net->ipv4.rt_genid) >>>> >>>> And so adding a new field in net->ipv6? >>> >>> or move net->ipv4.rt_genid to net->rt_genid >>> >>> Having separate field for IPv4/IPv6 is of little interest IMHO >>> >> Ok, I will wait feedback from other people and repost a patch after. > > A global net->rt_genid is definitely the way to go. > So it means that IPv6 dst entries will be invalidated by IPv4 route man= agement.=20 =46or example, calling rt_cache_flush() will flush IPv6 dst too. Is thi= s acceptable? I will send a new version.