From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexey Kuznetsov Subject: Re: [PATCH 2/2] ipv4: make do_ip_setsockopt for IP_MULTICAST_IF support ip_mreq struct Date: Thu, 17 Sep 2009 13:36:22 +0400 Message-ID: <20090917093622.GA19142@ms2.inr.ac.ru> References: <1253164811-15820-1-git-send-email-dfeng@redhat.com> <4AB1FE6E.9000201@cn.fujitsu.com> <4AB1FF14.6000801@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Shan Wei , davem@davemloft.net, kaber@trash.net, yoshfuji@linux-ipv6.org, jmorris@namei.org, pekkas@netcore.fi, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Marc Milgram To: Danny Feng Return-path: Received: from minus.inr.ac.ru ([194.67.69.97]:54194 "HELO ms2.inr.ac.ru" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with SMTP id S1751778AbZIQK3Z (ORCPT ); Thu, 17 Sep 2009 06:29:25 -0400 Content-Disposition: inline In-Reply-To: <4AB1FF14.6000801@redhat.com> Sender: netdev-owner@vger.kernel.org List-ID: Hello! On Thu, Sep 17, 2009 at 05:19:16PM +0800, Danny Feng wrote: > In fact, current implemetation supports: > > struct ip_mreqn mreqn; > setsockopt(socket, IPPROTO_IP, IP_MULTICAST_IF, &mreqn, sizeof(mreqn)); > > Then why not support mreq? Because support of ip_mreqn makes sense (it user interface index) and support of ip_mreq, which does not contain interface index, does not. "Polymorphic" IP_MULTICAST_IF was a stupid mistake (mine), which should not grow like a tumor. Ack to patch #1, nack to #2. Alexey