From mboxrd@z Thu Jan 1 00:00:00 1970 From: Steffen Klassert Subject: Re: [PATCH] route:ip_rt_frag_needed always return unzero Date: Wed, 19 Oct 2011 10:38:32 +0200 Message-ID: <20111019083832.GT1830@secunet.com> References: <1318921469-25599-1-git-send-email-gaofeng@cn.fujitsu.com> <1318929797.2657.21.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC> <4E9E2929.7070701@cn.fujitsu.com> <1318996187.19139.8.camel@edumazet-laptop> <4E9E5E1C.5020603@cn.fujitsu.com> <20111019072628.GS1830@secunet.com> <4E9E8552.1050405@cn.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Eric Dumazet , davem@davemloft.net, kuznet@ms2.inr.ac.ru, jmorris@namei.org, netdev@vger.kernel.org To: Gao feng Return-path: Received: from a.mx.secunet.com ([195.81.216.161]:57495 "EHLO a.mx.secunet.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752427Ab1JSIim (ORCPT ); Wed, 19 Oct 2011 04:38:42 -0400 Content-Disposition: inline In-Reply-To: <4E9E8552.1050405@cn.fujitsu.com> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, Oct 19, 2011 at 04:07:46PM +0800, Gao feng wrote: > 2011.10.19 15:26, Steffen Klassert wrote: > > > > > It is valid in the sense that we should not provide the user > > with a mtu information if we know that the value we got from > > the icmp packet ist bogus. But perhaps we can think about > > making the check for a valid mtu unconditionally and let > > ip_rt_frag_needed return a valid mtu in any case. > > > > > > I think we should return the pmtu in icmp packet to the raw socket, > and the valid mtu to tcp_v4_err or something else. > Why you want to handle raw sockets different here?