From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bijay Singh Subject: Re: TCP-MD5 checksum failure on x86_64 SMP Date: Mon, 10 May 2010 17:27:28 +0000 Message-ID: <884B3D18-4181-4D98-A8FD-8660CE658C0B@guavus.com> References: <1273085598.2367.233.camel@edumazet-laptop> <1273147586.2357.63.camel@edumazet-laptop> <20100506.220443.135536330.davem@davemloft.net> <1273210329.2222.42.camel@edumazet-laptop> <20100507101451.1b4286b7@nehalam> <1273252893.2261.84.camel@edumazet-laptop> <20100507103639.4f1a51fa@nehalam> <1273268446.2325.53.camel@edumazet-laptop> <1273504693.2221.17.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Stephen Hemminger , David Miller , "" , "" , "" To: Eric Dumazet Return-path: Received: from [204.232.241.167] ([204.232.241.167]:56916 "EHLO mx1.guavus.com" rhost-flags-FAIL-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1753778Ab0EJR1c convert rfc822-to-8bit (ORCPT ); Mon, 10 May 2010 13:27:32 -0400 In-Reply-To: <1273504693.2221.17.camel@edumazet-laptop> Content-Language: en-US Content-ID: <51878d71-deeb-4cea-aa77-afaa313f34c8> Sender: netdev-owner@vger.kernel.org List-ID: Hi Eric, Didn't intend to mix the issues. It was a hack intended to calm the ne= rves. I am going to apply the proper patches asap. About the latest problem of MD5 not working with MTU set to 4470. I not= iced this when i needed to change the MTU for some other purpose.=20 Since it was a production box, i have to first set up my box with the r= ight NIC card to reproduce this and try debugging it. In the meantime a= ny ques will help. Thanks, Bijay =20 On 10-May-2010, at 8:48 PM, Eric Dumazet wrote: > Le lundi 10 mai 2010 =E0 14:55 +0000, Bijay Singh a =E9crit : >> Hi, >> I had noticed the corruption in the context and actually did what is= mentioned. >>=20 >> I allocated the context on the stack and plugged in the md5.c functi= ons. I was able to temporarily solve the problem, all this before I got= a response on this thread. >>=20 >> But now I have seeing another problem, when i change the MTU on the = interface from 1500 to 4470 none of the message from the peer get thru = and I get hash failed message. I am wondering if this is another bug ge= tting hit in this scenario. >=20 > Thats very fine, but you mix very different problems. >=20 > Step by step resolution is required, and clean patches too, because > plugging md5.c functions is not an option for stable series :) >=20 > Obviously, nobody seriously used TCP-MD5 on linux, but you... >=20 >=20 >=20