From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH] tcp: ECN blackhole should not force quickack mode Date: Fri, 23 Sep 2011 13:47:34 -0400 (EDT) Message-ID: <20110923.134734.2267391217120737650.davem@davemloft.net> References: <1316757739.2560.12.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org, hkchu@google.com, ilpo.jarvinen@helsinki.fi, jhs@mojatatu.com, jg@freedesktop.org, dave.taht@gmail.com To: eric.dumazet@gmail.com Return-path: Received: from shards.monkeyblade.net ([198.137.202.13]:58816 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750789Ab1IWRr4 convert rfc822-to-8bit (ORCPT ); Fri, 23 Sep 2011 13:47:56 -0400 In-Reply-To: <1316757739.2560.12.camel@edumazet-laptop> Sender: netdev-owner@vger.kernel.org List-ID: =46rom: Eric Dumazet Date: Fri, 23 Sep 2011 08:02:19 +0200 > While playing with a new ADSL box at home, I discovered that ECN > blackhole can trigger suboptimal quickack mode on linux : We send one > ACK for each incoming data frame, without any delay and eventual > piggyback. >=20 > This is because TCP_ECN_check_ce() considers that if no ECT is seen o= n a > segment, this is because this segment was a retransmit. >=20 > Refine this heuristic and apply it only if we seen ECT in a previous > segment, to detect ECN blackhole at IP level. >=20 > Signed-off-by: Eric Dumazet > CC: Jamal Hadi Salim > CC: Jerry Chu > CC: Ilpo J=E4rvinen > CC: Jim Gettys > CC: Dave Taht > --- > Another possibility is to remove this (not in RFC 3168) heuristic, wh= at > do you think ? Jamal, please give this a look over. Thanks!