From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH] tcp: fixing TLP's FIN recovery Date: Thu, 12 Jun 2014 07:32:57 -0700 Message-ID: <1402583577.3645.477.camel@edumazet-glaptop2.roam.corp.google.com> References: <539319F6.2090907@cogentembedded.com> <1402151680-11434-1-git-send-email-per.hurtig@kau.se> <1402196305.3645.318.camel@edumazet-glaptop2.roam.corp.google.com> <539413BA.7060903@kau.se> <5399B762.5090002@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Nandita Dukkipati , Per Hurtig , Netdev , Anna =?ISO-8859-1?Q?Brunstr=F6m?= , mohammad.rajiullah@kau.se, Neal Cardwell , sergei.shtylyov@cogentembedded.com To: Weiping Pan Return-path: Received: from mail-pd0-f172.google.com ([209.85.192.172]:45553 "EHLO mail-pd0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933291AbaFLOc7 (ORCPT ); Thu, 12 Jun 2014 10:32:59 -0400 Received: by mail-pd0-f172.google.com with SMTP id fp1so1033072pdb.3 for ; Thu, 12 Jun 2014 07:32:59 -0700 (PDT) In-Reply-To: <5399B762.5090002@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 2014-06-12 at 22:21 +0800, Weiping Pan wrote: > When we queue an out of order pure FIN packet, we do not check whether > it has data or not. > tcp_rcv_established > -->tcp_data_queue > ---->tcp_data_queue_ofo > > Then the pure FIN packet can generate SACK, which will trigger fast > recovery or early retransmit on the sender. > > > > If you have verified that a pure FIN does indeed trigger recovery, can > > you tell me what part of the code makes that happen? > Here is the patch I use, I think the original if statement is useless, > so I delete it. Yes, this is exactly what we agreed, and what we tested as well here at Google. Per, can you submit an updated official version of the patch ? Thanks