From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH net] tcp: restore autocorking Date: Thu, 03 May 2018 13:55:46 +0000 Message-ID: References: <20180503032513.210324-1-edumazet@google.com> <298c3b30-d3d3-6069-dd46-bcf0ce315126@mellanox.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Cc: David Miller , Tal Gilboa , netdev , Michael Wenig , Eric Dumazet To: Tariq Toukan Return-path: Received: from mail-wr0-f179.google.com ([209.85.128.179]:42317 "EHLO mail-wr0-f179.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751064AbeECNz7 (ORCPT ); Thu, 3 May 2018 09:55:59 -0400 Received: by mail-wr0-f179.google.com with SMTP id v5-v6so17722420wrf.9 for ; Thu, 03 May 2018 06:55:58 -0700 (PDT) In-Reply-To: <298c3b30-d3d3-6069-dd46-bcf0ce315126@mellanox.com> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, May 3, 2018 at 4:06 AM Tariq Toukan wrote: > On 03/05/2018 6:25 AM, Eric Dumazet wrote: > > When adding rb-tree for TCP retransmit queue, we inadvertently broke > > TCP autocorking. > > > > tcp_should_autocork() should really check if the rtx queue is not empty= . > > > Hi Eric, > We are glad to see that the issue that Tal investigated and reported [1] > is now addressed. > Thanks for doing that! > Tal, let=E2=80=99s perf test to see the effect of this fix. > Best, > Tariq > [1] https://patchwork.ozlabs.org/cover/822218/ Yes, but you never really gave any insights of what exact tests you were running :/ Sometimes the crystal ball is silent, sometimes it gives a hint :)