From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neal Cardwell Subject: Re: [PATCH v2 net-next 3/4] tcp: make tcp_space() aware of socket backlog Date: Tue, 27 Nov 2018 16:22:55 -0500 Message-ID: References: <20181127155738.73068-1-edumazet@google.com> <20181127155738.73068-4-edumazet@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Cc: David Miller , Netdev , jean-louis@dupond.be, Yuchung Cheng , Eric Dumazet To: Eric Dumazet Return-path: Received: from mail-oi1-f194.google.com ([209.85.167.194]:46005 "EHLO mail-oi1-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726273AbeK1IWZ (ORCPT ); Wed, 28 Nov 2018 03:22:25 -0500 Received: by mail-oi1-f194.google.com with SMTP id b141so20643093oii.12 for ; Tue, 27 Nov 2018 13:23:12 -0800 (PST) In-Reply-To: <20181127155738.73068-4-edumazet@google.com> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Nov 27, 2018 at 10:57 AM Eric Dumazet wrote: > > Jean-Louis Dupond reported poor iscsi TCP receive performance > that we tracked to backlog drops. > > Apparently we fail to send window updates reflecting the > fact that we are under stress. > > Note that we might lack a proper window increase when > backlog is fully processed, since __release_sock() clears > sk->sk_backlog.len _after_ all skbs have been processed. > > This should not matter in practice. If we had a significant > load through socket backlog, we are in a dangerous > situation. > > Reported-by: Jean-Louis Dupond > Signed-off-by: Eric Dumazet > --- Acked-by: Neal Cardwell Nice. Thanks! neal