From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [RFC] napi: limit GRO latency Date: Mon, 08 Oct 2012 21:46:32 +0200 Message-ID: <1349725592.21172.3685.camel@edumazet-glaptop> References: <1349718955.21172.3534.camel@edumazet-glaptop> <1349719006.21172.3537.camel@edumazet-glaptop> <20121008115835.3a3bfed6@nehalam.linuxnetplumber.net> <20121008.151035.1665672334154033755.davem@davemloft.net> <20121008121248.783e6dc4@nehalam.linuxnetplumber.net> <1349724612.21172.3643.camel@edumazet-glaptop> <20121008124026.4057aba8@nehalam.linuxnetplumber.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: David Miller , rick.jones2@hp.com, herbert@gondor.apana.org.au, netdev@vger.kernel.org, jesse@nicira.com To: Stephen Hemminger Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:38788 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751072Ab2JHTqi (ORCPT ); Mon, 8 Oct 2012 15:46:38 -0400 Received: by mail-bk0-f46.google.com with SMTP id jk13so2249644bkc.19 for ; Mon, 08 Oct 2012 12:46:36 -0700 (PDT) In-Reply-To: <20121008124026.4057aba8@nehalam.linuxnetplumber.net> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2012-10-08 at 12:40 -0700, Stephen Hemminger wrote: > On Mon, 08 Oct 2012 21:30:12 +0200 > Eric Dumazet wrote: > > > System is under stress, you dont want to increase load at this point, > > but increase bandwidth. > > > > Really try this on a 40Gbe link and see how bad it is. > > Not everybody has 40Gbe hardware just lying around available for testing :-) ;) In fact, the current strategy to flush napi_gro when napi completes might be not good for moderate traffic (thats 99% of the linux machines I believe). We could have a deferred flush instead... I dont know what strategy is used by hardware. But surely hardware uses timers.