From mboxrd@z Thu Jan 1 00:00:00 1970 From: Herbert Xu Subject: Re: [0/14] GRO: Lots of microoptimisations Date: Fri, 29 May 2009 19:28:21 +1000 Message-ID: <20090529092821.GA11725@gondor.apana.org.au> References: <20090527044539.GA32372@gondor.apana.org.au> <20090527175223.GB7804@neterion.com> <20090527230858.GA24278@gondor.apana.org.au> <20090528152143.GA4501@neterion.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "David S. Miller" , netdev@vger.kernel.org To: Benjamin LaHaise Return-path: Received: from rhun.apana.org.au ([64.62.148.172]:58923 "EHLO arnor.apana.org.au" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751827AbZE2J2Z (ORCPT ); Fri, 29 May 2009 05:28:25 -0400 Content-Disposition: inline In-Reply-To: <20090528152143.GA4501@neterion.com> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, May 28, 2009 at 11:21:43AM -0400, Benjamin LaHaise wrote: > > With GRO off I'm getting ~4.7-5Gbps to the receiver which is CPU bound with > netperf. With GRO on, that drops to ~3.9-4.3Gbps. The only real difference > is the entry point into the net code being napi_gro_receive() vs > netif_receive_skb(). That doesn't sound right at all. Can you run tcpdump on it to see if it's actually aggregating? Thanks, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt