From mboxrd@z Thu Jan 1 00:00:00 1970 From: Rick Jones Subject: Re: e1000 full-duplex TCP performance well below wire speed Date: Thu, 31 Jan 2008 10:38:32 -0800 Message-ID: <47A215A8.2090104@hp.com> References: <36D9DB17C6DE9E40B059440DB8D95F52044F81DF@orsmsx418.amr.corp.intel.com> <36D9DB17C6DE9E40B059440DB8D95F52044F8BA3@orsmsx418.amr.corp.intel.com> <47A20E9E.7070503@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Cc: Bruce Allen , "Brandeburg, Jesse" , netdev@vger.kernel.org, Carsten Aulbert , Henning Fehrmann , Bruce Allen To: "Kok, Auke" Return-path: Received: from g1t0026.austin.hp.com ([15.216.28.33]:34214 "EHLO g1t0026.austin.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756658AbYAaSif (ORCPT ); Thu, 31 Jan 2008 13:38:35 -0500 In-Reply-To: <47A20E9E.7070503@intel.com> Sender: netdev-owner@vger.kernel.org List-ID: > A lot of people tend to forget that the pci-express bus has enough bandwidth on > first glance - 2.5gbit/sec for 1gbit of traffix, but apart from data going over it > there is significant overhead going on: each packet requires transmit, cleanup and > buffer transactions, and there are many irq register clears per second (slow > ioread/writes). The transactions double for TCP ack processing, and this all > accumulates and starts to introduce latency, higher cpu utilization etc... Sounds like tools to show PCI* bus utilization would be helpful... rick jones