From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from fed1rmmtao06.cox.net (fed1rmmtao06.cox.net [68.230.241.33]) by ozlabs.org (Postfix) with ESMTP id BE4AB2BC0F for ; Tue, 28 Sep 2004 23:50:29 +1000 (EST) Date: Tue, 28 Sep 2004 06:50:26 -0700 From: Matt Porter To: Marc Leeman Message-ID: <20040928065026.B31883@home.com> References: <1f729c4804092806332aaf052e@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <1f729c4804092806332aaf052e@mail.gmail.com>; from marc.leeman@gmail.com on Tue, Sep 28, 2004 at 03:33:16PM +0200 Cc: Linux PowerPC Embedded Mailinglist Subject: Re: throughput question List-Id: Linux on Embedded PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, Sep 28, 2004 at 03:33:16PM +0200, Marc Leeman wrote: > I see some mails passing over the list about gigabit interfaces, so I > wonder if there are any suggestions as to what processor/interface > should be used for high througput. > > I am currently working on a 8245 based board with 2.4.27 e100 driver > and when taking in (and processing streams), only about 50 Mbps can be > handled: about half of the time is based in kernel space (handling > interrupts). > > The eepro100 driver was much worse. > > As a result, I am a bit wondering about the higher bitrates. How much > do you achieve on tests with what hardware? What benchmark do you use to measure throughput? Since you are processing streams sounds like you have an application-based benchmark rather than one of the typical ones. You can probably show the good throughput if you pick another benchmark, but it probably won't answer your system design question. -Matt