From owner-freebsd-hackers@FreeBSD.ORG Fri Nov 3 06:55:39 2006 Return-Path: X-Original-To: freebsd-hackers@freebsd.org Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 48A9116A40F; Fri, 3 Nov 2006 06:55:39 +0000 (UTC) (envelope-from joerg@britannica.bec.de) Received: from antivirus.uni-rostock.de (mailrelay1.uni-rostock.de [139.30.8.201]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7498743D58; Fri, 3 Nov 2006 06:55:37 +0000 (GMT) (envelope-from joerg@britannica.bec.de) Received: from antivirus.exch.rz.uni-rostock.de ([127.0.0.1]) by antivirus.uni-rostock.de with Microsoft SMTPSVC(6.0.3790.1830); Fri, 3 Nov 2006 07:55:36 +0100 Received: from antivirus.uni-rostock.de (unverified) by antivirus.exch.rz.uni-rostock.de (Clearswift SMTPRS 5.2.5) with ESMTP id ; Fri, 3 Nov 2006 07:55:36 +0100 Received: from mail pickup service by antivirus.uni-rostock.de with Microsoft SMTPSVC; Fri, 3 Nov 2006 07:55:36 +0100 X-SCL: 1 42.44% Received: from mail.uni-rostock.de ([139.30.8.11]) by antivirus.uni-rostock.de with Microsoft SMTPSVC(6.0.3790.1830); Fri, 3 Nov 2006 07:55:28 +0100 Received: from conversion-daemon.mail2.uni-rostock.de by mail2.uni-rostock.de (iPlanet Messaging Server 5.2 HotFix 2.09 (built Nov 18 2005)) id <0J85003016XJFN@mail.uni-rostock.de> (original mail from joerg@britannica.bec.de); Fri, 03 Nov 2006 07:55:27 +0100 (MET) Received: from britannica.bec.de (wlan033094.uni-rostock.de [139.30.33.94]) by mail2.uni-rostock.de (iPlanet Messaging Server 5.2 HotFix 2.09 (built Nov 18 2005)) with ESMTP id <0J85005F876MFM@mail.uni-rostock.de>; Fri, 03 Nov 2006 07:54:22 +0100 (MET) Received: by britannica.bec.de (Postfix, from userid 1000) id B418E571E; Fri, 03 Nov 2006 07:54:21 +0100 (CET) Date: Fri, 03 Nov 2006 07:54:21 +0100 From: Joerg Sonnenberger In-reply-to: <9ab217670611021511l3120d58bhd0b61bf44f8ecc87@mail.gmail.com> To: freebsd-hackers@freebsd.org, hackers@freebsd.org Mail-followup-to: freebsd-hackers@freebsd.org, hackers@freebsd.org Message-id: <20061103065421.GA829@britannica.bec.de> MIME-version: 1.0 Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7BIT Content-disposition: inline User-Agent: Mutt/1.5.13 (2006-08-11) References: <9ab217670611021511l3120d58bhd0b61bf44f8ecc87@mail.gmail.com> X-OriginalArrivalTime: 03 Nov 2006 06:55:28.0146 (UTC) FILETIME=[0BA5F720:01C6FF15] Cc: Subject: Re: vr(4) performance X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 03 Nov 2006 06:55:39 -0000 On Thu, Nov 02, 2006 at 06:11:00PM -0500, Devon H. O'Dell wrote: > So, vr(4) kind of sucks, and it seems like this is mostly due to the > fact that we call m_defrag() on every mbuf that we send through it. > This seems to really screw performance on outgoing packets (something > like 33% the output efficiency of fxp(4), if I'm understanding this > all correctly). What hardware are you running at that it is a problem? Seriously -- copying 100Mbit/s in memory shouldn't fully busy any post-2k non-embedded CPU. You could just allocate a static output queue and copy into that, dropping at least the allocations of the mbuf clusters, but I don't think it should b worth it. And yes, if you can sustain Fast Ethernet speed with scp to localhost, m_defrag is absolutely not the main problem. Joerg