From owner-freebsd-hackers@FreeBSD.ORG Mon Aug 25 13:56:25 2003 Return-Path: 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 89FFF16A4BF; Mon, 25 Aug 2003 13:56:25 -0700 (PDT) Received: from dmlb.org (pc2-cmbg4-6-cust36.cmbg.cable.ntl.com [81.96.76.36]) by mx1.FreeBSD.org (Postfix) with ESMTP id F271743F85; Mon, 25 Aug 2003 13:56:23 -0700 (PDT) (envelope-from dmlb@dmlb.org) Received: from orac.my.domain ([192.168.200.75] helo=orac) by dmlb.org with smtp (Exim 3.36 #1) id 19rONy-000MIA-00; Mon, 25 Aug 2003 21:56:22 +0100 Message-ID: <000501c36b4b$57ba6b20$4bc8a8c0@orac> From: "Duncan Barclay" To: , Date: Mon, 25 Aug 2003 21:56:23 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 Subject: New BETA of Broadcom 440x chipset driver X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 25 Aug 2003 20:56:25 -0000 Hi I think I have fixed the RX packet loss and memory corruption problems with the previous version of the driver. Please try the code at http://people.freebsd.org/~dmlb/bcm-0308252140.tar.gz I have manged to get full link speed ftping large files in both direction. Also, I have successfully populated an empty /usr/src tree with cvsup whilst up and downloading 50MB files. Would people please try this and feed back good and bad experiences. I would be interested if people could run their favorite net bench marks with the hw.bcm_rx_quick sysctl set to 1 (default) and 0. Setting to zero forces the driver to copy data from the NIC, one causes the driver to do a small 44 byte memory shuffle. Whilst wire speed should not differ, time spent in the interrupt routine should be less with the sysctl set to 1. Please do not use on a "valuable" system. The problem that this release tries to fix sometimes resulted in kernel memory corruption. All the best Duncan