From owner-freebsd-bugs Mon May 14 8:12:29 2001 Delivered-To: freebsd-bugs@freebsd.org Received: from host213-123-129-154.btopenworld.com (host213-123-129-154.btopenworld.com [213.123.129.154]) by hub.freebsd.org (Postfix) with ESMTP id 340EB37B42C for ; Mon, 14 May 2001 08:12:25 -0700 (PDT) (envelope-from dominic@host213-123-129-154.btopenworld.com) Received: (from dominic@localhost) by host213-123-129-154.btopenworld.com (8.11.3/8.11.3) id f4EFCOx00955; Mon, 14 May 2001 16:12:24 +0100 (BST) (envelope-from dominic) Date: Mon, 14 May 2001 16:12:24 +0100 From: Dominic Marks To: freebsd-bugs@freebsd.org Cc: wpaul@ctr.columbia.edu Subject: vr0 network card failure Message-ID: <20010514161224.A944@host213-123-129-154.btopenworld> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hello, > # FreeBSD Version Information FreeBSD apollo 4.3-STABLE Mon May 14 08:37:38 BST 2001 < > # dmesg | grep vr Preloaded elf module "if_vr.ko" at 0xc02aa278. vr0: port 0x1000-0x10ff mem +0xf4000000-0xf40000ff irq 9 at device 16.0 on pci0 vr0: Ethernet address: 00:50:ba:f1:b0:44 miibus0: on vr0 vr0: watchdog timeout < I'm having problems using a D-Link network card, which is supported in the kernel as a VIA Rhine chipset (vr). I've searched the mailing list archives and it appears that I'm not the first to experience this flaw either. The problem occurs when I do intensive data traffic on the interface and this causes two effects. Firstly there are large durations when there is no traffic being sent, which is normally resumed after 20 seconds. Then there is second, which is the core problem. The card often dies completely setting the OACTIVE flag and becoming unresponsive. This is only fixed by downing and restarting the interface, which is very annoying as it normally only manages 20MB of data to be transferred. On low data load ( < 200KBps) the cards work fine. In my setup there are two FreeBSD boxes each with a vr card. One is a standard Intel BX440 based motherboard, the other an Abit KT7A. The problem seems to occur on both machines, although more frequently on the BX440 board. Having read various PR's on this problem there seems to be no good solution. It is reccomended that I move the card back to its pre-MII Bus state in the kernel, however is this really a good fix? If it is why isn't it either the default or mentioned in the 'man vr' page? Has anyone got a preferrable solution to this problem or am I going to have to buy a new set of network cards? Many Thanks -- Dominic Marks To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message