From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Hutchings Subject: Re: broken rtl8139 and dropped frames decreasing? Date: Mon, 27 Oct 2008 12:50:59 +0000 Message-ID: <1225111859.3018.4.camel@achroite> References: <49057814.2000403@caiway.nl> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: Jan Evert van Grootheest Return-path: Received: from smarthost01.mail.zen.net.uk ([212.23.3.140]:56925 "EHLO smarthost01.mail.zen.net.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751021AbYJ0MvF (ORCPT ); Mon, 27 Oct 2008 08:51:05 -0400 In-Reply-To: <49057814.2000403@caiway.nl> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2008-10-27 at 09:13 +0100, Jan Evert van Grootheest wrote: > Hi, > > I've got this old computer of a friend (to rescue files and put Ubuntu > on it (-: ). > And there's something strange going on in the network stack somewhere. > > I've booted it using an Ubuntu Hardy live CD (i.e. 2.6.24-something kernel). > There's a rtl8139 card in it that uses the 8139too module. But something > is not right with the hardware. Dmesg is full of reports with TX ring > pointers when I plug the cable and try to ping something. > > The interesting thing is that 'ifconfig eth0' shows dropped RX packets. > But each invocation shows one LESS dropped packet! It's never been above > zero (that I have seen). The counter currently stands at 4294967260 > (i.e. -36). [...] That might indicate that the RxMissed register is not implemented on this version of the chip. That could result in the driver adding a value of all-1s (effectively -1) to the rx_missed_errors statistic. Ben. -- Ben Hutchings, Senior Software Engineer, Solarflare Communications Not speaking for my employer; that's the marketing department's job. They asked us to note that Solarflare product names are trademarked.