From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261780AbTJXKMj (ORCPT ); Fri, 24 Oct 2003 06:12:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261827AbTJXKMh (ORCPT ); Fri, 24 Oct 2003 06:12:37 -0400 Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:29919 "EHLO www.linux.org.uk") by vger.kernel.org with ESMTP id S261780AbTJXKMf (ORCPT ); Fri, 24 Oct 2003 06:12:35 -0400 Message-ID: <3F98FB06.2070805@pobox.com> Date: Fri, 24 Oct 2003 06:12:22 -0400 From: Jeff Garzik User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030703 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Stephen Hemminger CC: linux-kernel@vger.kernel.org Subject: Re: r8169 bug in 2.4.22, too much work at interrupt indefinitely References: <1066952294.616b72c0sandos@home.se> <20031023165342.79e927b5.shemminger@osdl.org> In-Reply-To: <20031023165342.79e927b5.shemminger@osdl.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Stephen Hemminger wrote: > On Fri, 24 Oct 2003 01:38:14 +0200 > "John Bäckstrand" wrote: > > >>The r8169 driver in 2.4.22 with or without debian patches/acpi/usb devices sharing the same interrupt: the driver always ends up locking the machine by having indefinitely much to do in the interrupt handler? >> >>Commented the printk out, still hangs the machine anyway. It happens even without a cable in the card. I have found no references to any bug reports at all with this card, and I cant see any bugfixes being in 2.5/2.6 but not 2.4 either. I can see the card TX/RX:ing a few hundred packets or more (3-60 secs) before hanging. >> >>Any way to debug this, or should I just try 2.6? >> >>--- >>John Bäckstrand > > > 2.6 will have the same problem. I have a version that uses NAPI that shouldn't hang. > It seems to work fine, but didn't want to submit it without more testing. Are you sure you're not thinking about 8139too, not r8169?