From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alistair John Strachan Subject: Re: troubles with r8169 Date: Sun, 12 Aug 2007 21:38:58 +0100 Message-ID: <200708122138.58405.alistair@devzero.co.uk> References: <46BF61AB.5060400@shaw.ca> <46BF6853.4020800@lns.pnpi.spb.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Robert Hancock , netdev@vger.kernel.org, linux-kernel@vger.kernel.org To: Vadim Dyadkin Return-path: In-Reply-To: <46BF6853.4020800@lns.pnpi.spb.ru> Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Sunday 12 August 2007 21:06:43 Vadim Dyadkin wrote: > Robert Hancock =D0=BF=D0=B8=D1=88=D0=B5=D1=82: > > This could well be a problem with the nvidia driver as it shares th= e > > same IRQ. The first step would be to see if the problem still shows= up > > without the nvidia binary module loaded. > > Thank for your answer. This problem never shows up if I use only nvid= ia > driver (the work without network), or if I use only r8169 (without > x.org). If I use both of them I have the hang. Usually the hang appea= rs > if OpenGL is used or network rate is maximal. I tested this regimes m= any > times before. If you haven't already, I'd drop linux-bugs@nvidia.com an email to see = if they=20 have any insight. Not that this can't be a kernel bug, just that it's a= bit=20 difficult for kernel developers to debug when you're using a driver for= which=20 we don't have the sources. As for changing the IRQ assignments, I don't have any immediate suggest= ions. I=20 notice that this laptop has a dual core processor, so I'm guessing disa= bling=20 APIC isn't an option. Have you tried that anyway, just to see if the IR= Q=20 assignment differs? --=20 Cheers, Alistair. 137/1 Warrender Park Road, Edinburgh, UK.