From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932190AbVLUVRe (ORCPT ); Wed, 21 Dec 2005 16:17:34 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750996AbVLUVRe (ORCPT ); Wed, 21 Dec 2005 16:17:34 -0500 Received: from cust8446.nsw01.dataco.com.au ([203.171.93.254]:30114 "EHLO cunningham.myip.net.au") by vger.kernel.org with ESMTP id S1751093AbVLUVRd (ORCPT ); Wed, 21 Dec 2005 16:17:33 -0500 Subject: 2.6.15-rc5 and later: USB mouse IRQ post kills the computer post resume. From: Nigel Cunningham Reply-To: ncunningham@cyclades.com To: vojtech@ucw.cz, Greg KH Cc: Linux Kernel Mailing List Content-Type: text/plain Organization: Cyclades Message-Id: <1135199640.9616.21.camel@localhost> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6-1mdk Date: Thu, 22 Dec 2005 07:14:01 +1000 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi Vojtech. I have a HT box with USB mouse support built as modules. Beginning with 2.6.15-rc5 (maybe slightly earlier) a suspend/resume cycle makes the USB mouse get in an invalid state, such that I get a gazillion messages in the logs saying "unexpected IRQ trap at vector 99", or in some alternately a hard hang. No work around found yet. Are you the right man to talk to, or is Greg? (Spose I should cc him, so I'll add that now). I can use kdb if it's helpful. Would you like my kconfig? I should probably mention that I've also gained problems with DRM (it's Radeon based), but I don't think they're related. I'll email BenH on that separately. Regards, Nigel