From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264668AbUHGXZH (ORCPT ); Sat, 7 Aug 2004 19:25:07 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264085AbUHGXZH (ORCPT ); Sat, 7 Aug 2004 19:25:07 -0400 Received: from the-village.bc.nu ([81.2.110.252]:39364 "EHLO localhost.localdomain") by vger.kernel.org with ESMTP id S264668AbUHGXYZ (ORCPT ); Sat, 7 Aug 2004 19:24:25 -0400 Subject: Re: [linux-usb-devel] USB shared interrupt problem From: Alan Cox To: Pete Zaitcev Cc: Aleksey Gorelov , Linux Kernel Mailing List , linux-usb-devel@lists.sourceforge.net In-Reply-To: <20040806134230.5b05bca0@lembas.zaitcev.lan> References: <5F106036E3D97448B673ED7AA8B2B6B30154AFED@scl-exch2k.phoenix.com> <20040806134230.5b05bca0@lembas.zaitcev.lan> Content-Type: text/plain Content-Transfer-Encoding: 7bit Message-Id: <1091917299.19077.34.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6 (1.4.6-2) Date: Sat, 07 Aug 2004 23:21:41 +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Gwe, 2004-08-06 at 21:42, Pete Zaitcev wrote: > I have a patch which prevents SMM BIOS from doing this to us > by requesting unconditional handoff (it comes from Vojtech @SuSE, > modified by John Stulz from IBM for 2.4). However, if you have a "normal" > BIOS doing this, I'm a little lost as to what to do. It can still honor > the handoff, if you're lucky. In SMM mode the USB controller isn't connected to the PCI IRQ line. That is one thing that is done in the changeover. It also seems to be causing problems with EHCI hand over on NVidia boards still. BTW the patch wants extending by someone to cover EHCI - modern boards are doing EHCI USB legacy now. Alan