From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265124AbUGGNFo (ORCPT ); Wed, 7 Jul 2004 09:05:44 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265109AbUGGM7h (ORCPT ); Wed, 7 Jul 2004 08:59:37 -0400 Received: from holomorphy.com ([207.189.100.168]:28125 "EHLO holomorphy.com") by vger.kernel.org with ESMTP id S265124AbUGGM4J (ORCPT ); Wed, 7 Jul 2004 08:56:09 -0400 Date: Wed, 7 Jul 2004 05:55:59 -0700 From: William Lee Irwin III To: Dmitry Torokhov Cc: linux-kernel@vger.kernel.org, Andrew Morton Subject: Re: 2.6.7-mm6 Message-ID: <20040707125559.GF21066@holomorphy.com> Mail-Followup-To: William Lee Irwin III , Dmitry Torokhov , linux-kernel@vger.kernel.org, Andrew Morton References: <20040705023120.34f7772b.akpm@osdl.org> <200407070015.39507.dtor_core@ameritech.net> <20040707063733.GD21066@holomorphy.com> <200407070747.16512.dtor_core@ameritech.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <200407070747.16512.dtor_core@ameritech.net> User-Agent: Mutt/1.5.5.1+cvs20040105i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wednesday 07 July 2004 01:37 am, William Lee Irwin III wrote: >> This suspicion is correct. It boots normally with the patch you posted >> to do that registration outside the interrupts-off critical section >> applied. Bootlog below. On Wed, Jul 07, 2004 at 07:47:16AM -0500, Dmitry Torokhov wrote: > Great! I am still somewhat confused why it started locking up with sysfs > patch - even before sunzilog was calling serio_register_port with interrupts > off and serio core was downing it's serio_sem as the very first thing. Since > at the time sunzilog registers its ports no serio drivers have been registered > yet, effectively the only change introduced by sysfs patch is the call to > device_register which takes bus' subsystem rwsem and there really should not > be any congestion. > Maybe rwsems can not be touched with interrupts off? Sparc only? Everywhere? > (I know that you should not normally call functions that may sleep with > interrupts off). CONFIG_PREEMPT enables this to be warned on appropriately. It should basically never happen unless it's a down_trylock() etc. -- wli