From mboxrd@z Thu Jan 1 00:00:00 1970 From: Boaz Harrosh Subject: Re: [PATCH] gdth: bugfix for the at-exit problems Date: Thu, 14 Feb 2008 18:18:43 +0200 Message-ID: <47B469E3.4080403@panasas.com> References: <47A19E26.30107@panasas.com> <47B1D7A8.8010108@panasas.com> <47B1DA2A.1060904@panasas.com> <1202917468.3109.5.camel@localhost.localdomain> <47B312B3.3010200@panasas.com> <47B31BDE.2030408@panasas.com> <1202921122.3109.31.camel@localhost.localdomain> <47B31FC2.4040206@panasas.com> <1202922226.3109.36.camel@localhost.localdomain> <47B324E7.9040502@panasas.com> <1202924198.3109.52.camel@localhost.localdomain> <47B41CBD.8040701@panasas.com> <47B42CF4.2050008@panasas.com> <1203005443.3189.23.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from gw-colo-pa.panasas.com ([66.238.117.130]:6240 "EHLO cassoulet.panasas.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753520AbYBNQT5 (ORCPT ); Thu, 14 Feb 2008 11:19:57 -0500 In-Reply-To: <1203005443.3189.23.camel@localhost.localdomain> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: James Bottomley Cc: =?UTF-8?B?U3ZlbiBLw7ZobGVy?= , Christoph Hellwig , Jeff Garzik , linux-scsi , linux-kernel@vger.kernel.org, Joerg Dorchain , Jon Chelton , Stefan Priebe - allied internet ag On Thu, Feb 14 2008 at 18:10 +0200, James Bottomley wrote: > On Thu, 2008-02-14 at 13:58 +0200, Boaz Harrosh wrote: >> This is a bugfix for the 2.6.24.x stable releases. >> >> gdth_exit would first remove all cards then stop the timer >> and would not sync with the timer function. This caused a crash >> in gdth_timer() when module was unloaded. >> So del_timer_sync the timer before we delete the cards. >> >> also the reboot notifier function would crash. So unify >> the exit and halt functions with a gdth_shutdown() that's >> called by both. > > The patch looks fine now, thanks. Can we actually get a tester just to > make sure there's nothing I missed. > > James > > Yes, and the tester reported, a breakage. We are on it. Apparently, you cannot do a full deallocation of resources at reboot notifier, nor would you want to I guess. But you can do the flush. The exit call is never called on a reboot and the card access is valid to the end. Please comment? So I pretty much reverted that patch, but did leave some cleanups. Also we found the other problems reported with user-mode tools and cat /proc/sys/gdth/0 so 2 patches on the way above reverted. Give us a few ours to test every thing. Thanks Boaz