From mboxrd@z Thu Jan 1 00:00:00 1970 From: Brian King Subject: Re: [PATCH] IBM Power RAID driver - third version Date: Fri, 27 Feb 2004 13:39:30 -0600 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <403F9CF2.7000003@us.ibm.com> References: <403D0DBE.4020207@us.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from e3.ny.us.ibm.com ([32.97.182.103]:64708 "EHLO e3.ny.us.ibm.com") by vger.kernel.org with ESMTP id S263002AbUB0Tjd (ORCPT ); Fri, 27 Feb 2004 14:39:33 -0500 List-Id: linux-scsi@vger.kernel.org To: Brian King Cc: James.Bottomley@steeleye.com, linux-scsi@vger.kernel.org There was a bug in this version causing system hangs on certain machines as I was doing memory space accesses with memory space disabled, please use http://www-124.ibm.com/storageio/ipr/patch-2.6.3-ipr-2.0.0-7.gz instead. Also, you will need: http://www-124.ibm.com/storageio/ipr/patch-2.6.3-ipr-pci-ids as well. Thanks -Brian Brian King wrote: > I have uploaded the third version of the ipr device driver to: > > http://www-124.ibm.com/storageio/ipr/patch-2.6.3-ipr-2.0.0-6.gz > > Major changes: > > 1. Removed use of bit fields. > 2. Removed show_queue_depth since we can use default one. > 3. Added FIXME comments to issues still outstanding that > require midlayer changes > 4. Several bug fixes. > > Outstanding issues: > > 1. Driver needs the start_unit patch I submitted in order to function > properly. This has still not been accepted. A 2.6.3 patch can be > found at: http://www-124.ibm.com/storageio/ipr/patch-2.6.3-start_unit.gz > 2. Driver needs the array_controller patch, which still has not been > accepted. > http://www-124.ibm.com/storageio/ipr/patch-2.6.3-array_controller.gz > 3. Still checking allow_cmds flag in queuecommand. ERP ops (e.g. TUR), > still do not check host_blocked. Patch to force retries to obey > host_blocked flag has been accepted, will be submitting another > patch to fix ERP ops. Once this is accepted, I will remove the check. > 4. ioa_is_dead flag still exists and is checked in queuecommand. Need to > add a scsi_set_host_offline interface to change this. > 5. res->del_from_ml flag still exists. To remove it requires a > scsi_set_device_offline interface. > 6. Driver is still modifying the command timeout. Once an acceptable > alternative is agreed upon, this can be cleaned up. > > Please apply. I will continue working the outstanding issues and > update the ipr driver along with the rest of the patch. > > -- Brian King eServer Storage I/O IBM Linux Technology Center