From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: max GiB written per boot Date: Wed, 8 Oct 2003 08:44:38 +0200 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <20031008064438.GO1704@suse.de> References: <1064855570.3184.3.camel@patehci2> <1064865746.4262.63.camel@patehci2> <1065460336.10804.1.camel@patehci2> <1065474044.6835.44.camel@patehci2> <1065484077.2899.1.camel@patehci2> <20031007055714.GO972@suse.de> <20031007151225.0a65dcb7.rddunlap@osdl.org> <20031007225708.GK5662@linnie.riede.org> <20031007182714.0eb330c2.rddunlap@osdl.org> <20031007213443.56dd9b58.rddunlap@osdl.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from ns.virtualhost.dk ([195.184.98.160]:14513 "EHLO virtualhost.dk") by vger.kernel.org with ESMTP id S261328AbTJHGov (ORCPT ); Wed, 8 Oct 2003 02:44:51 -0400 Content-Disposition: inline In-Reply-To: <20031007213443.56dd9b58.rddunlap@osdl.org> List-Id: linux-scsi@vger.kernel.org To: "Randy.Dunlap" Cc: wrlk@riede.org, linux-scsi@vger.kernel.org On Tue, Oct 07 2003, Randy.Dunlap wrote: > On Tue, 7 Oct 2003 18:27:14 -0700 "Randy.Dunlap" wrote: > > | On Tue, 7 Oct 2003 18:57:08 -0400 Willem Riede wrote: > | > | | On 2003.10.07 18:12, Randy.Dunlap wrote: > | | > On Tue, 7 Oct 2003 07:57:14 +0200 Jens Axboe wrote: > | | > > | | > | > | | > | Yeah, ide-scsi is broken so you probably shouldn't be using it. > | | > > | | > Is there any sense in trying to repair it? > | | > I could spend some time on it... > | | > | | Yes, there is IMHO. The Onstream tape driver (osst) depends on ide-scsi > | | for controlling the IDE version of the drive (DI-30). > | | > | | Fixing ide-scsi is not trivial, though. I've spent some time on it > | | on and off and have not been all that successful. The biggest problem > | | I have is the error recovery -- making ide subsystem error handling > | | happen when the scsi error handler needs it... > | > | Alan said (a few months back) that he was fixing ide-scsi in 2.4.x-ac > | and that it should be forward-ported to 2.6.x, so that is where I > | would begin... I haven't checked it for progress yet. > > but this hasn't really happened (2.4.x-ac ide-scsi) afaict. 2.4 and 2.6 also uses different error handling for ide-scsi, so it would probably not have helped you a lot. -- Jens Axboe