From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH] fusion: streamline ->queuecommand Date: Wed, 6 Oct 2004 17:41:36 +0200 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <20041006154136.GA28333@lst.de> References: <0E3FA95632D6D047BA649F95DAB60E5705262196@exa-atlanta> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from verein.lst.de ([213.95.11.210]:29604 "EHLO mail.lst.de") by vger.kernel.org with ESMTP id S269314AbUJFPmH (ORCPT ); Wed, 6 Oct 2004 11:42:07 -0400 Content-Disposition: inline In-Reply-To: <0E3FA95632D6D047BA649F95DAB60E5705262196@exa-atlanta> List-Id: linux-scsi@vger.kernel.org To: "Moore, Eric Dean" Cc: Christoph Hellwig , Matthew Wilcox , Christoph Hellwig , linux-scsi@vger.kernel.org On Mon, Oct 04, 2004 at 05:33:41PM -0400, Moore, Eric Dean wrote: > (3) How will a generic negotiation be done for raid volumes? The MPT driver > supports RAID levels > 0, 1, and 0+1. The driver does DV via RAID passthru to the individual > devices. I've taken a quick look at fusion in that respect. I think the cleanest thing would be to add "virtual" scsi_device structures for raid physical disks, and set the BLIST_NO_ULD_ATTACH flag so we don't get the disk driver to attach to it. This would have the added benefit of beeing able to use the SG_IO passthrough interface even for physical disk in a raid volume. To make this work we'd obviously have to add some code to ->queuecommand to handle this special case. I think all the other questions have been answered already.