From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [PATCH] tmscsim: INQUIRY always only untagged Date: 05 Sep 2004 18:22:27 -0400 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1094422949.10977.15.camel@mulgrave> References: <20040904140102.A15247@infradead.org> <20040904233835.A18659@infradead.org> <1094398979.10977.9.camel@mulgrave> <20040905173134.A31521@infradead.org> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from stat16.steeleye.com ([209.192.50.48]:48358 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S267298AbUIEWWe (ORCPT ); Sun, 5 Sep 2004 18:22:34 -0400 In-Reply-To: <20040905173134.A31521@infradead.org> List-Id: linux-scsi@vger.kernel.org To: Christoph Hellwig Cc: Guennadi Liakhovetski , SCSI Mailing List On Sun, 2004-09-05 at 12:31, Christoph Hellwig wrote: > Umm yes. But lets get the driver no oops first and then move forward. > There's still a lot of work to do on the driver. Well, OK. I agree to putting a palliative fix in, since what I suggested will probably take more thought, but I don't think this one is actually it. Mixing tagged and untagged commands is illegal under the SCSI spec. Simply untagging the INQUIRY command may work to fix this particular bug, but I bet it will introduce others when a scsi utility sends an INQUIRY down to an in-use device. I think perhaps the correct palliative fix is simply to list this device (whatever it is, it's not listed in the bug report) as BLIST_NOTQ, then the problem won't arise (and of course, tmscsim would have actually to use the mid-layer black list rather than its own much shorter internal list). James