From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Martin K. Petersen" Subject: Re: [PATCH 04 of 10] scsi: Support devices with protection information (DIF) Date: Mon, 30 Jun 2008 11:24:35 -0400 Message-ID: References: <2fd66ca994dbf1ed6d14.1214407363@sermon.lab.mkp.net> <1214665658.3658.18.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from rgminet01.oracle.com ([148.87.113.118]:36156 "EHLO rgminet01.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760311AbYF3PYr (ORCPT ); Mon, 30 Jun 2008 11:24:47 -0400 In-Reply-To: <1214665658.3658.18.camel@localhost.localdomain> (James Bottomley's message of "Sat\, 28 Jun 2008 10\:07\:38 -0500") Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: James Bottomley Cc: "Martin K. Petersen" , linux-scsi@vger.kernel.org >>>>> "James" == James Bottomley writes: [Protection sgl alloc + init] James> I'm afraid you can't do it like this because it will violate James> our forward progress guarantees. If this is the last spare James> command required for writeout, you need to guarantee either James> this allocation will succeed, or we can proceed without the DIF James> data. We can't just turn off DIF -- that would violate the integrity guarantees. But I'll move the protection sgl to the host command pool. >> + if (inq_result[5] & 0x1) + sdev->protection = 1; James> Is there a reason to have a separate flag here Not really. I used to key off of sdp->protection quite a bit but I think the only place I use it now is during sd discovery. So this can simply go away... -- Martin K. Petersen Oracle Linux Engineering