From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Darrick J. Wong" Subject: Re: Aic94xx and Linux kernel 2.6.19 Date: Mon, 13 Nov 2006 17:53:42 -0800 Message-ID: <455921A6.8080102@us.ibm.com> References: <834553.41356.qm@web31804.mail.mud.yahoo.com> <4557707B.8040701@us.ibm.com> <45577499.1010101@garzik.org> Reply-To: "Darrick J. Wong" Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from e36.co.us.ibm.com ([32.97.110.154]:61670 "EHLO e36.co.us.ibm.com") by vger.kernel.org with ESMTP id S1755383AbWKNBxo (ORCPT ); Mon, 13 Nov 2006 20:53:44 -0500 Received: from westrelay02.boulder.ibm.com (westrelay02.boulder.ibm.com [9.17.195.11]) by e36.co.us.ibm.com (8.13.8/8.12.11) with ESMTP id kAE1riTG001057 for ; Mon, 13 Nov 2006 20:53:44 -0500 Received: from d03av03.boulder.ibm.com (d03av03.boulder.ibm.com [9.17.195.169]) by westrelay02.boulder.ibm.com (8.13.6/8.13.6/NCO v8.1.1) with ESMTP id kAE1riGG511532 for ; Mon, 13 Nov 2006 18:53:44 -0700 Received: from d03av03.boulder.ibm.com (loopback [127.0.0.1]) by d03av03.boulder.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id kAE1rhPO014038 for ; Mon, 13 Nov 2006 18:53:44 -0700 In-Reply-To: <45577499.1010101@garzik.org> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Jeff Garzik Cc: ltuikov@yahoo.com, mike.redan@bell.ca, James.Bottomley@SteelEye.com, alexisb@us.ibm.com, linux-scsi Jeff Garzik wrote: > I'll queue it, but let me know results from your testing regardless. I no longer see disk errors; it looks like the commands are being retried at some point in the future, as I'd expect. However, now I periodically see this: BUG: warning at drivers/ata/libata-core.c:4597/ata_qc_issue() One cause of this is that sas_ata was writing the aic94xx sequencer's version of sactive into the libata's ata_port->sactive shadow register. Since we now know that aic94xx assigns its own NCQ tags, the solution is to avoid this copy, since libata manages ata_port->sactive anyway. Unfortunately, even with that patched up I still see the message, but only about 10% as frequently as I used to. I suspect that I haven't the right locking around ata_sas_queuecmd and ata_qc_complete; I'm testing that fix overnight. --D