From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andy Warner Subject: Re: aic94xx IO errors with "escb_tasklet_complete: phy0: REQ_TASK_ABORT" Date: Wed, 4 Oct 2006 14:31:54 -0500 Message-ID: <20061004143154.A29888@florence.linkmargin.com> References: <20061004164438.GC5091@rhun.haifa.ibm.com> <20061004132929.A29689@florence.linkmargin.com> <452415AE.7030202@us.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from ms-smtp-01.rdc-kc.rr.com ([24.94.166.115]:19645 "EHLO ms-smtp-01.rdc-kc.rr.com") by vger.kernel.org with ESMTP id S1751084AbWJDUcE (ORCPT ); Wed, 4 Oct 2006 16:32:04 -0400 Content-Disposition: inline In-Reply-To: <452415AE.7030202@us.ibm.com>; from djwong@us.ibm.com on Wed, Oct 04, 2006 at 01:12:30PM -0700 Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: "Darrick J. Wong" Cc: Andy Warner , Muli Ben-Yehuda , linux-scsi , James Bottomley Darrick J. Wong wrote: > Andy Warner wrote: > > > So you are seeing legitimate REQ_TASK_ABORT values, but need to look > > at the remaining data to see what the chip is trying to tell you. > > For REQ_TASK_ABORT, status_block[1..2] is the transaction context, > > and status_block[3] is the reason (TC_NO_ERROR etc from aic94xx_sas.h) > > Funny, I've been trying to write at least a REQ_TASK_ABORT handler all > week. I'm having a bit of trouble using the transaction context... > mostly because the value I'm getting doesn't correspond to any ascb. > Assuming, of course, that ascb->scb->header.index/ascb->tc_index are > indeed transaction contexts. Check you're actually getting REQ_TASK_ABORT, and not REQ_DEVICE_RESET or SIGNAL_NCQ_ERROR, where the same field is a connection handle, and not a transaction context. -- andyw@pobox.com Andy Warner Voice: (612) 801-8549 Fax: (208) 575-5634