From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Lord Subject: Re: [PATCH 6/8] libata: adjust speed down rules Date: Tue, 27 Nov 2007 20:10:47 -0500 Message-ID: <474CC017.3040305@rtr.ca> References: <11961593392491-git-send-email-htejun@gmail.com> <11961593403415-git-send-email-htejun@gmail.com> <474C9B10.4060500@rtr.ca> <474CA31E.4000508@gmail.com> <474CA94E.9080603@rtr.ca> <474CA9C4.9040902@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from rtr.ca ([76.10.145.34]:3700 "EHLO mail.rtr.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759985AbXK1BKt (ORCPT ); Tue, 27 Nov 2007 20:10:49 -0500 In-Reply-To: <474CA9C4.9040902@gmail.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Tejun Heo Cc: jeff@garzik.org, linux-ide@vger.kernel.org, bzolnier@gmail.com Tejun Heo wrote: > Mark Lord wrote: >> Hi Tejun, >> >> Tejun Heo wrote: >>> Mark Lord wrote: >> .. >>>> -- deliberate (or otherwise) errors via SG_IO will cause inappropriate >>>> speed-downs. >>> No, only device errors issued via block layer are considered for speed >>> down (ATA_QCFLAG_IO). SG_IO commands will never trigger speed down >>> unless the error is HSM or timeout. >> .. >> >> Mmmm... what category is "device rejected the command" (ABRT) ? >> Is that HSM ? > > Nope, that's UNK_DEV if the command is IO (came from block layer) and > cat-0 (don't care) if it has come from SG_IO. .. Perfect, thanks! Last question: what about CD/DVD writers? Are the data-out commands for those done via block, or via SG_IO ? (my memory fails me here). Do they ever get speed up/down based on WRITE errors ? Cheers