From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: [ACPI] Re: RFC: ACPI/scsi/libata integration and hotswap Date: Thu, 08 Dec 2005 09:14:30 -0500 Message-ID: <43983FC6.6050108@pobox.com> References: <20051208030242.GA19923@srcf.ucam.org> <20051208091542.GA9538@infradead.org> <20051208132657.GA21529@srcf.ucam.org> <20051208133308.GA13267@infradead.org> <20051208133945.GA21633@srcf.ucam.org> <20051208135225.GA13122@havoc.gtf.org> <1134050863.17102.5.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail.dvmed.net ([216.237.124.58]:22952 "EHLO mail.dvmed.net") by vger.kernel.org with ESMTP id S932131AbVLHOOt (ORCPT ); Thu, 8 Dec 2005 09:14:49 -0500 In-Reply-To: <1134050863.17102.5.camel@localhost.localdomain> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Alan Cox Cc: Matthew Garrett , Christoph Hellwig , randy_d_dunlap@linux.intel.com, linux-ide@vger.kernel.org, linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org, acpi-devel@lists.sourceforge.net Alan Cox wrote: > On Iau, 2005-12-08 at 08:52 -0500, Jeff Garzik wrote: > >>On Thu, Dec 08, 2005 at 01:39:45PM +0000, Matthew Garrett wrote: >> >>>On Thu, Dec 08, 2005 at 01:33:08PM +0000, Christoph Hellwig wrote: >>> >>> >>>>Don't do it at all. We don't need to fuck up every layer and driver for >>>>intels braindamage. >>> >>>Doing SATA suspend/resume properly on x86 depends on knowing the ACPI >>>object that corresponds to a host or target. >> >>Not true. > > > > Actually he is right. You have to know the ACPI object in order to run > the _GTM/_STM etc functions. If you don't run those your suspend/resume These are only for PATA. We don't care about _GTM/_STM on SATA. Further, SATA completely resets and re-initializes the device as if from a hardware reset (except on ata_piix, which doesn't support COMRESET, and PATA). This makes _GTF uninteresting, as well. > may not work, may corrupt and so on. The only safe alternative is to > disable acpi which, while it would have been a good idea before the spec > ever got out, is a bit late now. suspend/resume works just fine with Jens' out-of-tree patch. > If you don't run the resume methods your disk subsystem status after a > resume is simply undefined and unsafe. I initialize the hardware to a defined state. Jeff