From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Garrett Subject: Re: [patch 3/3] Enable Aggressive Link Power management for AHCI controllers. Date: Tue, 12 Jun 2007 16:56:29 +0100 Message-ID: <20070612155629.GA14044@srcf.ucam.org> References: <466DFE09.3020408@garzik.org> <20070612035928.GB9808@khazad-dum.debian.net> <466E1A32.90306@linux.intel.com> <20070612090940.GA9719@srcf.ucam.org> <20070612121819.GA20408@khazad-dum.debian.net> <20070612135010.GA12601@srcf.ucam.org> <20070612141714.GB20408@khazad-dum.debian.net> <20070612153828.GA13888@srcf.ucam.org> <466EBF91.8060508@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <466EBF91.8060508@gmail.com> Sender: linux-ide-owner@vger.kernel.org To: Tejun Heo Cc: Henrique de Moraes Holschuh , Arjan van de Ven , Jeff Garzik , Dagfinn Ilmari =?iso-8859-1?Q?Manns=E5ker?= , Kristen Carlson Accardi , james.bottomley@steeleye.com, linux-ide@vger.kernel.org, linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: linux-scsi@vger.kernel.org On Wed, Jun 13, 2007 at 12:45:21AM +0900, Tejun Heo wrote: > Matthew Garrett wrote: > > Yes, but they'll also send an ACPI interrupt even if the SATA host > > controller doesn't - it's part of the spec for bays. > > Does the spec mandate that the ACPI interrupt shouldn't depend on SATA > phy status? I don't think vendors are likely to implement separate > mechanism when SATA phy status can do the job fine. I suspect that the spec allows them to do that, but think that it's unlikely to actually happen in most cases. Bear in mind that Windows doesn't tend to drive the hardware in AHCI mode, and that their implementation is likely to be very similar to the implementation they used for PATA devices. -- Matthew Garrett | mjg59@srcf.ucam.org