From mboxrd@z Thu Jan 1 00:00:00 1970 From: Benjamin Herrenschmidt Subject: Re: [PATCH] fix sata_sil compilation on non-DMI platforms Date: Tue, 12 May 2009 08:31:44 +1000 Message-ID: <1242081104.304.12.camel@pasglop> References: <200905062009.28087.markos@codex.gr> <4A086A76.3090008@garzik.org> <20090511194410.46641214@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20090511194410.46641214@lxorguk.ukuu.org.uk> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linuxppc-dev-bounces+glppe-linuxppc-embedded-2=m.gmane.org@ozlabs.org Errors-To: linuxppc-dev-bounces+glppe-linuxppc-embedded-2=m.gmane.org@ozlabs.org To: Alan Cox Cc: Jeff Garzik , linuxppc-dev@ozlabs.org, Konstantinos Margaritis , linux-ide@vger.kernel.org, rmk@arm.linux.org.uk, Lennart Sorensen , Lennert Buytenhek List-Id: linux-ide@vger.kernel.org On Mon, 2009-05-11 at 19:44 +0100, Alan Cox wrote: > > Ideally the DMI subsystem should be provided wrappers for platforms > > without DMI, rendering patches like this unnecessary. > > Interesting question - is the PPC OpenFirmware machine information > mappable onto the DMI space ? or the various static bits of name > information in the various ARM and the like machine descriptions. > > If not which bits are similar enough we could replace dmi at the high > level with an abstract interface for system/vendor/... that was ? The one thing we could try to map would be the device-tree "compatible" property which would contain vendor,system tuples from more specific to more generic with which the machine is compatible with. Of course I would argue the other way around and create a device-tree from the DMI data :-) Cheers, Ben.