From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from wf-out-1314.google.com (wf-out-1314.google.com [209.85.200.169]) by ozlabs.org (Postfix) with ESMTP id 6E491DDEE3 for ; Sat, 5 Jul 2008 16:53:08 +1000 (EST) Received: by wf-out-1314.google.com with SMTP id 24so1537146wfg.15 for ; Fri, 04 Jul 2008 23:53:07 -0700 (PDT) Date: Sat, 5 Jul 2008 00:53:04 -0600 From: Grant Likely To: Segher Boessenkool , Matt Sealey , linuxppc-dev list Subject: Re: New fsl device bindings file Message-ID: <20080705065304.GA20762@secretlab.ca> References: <40B88BFD-1FEB-4E36-9917-54C7380FF801@kernel.crashing.org> <20080703185312.GA6043@polina.dev.rtsoft.ru> <486E26C0.9060604@genesi-usa.com> <20080705012624.GA3032@yookeroo.seuss> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 In-Reply-To: <20080705012624.GA3032@yookeroo.seuss> Sender: Grant Likely List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Sat, Jul 05, 2008 at 11:26:24AM +1000, David Gibson wrote: > On Sat, Jul 05, 2008 at 01:12:31AM +0200, Segher Boessenkool wrote: > >>> /* deprecated; */ > >>> device_type = "i2c"; > >> > >> How about "deprecated but kept for compatibility with true Open > >> Firmware > >> implementations"? > > > > Well, except a flat tree isn't compatible with OF at all here. > > A "device_type" promises a certain interface; a flat tree doesn't > > even have the "open" method. From the OF base spec: > > > > “device_type” S > > > > Standard property name to specify the implemented interface. > > > > prop-encoded-array: Text string encoded with encode-string. > > > > Specifies the “device type” of this package, thus implying a > > specific set of package class methods implemented by this > > package. > > > >> Seriously, you can't have a binding for "OF" and then cut out that > >> part of the > >> standard at a whim. > > > > Nothing is cut out. There never was a device binding for device_type > > i2c; creating one would be a considerable effort, and since flat tree > > users wouldn't use it anyway, you can't be seriously suggesting they > > should do this. > > > >> It should be there (at least for those parts which are > >> governed by a client interface API, like display, serial etc. > > > > Huh? Nothing in the client interface mentions display or serial > > as far as I know. > > > >> but cutting it off takes away all it's meaning, > > > > So what? There _is_ no "real" device interface, when a flat tree is > > used. > > > >> plus Linux implementations STILL keep searching > >> that property along with "compatible", > > > > That's a bug. > > Thank you Segher, you saved me the trouble of saying exactly all that. Ditto. g.