From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from imap.sh.mvista.com (unknown [63.81.120.155]) by ozlabs.org (Postfix) with ESMTP id 6734EDE058 for ; Thu, 3 May 2007 23:27:08 +1000 (EST) Message-ID: <4639E389.4040604@ru.mvista.com> Date: Thu, 03 May 2007 17:28:41 +0400 From: Sergei Shtylyov MIME-Version: 1.0 To: David Gibson Subject: Re: powerpc_flash_init(), wtf!? References: <20070501051804.GB3881@localhost.localdomain> <20070503103534.63ff67b6@localhost.localdomain> <20070503070358.GA9430@localhost.localdomain> <4639CF44.6020608@ru.mvista.com> <20070503122213.GD26659@localhost.localdomain> In-Reply-To: <20070503122213.GD26659@localhost.localdomain> Content-Type: text/plain; charset=us-ascii; format=flowed Cc: linuxppc-dev@ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hello. David Gibson wrote: >>>>>Unless someone who actually knows how this code was intended to be >>>>>used can suggest a more polite way of fixing it. >>>>I guess, the idea was for this stuff to be updated once one of the >>>>dts inside boot/ would have physmap nodes added. I have >>>>rom/physmap[dts] rehaul in my TODO list, but it has (so far at >>>>least) little chance to happen during this merge window. Yet, if >>>>someone has suggestions and/or some interest for this to be cured, >>>>it will gain priority. Otherwise, I'll replace actual erroneous code >>>>with kind of rant that it's up to BSP code to take care of >>>>of_devices to be registered, using of_platform_bus_probe() or other >>>>way. >>>I'm having some trouble parsing that paragraph. At this stage I don't >>>see any reason to hold off on tearing out arch/powerpc/sysdev/rom.c, >>>any necessary changes to replace it will go in the platform code or >> It doesn't seem a flexible enough approach. We could continue using >>platform devices then. > Sorry, I don't follow you. Well, IIUC, the idea behind the device is to free the platform code of as much burden of registering the platform devices itself as possible, isn't it? >>>other places. >> Any ideas where else? > Not really. I don't immediately see a case where doing it from the > platform code wouldn't be right. BTW, is it legal/appropriate to specify device (not bus) types for of_platform_bus_probe()? WBR, Sergei