From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dong Aisheng Subject: Re: [PATCH v1 1/5] ARM: imx28: add basic dt support Date: Fri, 16 Mar 2012 16:22:44 +0800 Message-ID: <20120316082243.GA13065@shlinux2.ap.freescale.net> References: <1331628428-24017-2-git-send-email-b29396@freescale.com> <20120313172351.97C5E3E053B@localhost> <20120314124512.GG1180@shlinux2.ap.freescale.net> <20120314141643.GP3852@pengutronix.de> <20120315030206.GB13022@shlinux2.ap.freescale.net> <20321.37353.63115.664975@ipc1.ka-ro> <20120315105927.GE13022@shlinux2.ap.freescale.net> <20321.53468.310199.875966@ipc1.ka-ro> <20120316030134.GA5161@shlinux2.ap.freescale.net> <20322.61501.432329.124466@ipc1.ka-ro> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from ch1ehsobe006.messaging.microsoft.com ([216.32.181.186]:58728 "EHLO ch1outboundpool.messaging.microsoft.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S965186Ab2CPIVg convert rfc822-to-8bit (ORCPT ); Fri, 16 Mar 2012 04:21:36 -0400 Content-Disposition: inline In-Reply-To: <20322.61501.432329.124466@ipc1.ka-ro> Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Lothar =?iso-8859-1?Q?Wa=DFmann?= Cc: Dong Aisheng-B29396 , "s.hauer@pengutronix.de" , Guo Shawn-R65073 , "vinod.koul@linux.intel.com" , "devicetree-discuss@lists.ozlabs.org" , "linux-mmc@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "rob.herring@calxeda.com" , Grant Likely , "rdunlap@xenotime.net" , "kernel@pengutronix.de" , "cjb@laptop.org" , "linux-arm-kernel@lists.infradead.org" On Fri, Mar 16, 2012 at 03:48:13PM +0800, Lothar Wa=DFmann wrote: > Hi, >=20 > Dong Aisheng writes: > > On Thu, Mar 15, 2012 at 07:22:04PM +0800, Lothar Wa=DFmann wrote: > [...] > > My proposal is only set the fixed part(first two octets) in board d= ts file, > > each board knows it's value, and read the left 4 octets from OCOTP = dynamically. > > > The OUI part is three bytes, not two! But anyway, since there is no > definition on how the MAC has to be stored in OCOTP there is no way > for the driver to know how to interpret the data it may read from > there. >=20 I may miss this, i will double check it. Thanks for the info. :-) > > > Anyway there is no definite spec how the MAC address(es) are stor= ed > > > in the fuse map. Thus reading the MAC from there is more or less > > > platform specific. > > >=20 > > It's just provide one more option since there are customers storing= the MAC > > in the fuse map. > >=20 > How would the driver know, whether and how the customer has stored th= e > MAC address in OCOTP? E.g. on our TX28 modules the FULL MAC is stored > in the CUSTOM registers in OCOTP. >=20 > > > Currently I'm setting up the MAC address for our TX28 from the fu= ses > > > in the platform code passed via platform_data, but that will obvi= ously > > > not work with DT. > > >=20 > > Non-dt can also use my proposal, then you only need to pass the fix= ed part of > > MAC via platfrom data, the left will be read by fec driver. > >=20 > Reading the MAC from fuses is platform sepcific. The driver cannot > know how the MAC is stored there and needs to rely on platform > specific code to do this. >=20 > > > The correct way would probably be to pass the MAC from the bootlo= ader > > > via a DT blob. But that would require all bootloaders to be updat= ed > > > first to support DTS. :( > > >=20 > > But uboot will face the same problem that can not define a valid MA= C > > in dts, did i undertand correctly? > >=20 > That's why I said "would require all bootloaders to be updated first > to support DTS". Regards Dong Aisheng