From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from wolverine02.qualcomm.com ([199.106.114.251]:60755 "EHLO wolverine02.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753053AbaHSHi2 (ORCPT ); Tue, 19 Aug 2014 03:38:28 -0400 From: Kalle Valo To: Srinivas Kandagatla CC: , , Subject: Re: ath6kl DT support and multiple instances. References: <53F2FB31.3090802@linaro.org> Date: Tue, 19 Aug 2014 10:38:19 +0300 In-Reply-To: <53F2FB31.3090802@linaro.org> (Srinivas Kandagatla's message of "Tue, 19 Aug 2014 08:22:25 +0100") Message-ID: <87oavhapzo.fsf@kamboji.qca.qualcomm.com> (sfid-20140819_093853_542833_A3FD38CA) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-wireless-owner@vger.kernel.org List-ID: Hi Srinivas, Srinivas Kandagatla writes: > I noticed that ath6kl partially supports Device trees, via > check_device_tree() function in init.c. That's really old code. Sam Leffler wrote that back in 2011 and IIRC nobody has touched it since. > I have a use case for adding reset line and pinctrl properties to this > dt bindings, but soon realized that this function would not work for > multiple instances of the driver, as there is no direct > association/binding of dt node and driver instance. > > Just wondering if this is a know limitation and there will be no > possible multiple instances of this driver? > > is check_device_tree function obsolete? I'm not familiar with the device tree at all so I can't really comment much, except that I'm happy to take patches :) > should I care about multiple instances? Yes, I think you should. > Other than adding static code in the board file, are there any other > options to pass platform data like reset gpio lines and pinctrl > properties to the driver? Few years back we talked about adding function pointers to the board files for that, but there never were any patches. -- Kalle Valo