From mboxrd@z Thu Jan  1 00:00:00 1970
From: marex@denx.de (Marek Vasut)
Date: Wed, 20 Mar 2013 15:03:20 +0100
Subject: [PATCH V4 6/9] ARM: mxs: Provide regulator to pwm-backlight
In-Reply-To: <20130320082318.GA28775@opensource.wolfsonmicro.com>
References: <1363719573-20926-1-git-send-email-achew@nvidia.com>
 <5148E252.9050900@wwwdotorg.org>
 <20130320082318.GA28775@opensource.wolfsonmicro.com>
Message-ID: <201303201503.20480.marex@denx.de>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

Hi Mark,

> On Tue, Mar 19, 2013 at 04:10:26PM -0600, Stephen Warren wrote:
> > On 03/19/2013 03:27 PM, Marek Vasut wrote:
> > > Do we really need a mandatory regulator? Why can't it be optional?
> > 
> > IIRC, the previous advice I've seen is that if a device (driver) uses a
> > regulator, it must /require/ a regulator, and if a particular board
> > doesn't actually have a SW-controlled regulator, then a fixed- or dummy-
> > regulator should be provided to satisfy this requirement.
> > 
> > CC'ing Mark Brown to make sure I really do Recall Correctly.
> 
> Yes, and it should be fixed rather than dummy.  The issue is partly that
> it's probably important that the device has power so we don't want to
> just ignore errors and partly that this is something which applies to
> essentially all devices so whatever we do for this case ought to be done
> by the core so all devices can benefit and we don't have to duplicate
> lots of code in individual drivers.

Thanks for clearing this!

Best regards,
Marek Vasut