[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAMRc=MenTgffszv4NsbCKRhH0TcRPSTLbeP3BttW9fmFBjLdCA@mail.gmail.com>
Date: Tue, 10 Dec 2019 14:10:15 +0100
From: Bartosz Golaszewski <brgl@...ev.pl>
To: Mark Brown <broonie@...nel.org>
Cc: Liam Girdwood <lgirdwood@...il.com>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Bartosz Golaszewski <bgolaszewski@...libre.com>
Subject: Re: [PATCH] regulator: max77650: add of_match table
wt., 10 gru 2019 o 14:02 Mark Brown <broonie@...nel.org> napisał(a):
>
> On Tue, Dec 10, 2019 at 01:51:38PM +0100, Bartosz Golaszewski wrote:
> > wt., 10 gru 2019 o 13:12 Mark Brown <broonie@...nel.org> napisał(a):
>
> > > Why would we need to use a compatible string in a child node to load the
> > > regulator driver, surely we can just register a platform device in the
> > > MFD?
>
> > The device is registered all right from MFD code, but the module won't
> > be loaded automatically from user-space even with the right
> > MODULE_ALIAS() for sub-nodes unless we define the
> > MODULE_DEVICE_TABLE().
>
> This seems to work fine for other drivers and the platform bus has to be
> usable on systems that don't use DT so that doesn't sound right. Which
> MODULE_ALIAS() are you using exactly?
>
MODULE_ALIAS("platform:max77650-regulator");
> > Besides: the DT bindings define the compatible for sub-nodes already.
> > We should probably conform to that.
>
> I would say that's a mistake and should be fixed, this particular way of
> loading the regulators is a Linux implementation detail.
Fixed by removing this from the bindings?
Bartosz
Powered by blists - more mailing lists