Re: [PATCH 08/13] regulator: max77650: add regulator support

[Bartosz Golaszewski]

pt., 18 sty 2019 o 19:01 Mark Brown <broonie@xxxxxxxxxx> napisaÅ(a):
>
> On Fri, Jan 18, 2019 at 02:42:39PM +0100, Bartosz Golaszewski wrote:
>
> > Add regulator support for max77650. We support all four variants of this
> > PMIC including non-linear voltage table for max77651 SBB1 rail.
>

Sorry for spamming, I missed it the last time.

> Looks good, the ramping stuff might be a candidate for core (TBH I was
> sure we'd got that implemented already but we don't seem to) but that
> can be done later and the more complex one with non-linear steps does
> feel like it might have to stay in the driver anyway.
>

I'm thinking about implementing that in the core for the next release
cycle and then making this driver the first user.

Bart

> A couple of small nits:
>
> > +++ b/drivers/regulator/max77650-regulator.c
> > @@ -0,0 +1,537 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Copyright (C) 2018 BayLibre SAS
> > + * Author: Bartosz Golaszewski <bgolaszewski@xxxxxxxxxxxx>
>
> Please make the entire header C++ style so it looks more intentional.
>
> > +             for_each_child_of_node(dev->of_node, child) {
> > +                     if (!of_node_name_eq(child, rdesc->desc.name))
> > +                             continue;
> > +
> > +                     init_data = of_get_regulator_init_data(dev, child,
> > +                                                            &rdesc->desc);
> > +                     if (!init_data)
> > +                             return -ENOMEM;
> > +
> > +                     config.of_node = child;
> > +                     config.init_data = init_data;
> > +             }
>
> You don't need to do this, the core will do it for you (it will actually
> still do it even with the above, it'll only fall back to using
> config->init_data if it's own lookup fails).