Re: [PATCH v6 10/10] clk: bd718x7: Initial support for ROHM bd71837/bd71847 PMIC clock

[Matti Vaittinen]

Thanks again Stephen.

On Wed, Dec 05, 2018 at 09:28:21AM -0800, Stephen Boyd wrote:
> Quoting Matti Vaittinen (2018-12-04 03:39:38)
> > diff --git a/drivers/clk/clk-bd718x7.c b/drivers/clk/clk-bd718x7.c
> > new file mode 100644
> > index 000000000000..d486859526ed
> > --- /dev/null
> > +++ b/drivers/clk/clk-bd718x7.c
> > @@ -0,0 +1,131 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +// Copyright (C) 2018 ROHM Semiconductors
> > +// bd71837.c  -- ROHM BD71837MWV clock driver
> 
> This isn't even the name of the file. Please remove. Also, only the SPDX
> tag is supposed to have // on it and the other things should be normal
> /* */ comment style.

I'll drop this name.

> 
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/module.h>
> > +#include <linux/init.h>
> > +#include <linux/err.h>
> > +#include <linux/platform_device.h>
> > +#include <linux/slab.h>
> > +#include <linux/mfd/rohm-bd718x7.h>
> > +#include <linux/clk-provider.h>
> > +#include <linux/clkdev.h>
> > +#include <linux/regmap.h>
> [....]
> > +
> > +static int bd71837_clk_probe(struct platform_device *pdev)
> > +{
> > +       struct bd718xx_clk *c;
> > +       int rval = -ENOMEM;
> > +       const char *parent_clk;
> > +       struct device *parent = pdev->dev.parent;
> > +       struct bd718xx *mfd = dev_get_drvdata(parent);
> > +       struct clk_init_data init = {
> > +               .name = "bd718xx-32k-out",
> > +               .ops = &bd71837_clk_ops,
> > +       };
> > +
> > +       c = devm_kzalloc(&pdev->dev, sizeof(*c), GFP_KERNEL);
> > +       if (!c)
> > +               return -ENOMEM;
> > +
> > +       init.num_parents = 1;
> > +       parent_clk = of_clk_get_parent_name(parent->of_node, 0);
> > +
> > +       init.parent_names = &parent_clk;
> > +       if (!parent_clk) {
> > +               dev_err(&pdev->dev, "No parent clk found\n");
> > +               return -EINVAL;
> > +       }
> > +
> > +       c->reg = BD718XX_REG_OUT32K;
> > +       c->mask = BD718XX_OUT32K_EN;
> > +       c->mfd = mfd;
> > +       c->pdev = pdev;
> > +       c->hw.init = &init;
> > +
> > +       of_property_read_string_index(parent->of_node,
> > +                                     "clock-output-names", 0, &init.name);
> > +
> > +       rval = devm_clk_hw_register(&pdev->dev, &c->hw);
> > +       if (!rval) {
> > +               rval = devm_clk_hw_register_clkdev(&pdev->dev,
> > +                                                  &c->hw, init.name, NULL);
> 
> Do you plan to use the clkdev lookup? This driver looks fairly DT
> dependent, so I would prefer to remove the clkdev part and only add it
> later if anyone needs it.

Right. We depend heavily on DT so I guess I can drop the clkdev portion.

> > +               if (rval)
> > +                       dev_warn(&pdev->dev, "Failed to register clkdev\n");
> > +               if (parent->of_node) {
> > +                       rval = devm_of_clk_add_hw_provider(&pdev->dev,
> > +                                            of_clk_hw_simple_get, &c->hw);
> > +                       if (rval)
> > +                               dev_err(&pdev->dev,
> > +                                       "adding clk provider failed\n");
> > +               }

-- 
Matti Vaittinen
ROHM Semiconductors

~~~ "I don't think so," said Rene Descartes.  Just then, he vanished ~~~