Re: [PATCH v3 05/11] mfd: lm3533: Convert to use OF bindings
From: Svyatoslav Ryhel
Date: Tue Jun 02 2026 - 10:41:46 EST
вт, 2 черв. 2026 р. о 17:20 Jonathan Cameron <jic23@xxxxxxxxxx> пише:
>
> On Tue, 2 Jun 2026 16:50:16 +0300
> Svyatoslav Ryhel <clamor95@xxxxxxxxx> wrote:
>
> > вт, 2 черв. 2026 р. о 16:46 Jonathan Cameron <jic23@xxxxxxxxxx> пише:
> > >
> > > On Mon, 1 Jun 2026 18:18:25 +0300
> > > Svyatoslav Ryhel <clamor95@xxxxxxxxx> wrote:
> > >
> > > > Since there are no users of this driver via platform data, remove the
> > > > platform data support and switch to using Device Tree bindings.
> > > >
> > > > Signed-off-by: Svyatoslav Ryhel <clamor95@xxxxxxxxx>
> > >
> > >
> > > > index 52136ca1abc9..55b35467a722 100644
> > > > --- a/drivers/iio/light/lm3533-als.c
> > > > +++ b/drivers/iio/light/lm3533-als.c
> > > > @@ -16,16 +16,19 @@
> > > > #include <linux/module.h>
> > > > #include <linux/mutex.h>
> > > > #include <linux/mfd/core.h>
> > > > +#include <linux/mod_devicetable.h>
> > > > #include <linux/platform_device.h>
> > > > +#include <linux/property.h>
> > > > #include <linux/regmap.h>
> > > > #include <linux/slab.h>
> > > > #include <linux/uaccess.h>
> > > > +#include <linux/units.h>
> > > >
> > > > #include <linux/mfd/lm3533.h>
> > > >
> > > >
> > > > -#define LM3533_ALS_RESISTOR_MIN 1
> > > > -#define LM3533_ALS_RESISTOR_MAX 127
> > > > +#define LM3533_ALS_RESISTOR_MIN 1575
> > > > +#define LM3533_ALS_RESISTOR_MAX 200000
> > > > #define LM3533_ALS_CHANNEL_CURRENT_MAX 2
> > > > #define LM3533_ALS_THRESH_MAX 3
> > > > #define LM3533_ALS_ZONE_MAX 4
> > > > @@ -57,6 +60,9 @@ struct lm3533_als {
> > > >
> > > > atomic_t zone;
> > > > struct mutex thresh_mutex;
> > > > +
> > > > + bool pwm_mode;
> > > > + u32 r_select;
> > > > };
> > > >
> > > >
> > > > @@ -411,7 +417,7 @@ static ssize_t show_thresh_either_en(struct device *dev,
> > > > int enable;
> > > > int ret;
> > > >
> > > > - if (als->irq) {
> > > > + if (als->irq > 0) {
> > > > ret = lm3533_als_get_int_mode(indio_dev, &enable);
> > > > if (ret)
> > > > return ret;
> > > > @@ -716,30 +722,34 @@ static const struct attribute_group lm3533_als_attribute_group = {
> > > > .attrs = lm3533_als_attributes
> > > > };
> > > >
> > > > -static int lm3533_als_setup(struct lm3533_als *als,
> > > > - const struct lm3533_als_platform_data *pdata)
> > > > +static int lm3533_als_setup(struct lm3533_als *als)
> > > > {
> > > > struct device *dev = &als->pdev->dev;
> > > > int ret;
> > > >
> > > > + als->pwm_mode = device_property_read_bool(dev, "ti,pwm-mode");
> > > > +
> > > > ret = regmap_update_bits(als->lm3533->regmap, LM3533_REG_ALS_CONF,
> > > > LM3533_ALS_INPUT_MODE_MASK,
> > > > - pdata->pwm_mode ? LM3533_ALS_INPUT_MODE_MASK : 0);
> > > > + als->pwm_mode ? LM3533_ALS_INPUT_MODE_MASK : 0);
> > > > if (ret)
> > > > return dev_err_probe(dev, ret, "failed to set input mode %d\n",
> > > > - pdata->pwm_mode);
> > > > -
> > > > + als->pwm_mode);
> > > >
> > > > /* ALS input is always high impedance in PWM-mode. */
> > > > - if (!pdata->pwm_mode) {
> > > > - if (pdata->r_select < LM3533_ALS_RESISTOR_MIN ||
> > > > - pdata->r_select > LM3533_ALS_RESISTOR_MAX) {
> > > > - dev_err(&als->pdev->dev, "invalid resistor value\n");
> > > > - return -EINVAL;
> > > > - }
> > > > + if (!als->pwm_mode) {
> > > > + ret = device_property_read_u32(dev, "ti,resistor-value-ohms",
> > > > + &als->r_select);
> > > > + if (ret)
> > > > + return dev_err_probe(dev, ret,
> > > > + "failed to ger resistor value\n");
> > > > +
> > > > + als->r_select = clamp(als->r_select, LM3533_ALS_RESISTOR_MIN,
> > > > + LM3533_ALS_RESISTOR_MAX);
> > >
> > > If we are getting garbage from DT I think I'd rather error out that paper over
> > > that problem. So similar to before, check valid value and if not fail probe
> > > so that hopefully someone goes and fixes it!
> > >
> >
> > sure
> >
> > > > + als->r_select = DIV_ROUND_UP(2 * MICRO, 10 * als->r_select);
> > > Why do we need this when we didn't before? The range checks are the same
> > > so it smells like it shouldn't need transforming. I'd also rather we didn't do
> > > rewriting of the meaning of r_select like this. Just use a local variable for
> > > the intermediate result.
> > >
> >
> > before pdata passed resistor value as actual register value, not we
> > are getting the actual resistance in ohms from the tree and must
> > convert it into register value.
>
> ah. I missed the change of values. Can you make them explicitly now _OHMS or something
> along those lines rather than reusing the macro name for a different thing.
>
Acknowledged.
> >
> > > >
> > > > ret = regmap_write(als->lm3533->regmap, LM3533_REG_ALS_RESISTOR_SELECT,
> > > > - pdata->r_select);
> > > > + als->r_select);
> > > > if (ret)
> > > > return dev_err_probe(dev, ret, "failed to set resistor\n");
>