Re: [PATCH] iio: light: add support for ROHM BH1710/BH1715/BH1721/BH1750/BH1751 ambient light sensors

[Tomasz Duszynski <tduszyns@gmail.com>]

On Thu, Apr 23, 2015 at 10:30:09AM +0300, Daniel Baluta wrote:
> On Tue, Apr 21, 2015 at 9:43 PM, Tomasz Duszynski <tduszyns@gmail.com> wrote:
> > Add support for ROHM BH1710/BH1715/BH1721/BH1750/BH1751 ambient light
> > sensors.
> >
> > Signed-off-by: Tomasz Duszynski <tduszyns@gmail.com>
> > ---
> >  drivers/iio/light/Kconfig  |  10 ++
> >  drivers/iio/light/Makefile |   1 +
> >  drivers/iio/light/bh1750.c | 322 +++++++++++++++++++++++++++++++++++++++++++++
> >  3 files changed, 333 insertions(+)
> >  create mode 100644 drivers/iio/light/bh1750.c
> >
> > diff --git a/drivers/iio/light/Kconfig b/drivers/iio/light/Kconfig
> > index 5a3237b..9fb79ca 100644
> > --- a/drivers/iio/light/Kconfig
> > +++ b/drivers/iio/light/Kconfig
> > @@ -37,6 +37,16 @@ config APDS9300
> >          To compile this driver as a module, choose M here: the
> >          module will be called apds9300.
> >
> > +config BH1750
> > +       tristate "BH1750 ambient light sensor"
>
> Better use ROHM BH1750 .. here.
>
>
Did git blame on Kconfig and among newest sensors some
have this manufacturer prefix while others don't, but still no problem
for me to add that.
> > +static const struct iio_chan_spec bh1750_channels[] = {
> > +       {
> > +               .type = IIO_INTENSITY,
>
> Shouldn't this be IIO_LIGHT channel type?
>
> Users will get illuminance after multiplying raw value with scale.
>
That naming is misleading. Thanks for pointing this out.
> > +        .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
> > +                                     BIT(IIO_CHAN_INFO_SCALE) |
> > +                                     BIT(IIO_CHAN_INFO_INT_TIME)
> > +       }
> > +};
> > +
> > +static int bh1750_probe(struct i2c_client *client,
> > +                       const struct i2c_device_id *id)
> > +{
> > +       int ret, usec;
> > +       struct bh1750_data *data;
> > +       struct iio_dev *indio_dev;
> > +
> > +       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C |
> > +                               I2C_FUNC_SMBUS_WRITE_BYTE))
> > +               return -ENODEV;
> > +
> > +       indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*data));
> > +       if (!indio_dev)
> > +               return -ENOMEM;
> > +
> > +       data = iio_priv(indio_dev);
> > +       i2c_set_clientdata(client, indio_dev);
> > +       data->client = client;
> > +       data->chip_info = &chip_info_tbl[id->driver_data];
> > +
> > +       usec = data->chip_info->mtreg_to_usec * data->chip_info->mtreg_default;
> > +       ret = bh1750_change_int_time(data, usec);
> > +       if (ret < 0)
> > +               return ret;
> > +
> > +       mutex_init(&data->lock);
> > +       indio_dev->dev.parent = &client->dev;
> > +       indio_dev->info = &bh1750_info;
> > +       indio_dev->name = id->name;
> > +       indio_dev->channels = bh1750_channels;
> > +       indio_dev->num_channels = ARRAY_SIZE(bh1750_channels);
> > +       indio_dev->modes = INDIO_DIRECT_MODE;
> > +
> > +       return devm_iio_device_register(&client->dev, indio_dev);
>
> You need also to add a remove function where the chip is powered down. Then
> you shouldn't use devm_iio_device_register here, because you need to control
> the order of chip powerdown/ device unregister operations at remove.
>
Right.
> thanks,
> Daniel.

Thanks for review!