Re: [PATCH 1/1] iio adc: add driver for Texas Instruments TLA2528 adc

[Jonathan Cameron <Jonathan.Cameron@Huawei.com>]

...

> > +
> > +static int tla2528_probe(struct i2c_client *client,
> > +                        const struct i2c_device_id *id)
> > +{
> > +       struct iio_dev *indio_dev;
> > +       struct tla2528_st *st;
> > +       int ret;
> > +
> > +       if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C |
> > +                                    I2C_FUNC_SMBUS_WRITE_BYTE))
> > +               return -EOPNOTSUPP;
> > +
> > +       indio_dev = devm_iio_device_alloc(&client->dev, sizeof(*st));
> > +       if (!indio_dev)
> > +               return -ENOMEM;
> > +
> > +       st = iio_priv(indio_dev);
> > +       i2c_set_clientdata(client, indio_dev);
> > +       st->client = client;
> > +
> > +       indio_dev->name = id->name;
> > +       indio_dev->info = &tla2528_info;
> > +       indio_dev->modes = INDIO_DIRECT_MODE;
> > +       indio_dev->channels = tla2528_channel;
> > +       indio_dev->num_channels = ARRAY_SIZE(tla2528_channel);
> > +
> > +       st->ref = devm_regulator_get(&client->dev, "vref");
> > +       if (IS_ERR(st->ref))
> > +               return PTR_ERR(st->ref);
> > +
> > +       ret = regulator_enable(st->ref);
> > +       if (ret < 0)
> > +               return ret;
> > +  
> 
> I was thinking about whether to mention this earlier.
> But, an idea, is to do:
> 
>         ret = devm_add_action_or_reset(&spi->dev, tla2528_reg_disable,
> st->vref);
>         if (ret)
>                 return ret;
> 
> Then the i2c_set_clientdata() call can be removed, and the
> tla2528_remove() hook as well.
> A simple example can be seen in drivers/iio/adc/ti-ads8344.c
> 
> I'm not insisting on it. Since the driver is simple enough.

I'd prefer this, because otherwise someone will come along and change
it later creating more patches for me to review :)

Jonathan