Re: [RFC V2] DA9210 driver files

[Mark Brown]

On Mon, Jul 01, 2013 at 03:10:28PM +0100, Steve Twiss wrote:
> From: Steve Twiss <stwiss.opensource@xxxxxxxxxxx>
> 
> This is the regulator driver for the Dialog DA9210 Multi-phase Buck.
> The patch is relative to linux-next next-20130701
> and this is RFC attempt number 2

Please follow the patch submission process in SubmittingPatches.  This
doesn't visually resemble most patch submissions...

>  >This looks like you should be using a regmap range.

> The use of regmap_range is not being considered because I am not intending
> to use PAGE_CON register page selection in any of the driver development.

Makes sense to map things in for diagnostics...

> +config REGULATOR_DA9210
> +	tristate "Dialog Semiconductor DA9210 Regulator"

Capitalisation is wrong Here.

> +static int da9210_set_current_limit(struct regulator_dev *rdev, int min_uA,
> +				    int max_uA)
> +{
> +	struct da9210 *chip = rdev_get_drvdata(rdev);
> +	unsigned int sel;
> +	int i;
> +
> +	/* search for closest to maximum */
> +	for (i = N_CURRENT_LIMITS-1; i >= 0; i--) {

Coding style.

> +	ret = regmap_read(chip->regmap, DA9210_REG_BUCK_ILIM, &data);
> +	if (ret < 0)
> +		return ret;
> +
> +	sel = (data & DA9210_BUCK_ILIM_MASK) >> DA9210_BUCK_ILIM_SHIFT;
> +
> +	return da9210_buck_limits[sel];

There's no unused values in the selector?

> +	chip->desc.id = 0;
> +	chip->desc.type = REGULATOR_VOLTAGE;
> +	chip->desc.n_voltages = ((DA9210_MAX_MV - DA9210_MIN_MV)
> +				/ DA9210_STEP_MV) + 1;
> +	chip->desc.ops = &da9210_buck_ops;
> +	chip->desc.owner = THIS_MODULE;
> +	chip->desc.name = "DA9210";
> +	chip->desc.enable_reg = DA9210_REG_BUCK_CONT;
> +	chip->desc.enable_mask = DA9210_BUCK_EN;
> +	chip->desc.vsel_reg = DA9210_REG_VBUCK_A;
> +	chip->desc.vsel_mask = DA9210_VBUCK_MASK;
> +	chip->desc.min_uV = (DA9210_MIN_MV * 1000);
> +	chip->desc.uV_step = (DA9210_STEP_MV * 1000);

Why is this not just global static data?  There's nothing variable
here...

> +	dev_info(&i2c->dev,
> +		"DA9210 device detected\n");
> +

Remove this - it's just noise, apart from anything else nothing here has
actually verified that the chip exists.

Attachment: signature.asc
Description: Digital signature