Re: [NEW DRIVER V3 1/8] DA9058 MFD core driver

[Mark Brown <broonie@opensource.wolfsonmicro.com>]

On Wed, Aug 15, 2012 at 04:05:21PM +0100, Anthony Olech wrote:

>  
>  if HAS_IOMEM
> +
>  menu "Multifunction device drivers"

This random change is still present from the first version....

> +	/*
> +	 * the init_board_irq() call-back function should be defined in
> +	 * the machine driver initialization code and is used to set up
> +	 * the actual (probably GPIO) line as an interrupt line.
> +	 */
> +	if (pdata->init_board_irq) {
> +		ret = pdata->init_board_irq();
> +		if (ret)
> +			goto failed_to_setup_the_actual_i2c_hw_irq;
> +	}

You appear to have ignored my previous review comments about this...  it
still shouldn't be needed with modern kernels.

> +static bool da9058_register_volatile(struct device *dev, unsigned int reg)
> +{
> +	switch (reg) {
> +	case DA9058_ADCMAN_REG:
> +	case DA9058_ADCRESH_REG:
> +	case DA9058_ADCRESL_REG:
> +	case DA9058_ALARMD_REG:
> +	case DA9058_ALARMH_REG:
> +	case DA9058_ALARMMI_REG:
> +	case DA9058_ALARMMO_REG:
> +	case DA9058_ALARMS_REG:
> +	case DA9058_ALARMY_REG:

Are all the alarm registers really volatile?

> +	case DA9058_LDO9_REG:
> +	case DA9058_TOFFSET_REG:
> +	default:
> +		return false;

Just use the default.

> +static struct regulator_consumer_supply platform_vddarm_consumers[] = {
> +	{.supply = "vcc",}
> +};

No, this and all your other regulator configuration is board specific.