On Thu, Nov 15, 2012 at 01:46:52PM +0100, ciminaghi@gnudd.com wrote:

> I haven't seen any reply up to now, so I'm just resending the original patch
> (see https://lkml.org/lkml/2012/10/25/224).
> I just applied it to today's next and did a quick re-test.

The fact that you don't send a real name in your patch submission mails
might be an issue here - I suspect you've fallen foul of spam filtering,
either automatic or human, there's no sign of this in my pending queue.

> +static inline bool _reg_in_ranges(unsigned int reg,
> +				  const struct regmap_range *ranges,
> +				  unsigned int nranges)
> +{
> +	const struct regmap_range *r;
> +	int i;
> +
> +	for (i = 0, r = ranges; i < nranges; i++, r++)
> +		if (_reg_in_range(reg, r))
> +			return true;
> +	return false;
> +}

It would be better to make this an externally visible function (perhaps
just move it into the header with a regmap on the front of the name) so
that drivers can use this as part of the implementation of more complex
checks.  Otherwise this looks fine.