Re: [PATCH 1/1] regulator: return set_mode is same mode is requested

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

On Mon, May 17, 2010 at 07:39:53PM +0530, Sundar Iyer wrote:
> From: Sundar R Iyer <sundar.iyer@stericsson.com>

The commit message reads "regulator: return set_mode is same mode is
requested".  I'm having a hard time parsing what that actually means,
you probably need a "when" in there...

> +	/* return if the same mode is requested */
> +	regulator_curr_mode = rdev->desc->ops->get_mode(rdev);
> +	if (regulator_curr_mode == mode) {
> +		ret = 0;
> +		goto out;
> +	}
> +

This is going to oops if the regulator doesn't implement a get_mode()
operation.  

I'm also a little ambivalent on the benefit of it - if the goal is to
save I/O costs (you didn't say...) it's not clear to me that the effort
of checking the current mode is going to be a win in situations where
the mode is actually being changed a lot.

As I said in reply to your previous message the trend is away from
having any mode configration at all, with regulators being able to adapt
to their current load without any software assistance.