On Tue, 2012-09-18 at 17:09 +0100, Mark Brown wrote:

On Tue, Sep 18, 2012 at 04:44:16PM +0100, Pawel Moll wrote:

quoted

Well, that's what it really is. The config API sends a request "set xyz
uV" to the microcontrollers. And the micro can (at least in theory) get
you any voltage within the min/max limits by whatever means it has (at
least some of the daugtherboards use micro's DAC to adjust reference
voltage for a DC/DC converter with a feedback loop using ADC).

And the microcontroller is incapable of telling us what it supports, 

No, there is no such . Device Tree defines per-device min/max
constraints and any value between these values should be supported.

or even what's physically present?

You mean read back what voltage is set now? By all means - that's what
it is doing for "read only" devices.

quoted

But fair enough, I should have done better work in describing this.

So this is going to break interoperation with a bunch of consumer
drivers that rely on being able to tell what voltages are supported.
The key thing for them would be that regulator_is_supported_voltage()
works, currently it relies on list_voltage() as that's the only way to
do that right now.

Ok, I guess I should use regulator_list_voltage_linear() and
regulator_map_voltage_linear() then? I'll just have to carefully think
what step to choose.

quoted

v3 to follow.

v3?  I didn't see V2.  Please also CC posts to the relevant mailing
lists for the subsystem (lkml in this case).

Sure, will do.

Pawel