Re: [RFC] Regulator state after regulator_get

[Sakari Ailus <sakari.ailus@nokia.com>]

Mark Brown wrote:
> On Thu, Apr 28, 2011 at 09:44:10AM +0000, kalle.jokiniemi@nokia.com wrote:
> 
>>  > Another alternative to the first option you proposed could be to add a
>>  > flags field to regulator_consumer_supply, and use a flag to recognise
>>  > regulators which need to be disabled during initialisation. The flag
>>  > could be set by using a new macro e.g. REGULATOR_SUPPLY_NASTY() when
>>  > defining the regulator.
> 
>> This sounds like a good option actually. Liam, Mark, any opinions?
> 
> I'm not sure what "supply_nasty" would mean?  This also doesn't seem
> like something that we can set up per supply - it's going to affect the
> whole regulator state, it's not something that only affects a single
> supply.

supply_nasty() would be used to define a regulator which is enabled by
the boot loader when it shouldn't be, which is the actual problem.

We have a regulator which is enabled by the boot loader. However, this
regulator shouldn't be on at boot since it's not needed by any devices
--- the drivers for those devices will use proper regulator framework
calls to use the regulator when it's needed. There's no chance to have
the boot loader fixed, as stated by Kalle.

How should this regulator be turned off in the boot by the kernel?

Regards,

-- 
Sakari Ailus
sakari.ailus@maxwell.research.nokia.com