On Fri, Mar 06, 2015 at 03:35:36PM -0700, Eric Nelson wrote:
On 03/06/2015 01:16 PM, Mark Brown wrote:
Why not just fix the code, reinitializing the chip is usually pretty trivial?
Apparently not.
Well, I'm not sure if that's actually true or not - the current driver code is pretty bad and there's been a history of worrying submissions that people assure me are working well and address problems. All of this is indicating to me that the complexity here is largely due to code quality issues and that if we address those then things might get a lot easier.
Requiring a funny power on sequence shouldn't be hard and is needed for suspend and resume anyway, all having the regulator support does is mean we can potentially take advantage of that code path at runtime.