[PATCH v1 01/24] psci: Accessor for configured PSCI version

[David Brazdil]

> > +
> >  struct psci_operations {
> >  	u32 (*get_version)(void);
> >  	int (*cpu_suspend)(u32 state, unsigned long entry_point);
> 
> I still maintain that populating .get_version in all cases instead of
> duplicating an existing functionality is a better outcome. PSCI not
> supported would be implied by .get_version being NULL.
> 
> What breaks?

Must have missed that suggestion before. Good idea, I'll do that.