On Sat, Mar 24, 2012 at 03:52:38PM -0700, Guenter Roeck wrote: > On Sat, Mar 24, 2012 at 03:45:43PM -0400, Roland Stigge wrote: > > >> +#ifdef CONFIG_PM > > > Looking into other drivers, I think that should be CONFIG_PM_SLEEP > > Which other drivers did you mean? > > What about those drivers that all handle it via CONFIG_PM: > > abituguru.c > > abituguru3.c > > exynos4_tmu.c > > gpio-fan.c > > jc42.c > > lm75.c > > tmp102.c > Those drivers don't use SET_SYSTEM_SLEEP_PM_OPS. I checked several drivers > which do use SET_SYSTEM_SLEEP_PM_OPS, and they all use CONFIG_PM_SLEEP. It's nothing to do with SET_SYSTEM_SLEEP_PM_OPS - it's more about drivers that are following current best practice (which includes using the SET_ macros). A modern driver should be using CONFIG_PM_SLEEP for system sleep operations and CONFIG_PM_RUNTIME for runtime PM operations since they can be enabled independently. CONFIG_PM should only control having dev_pm_ops at all (if you want to bother with the ifdefery).
Attachment:
signature.asc
Description: Digital signature
_______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
