On Sat, 2004-11-13 at 20:53 +0000, Linux Kernel Mailing List wrote: > ChangeSet 1.2115, 2004/11/13 12:53:51-08:00, torvalds@xxxxxxxxxxxxxxx > > ppc: fix up pmac IDE driver for driver core changes > > device power state is in "dev.power.power_state" now, rather than > in "dev.power_state". > Hrm... Missed that core change, where does it come from ? doesn't quite go in the direction we have been discussing on linux-pm lately, which is rather to remove this power_state field entirely since it's mostly meaningless at this point (or rather, it's semantics are confuse and the userland interface to it is means pretty much nothing). While having a power_state field here for internal use of the driver may be "useful" (avoiding the need for drivers to maintain something equivalent locally), it's pretty much impossible to have a generic abstract power_state that has globally defined semantics, at least not with our current scheme. Ben.
