On Friday 29 October 2004 21:03, Benjamin Herrenschmidt wrote: > On Fri, 2004-10-29 at 20:50 -0700, David Brownell wrote: > > (I'm not sure what Benjamin means by "Dynamic PM", > > but suspect it has to do with "Device PM".) > > Yes, what I call Dynamic PM is basically driver local PM policy, that > is a driver locally deciding to put it's hw in a low power state due to > overall idleness, but with automatic wakeup upon arrival of an upstream > request. So not exactly the same as http://dynamicpower.sourceforge.net either. They call that "Dynamic PM" too, defining system-wide "operating points" that combine Device PM and CPU PM (cpufreq) through various constraints. Controlled from userspace. - Dave