On Sat, 11 Sep 2010, Rafael J. Wysocki wrote: > > > > First check for things that make the function impossible: > > > > power.runtime_error set, usage_count or child_count set > > > > wrong, or disable_depth > 0. If these checks fail, return > > > > immediately. > > > > > > OK > > > > > > > Then check the current status and the pending requests to > > > > see if they rule out the current function: > > > > > > > > idle is allowed only in RPM_ACTIVE and when there > > > > are no pending requests (but the suspend timer can > > > > be running); > > > > > > > > suspend is not allowed if the status is RPM_SUSPENDED > > > > or RPM_RESUMING, or if there is a pending resume > > > > (either a request or a deferred resume); > > > > > > > > resume is not allowed if the status is RPM_ACTIVE. > > > > > > OK, but that need not apply to the _ldle() variants. > > > > I don't understand. Are you referring to the "idle is allowed only..." > > paragraph above? What's the connection with resume? > > Sorry, I thought *_request_*() and wrote _idle(). Must be too tired. :-) I'm still not sure what you mean. I meant those conditions to apply to both the synchronous and asynchronous calls (except of course that we will allow async suspend if the status is RPM_RESUMING). Did you have something else is mind? Alan Stern _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linux-foundation.org/mailman/listinfo/linux-pm
