On Mon, 2004-11-01 at 23:45 +0100, Pavel Machek wrote: > Hi! > > > Anyway, I think what we have defined so far seem to be good enough to > > start the actual move. The only thing I'd like you to judge is wether we > > keep the original PM message semantics we discussed, or we move the > > freeze semantic to a separate flag as discussed in another message so > > that we have some kind of generic way of triggering drivers local PM... > > Actually it seems like an implementation detail to me... but I'd vote > for separate flags. Distinction between "IDLE" and "SUSPEND" seems > quite unclear to me. > > OTOH... there's preparation for shutdown, which does not fit here too > much, and preparation for apm sleep, which does not fit, either: > > For shutdown, power state is not interesting, and you do not even need > to freeze, but you'd better spin down the disks, or you loose your > data. If you don't freeze, a stal request may spin back up. Also, shutdown is used for kexec as well, which absolutely need DMA off, that is freeze semantics. I suggest we define a shutdown message too. By default, busses would call shutdown if it exist, if not, pass suspend with the shutdown message. Looks ok ? > For apm suspend... noone knows whats really required. Theoretically, > neither freeze nor powersaving is needed. Freeze would be a good thing tho. There are interesting "races" with APM suspend that would be fixed... Also, saving/restoring device states since APM tend to fail restore some stuffs here or there. Ben.
