Paul Gilmartin wrote: > I reported a similar problem a few months ago with either > OS X or XP host and Solaris guest. I observed it when I > put the host (laptop) to sleep; when I reawaken the host, > the guest clock is wrong but the host clock (of course) is > correct. > > Suggestions on this list were: > > o Install Guest Additions, which steers the clock better. > > o Run NTP on the guest. > > Neither of these is particularly satisfactory; both seem to > steer the clock by only a small amount, 1% or so, so if my > host sleeps for several hours, it will take the guest a > few weeks to catch up. > > Then a Solaris developer told me the problem can't be solved > because the host passes no indication to the guest that the > host has slept; WAD. > > But I've subsequently observed that I can create the same > (mis-)behavior by Pausing (Host)-P the guest, then resuming. > The guest clock is then slow by the duration of the Pause. > In this case, VBox surely knows about the Pause; VBox itself > did it. I consider this a defect requiring repair: on Resume > VBox should reload the emulated BIOS clock (or whatever) from > the host BIOS clock; perhaps signal the guest that it has > slept, as on a laptop with the hinge closed. > For OSs that use API or ACPI, it should be possible to generate an event when you pause/resume the machine. As for knowing when the host pauses/resumes, could you add a section to the vbox driver that sends vbox a signal on resume? It should not be a problem on Linux, and I would asume that other OSs have something to re-initialize hardware that needs it when the machine resumes/wakes up.
Mikkel -- Do not meddle in the affairs of dragons, for thou art crunchy and taste good with Ketchup!
signature.asc
Description: OpenPGP digital signature
_______________________________________________ vbox-users mailing list vbox-users@virtualbox.org http://vbox.innotek.de/mailman/listinfo/vbox-users
