On Fri, Sep 23, 2016 at 01:27:19PM +1000, Benjamin Herrenschmidt wrote: > On Fri, 2016-09-23 at 11:37 +1000, David Gibson wrote: > > > > For KVM HV there's a bit of a nit: that would disallow migration > > between host cpus which aren't exactly the same model, but are close > > enough that migration will work in practice. > > In that case we should use the architected PVR
Uh... probably yes. Working out how to do that isn't totally trivial,
since for TCG mode the actualy PVR SPR that qemu tracks must contain
the real PVR value (to implement mfpvr), though the spapr code is also
aware of the architected one. We don't want to make things
gratuitously different for TCG. Plus we need to make sure it works
for TCG, PR and HV and also for no compat mode specified, compat mode
specified on the command line and compat mode negotiated by CAS.
I don't think there's any showstopper there, but it will require a bit
of thinking.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
signature.asc
Description: PGP signature
