On 05/07/2016 13:09, Michael S. Tsirkin wrote: > > How do you handle migration in the above scenario from say 46bit host to > > 39bit host, where the firmware has mapped (while running on the source) > > a 64-bit BAR above the destination's maximum physical address? > > Again management would specify how much 64 bit pci space firmware should use. > If more is specified than host can support we can error out.
Ok, so the destination would fail to start. It sounds good, and it may provide a good reason not to enable autodetection of phys-addr-bits for new machine types. We still want David's patches for -cpu host, and for improved backwards compatibility now that KVM_SET_MSR error detection is strict. Thanks, Paolo