On Tue, Jul 05, 2016 at 01:41:50PM +0100, Dr. David Alan Gilbert wrote: > * Michael S. Tsirkin (m...@redhat.com) wrote: > > On Tue, Jul 05, 2016 at 12:59:42PM +0200, Paolo Bonzini wrote: > > > > > > > > > On 05/07/2016 12:06, Michael S. Tsirkin wrote: > > > > > -m 2G,slots=16,maxmem=2T > > > > > > > > > > On a host with a 39bit physaddress limit do you error > > > > > on that or not? I think oVirt is currently doing something > > > > > similar to that, but I'm trying to get confirmation. > > > > > > > > That would only be a problem since pci is allocated above > > > > maxmem so 64 bit pci addresses aren't accessible. > > > > With my proposal we can actually force firmware to avoid > > > > using 64 bit memory for that config. > > > > Will work better than today. > > > > > > So you would remove completely the 64-bit _CRS in this case? > > > > Yes. > > > > > 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? > > > > > > Thanks, > > > > > > Paolo > > > > 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. > > What stops the guest OS mapping PCI stuff high up - however much you change > the firmware? > > Dave
It only maps what is described by _CRS. Mapping addresses outside CRS will crash baremetal too, with unpredictable results. > > > > -- > > MST > -- > Dr. David Alan Gilbert / dgilb...@redhat.com / Manchester, UK
