On Wed, Jul 24, 2013 at 03:18:01PM +0200, Paolo Bonzini wrote: > Il 24/07/2013 15:14, Michael S. Tsirkin ha scritto: > > On Wed, Jul 24, 2013 at 02:59:18PM +0200, Paolo Bonzini wrote: > >> Il 24/07/2013 11:51, Michael S. Tsirkin ha scritto: > >>>> Current seabios uses a slightly different approach: the 64bit hole is > >>>> present only in case it is actually used to map bars there, and seabios > >>>> tries to fit everything into the 32bit hole first. > >>> > >>> Yes. But this doesn't work with device hotplug. > >> > >> Are you preserving that behavior for older machine types? > > > > Older machine types (1.5 and back) don't report PCI hole size to guests > > so there's nothing to preserve. > > So pc-1.5 and older always use the SeaBIOS ACPI tables, even on new > QEMU? (I guess I should review your ACPI table patches). > > Paolo
Yes. > > This means there's not much I can do for pc-1.5: as before this > > patch, hotplug of devices with large BARs won't work, > > and some windows guests will crash if a device with a huge > > BAR is added statically. > > > >> (Also, what happens if you do -global i440FX-pcihost.pci_hole64_size=0? > >> Does that return a _CRS with only the 32-bit hole)? > >> > >> Paolo > > > > Yes. > >
