On Tue, Jun 28, 2011 at 06:49, Paolo Bonzini <pbonz...@redhat.com> wrote: > On 06/27/2011 08:07 PM, Peter Maydell wrote: >> >> > + int ioport_base; >> > + uint32_t ioport_reg[2]; >> >> I think ioport_reg[] needs to go in the VMStateDescription as well. >> I don't know enough about the PCI subsystem to know whether that's >> also true of ioport_base or whether the the map function is called >> again on a vmload. > > ioport_reg[1] is never written nor read, so this can be changed to a single > uint32_t ioport_addr. Since it will be almost always 0 (either because it > was never written, or because E1000_CTRL == 0), a subsection is useful to > have.
Indeed, I will remove ioport_reg[1] and ioport_addr. Thanks, -- Anthony PERARD