On Fri, Apr 11, 2014 at 09:40:19PM -0400, Paolo Bonzini wrote:
> Il 07/04/2014 11:19, Michael S. Tsirkin ha scritto:
> >This means we can't cleanly implement an option for guest to
> >disable ACPI and switch to native if supported,
> >like the ACPI spec allows.
> >
> >If we change hotplug code to walk the tree top down
> >and invoke all callbacks, then it will be fixed
> >in a cleaner way: bridges would just do:
> >
> > if (dev->bus != self) {
> > set_error
> > return;
> > }
> >
> >and suddently pci host can trap callbacks and redirect
> >to acpi if it wants to.
>
> I think this should be handled by making the q35 PCI host bridge
> implement HotplugHandler itself, possibly overriding the parent's
> implementation.
>
> Paolo
Yes but it needs to be dynamic, i.e. q35 should be able to say
at runtime "I can't handle this hotplug, pass it up to parent"
rather than just checking callback is NULL.
