On Thu, Dec 13, 2012 at 01:09:11AM +0000, Zhang, Yang Z wrote:
> >> +static void ioapic_update_eoi_exitmap_one(struct kvm_ioapic *ioapic, int pin)
> >> +{
> >> + union kvm_ioapic_redirect_entry *e;
> >> +
> >> + e = &ioapic->redirtbl[pin];
> >> +
> >> + if ((e->fields.trig_mode == IOAPIC_LEVEL_TRIG ||
> >> + kvm_irq_has_notifier(ioapic->kvm, KVM_IRQCHIP_IOAPIC, pin))) {
> >> + struct kvm_lapic_irq irqe;
> >> +
> >> + irqe.dest_id = e->fields.dest_id;
> >> + irqe.vector = e->fields.vector;
> >> + irqe.dest_mode = e->fields.dest_mode;
> >> + irqe.delivery_mode = e->fields.delivery_mode << 8;
> >> + kvm_update_eoi_exitmap(ioapic->kvm, &irqe);
> >> + }
> >> +}
> >> +
> > There is a bugs in exitbitmap recalculation that I've missed.
> > We need to zero all the exit bitmaps for all vcpus before we are
> > starting to recalculate them to get rid of stale bits.
> Right. I forget it too. So if consider to clear it, we should call ioapic_update_eoi_exitmap instead ioapic_update_eoi_exitmap_one when programming one program ioapic entry.
>
Yes.
--
Gleb.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
