On 19/04/16 11:26, Andre Przywara wrote: > Hi Vladimir, > > can you try the attached patch on top of the series to see if this > fixes the GICv3 emulation issues you see? >
With attached patch I don't see issue any more - all 255 vcpus are able to boot and send/receive self SGI. Thanks Vladimir > I think these lines got lost during the MMIO framework rebasing. > > Cheers, > Andre. > > --- > virt/kvm/arm/vgic/vgic_mmio.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/virt/kvm/arm/vgic/vgic_mmio.c b/virt/kvm/arm/vgic/vgic_mmio.c > index ec96188..92f39ba 100644 > --- a/virt/kvm/arm/vgic/vgic_mmio.c > +++ b/virt/kvm/arm/vgic/vgic_mmio.c > @@ -1228,6 +1228,7 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t > redist_base_address) > kvm_for_each_vcpu(c, vcpu, kvm) { > kvm_iodevice_init(®ion->dev, &kvm_io_v3redist_ops); > region->base_addr = redist_base_address; > + region->redist_vcpu = vcpu; > > mutex_lock(&kvm->slots_lock); > ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, > @@ -1241,6 +1242,7 @@ int vgic_register_redist_iodevs(struct kvm *kvm, gpa_t > redist_base_address) > region++; > kvm_iodevice_init(®ion->dev, &kvm_io_v3redist_private_ops); > region->base_addr = redist_base_address + SZ_64K; > + region->redist_vcpu = vcpu; > > mutex_lock(&kvm->slots_lock); > ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, > _______________________________________________ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm