> That being said, I'm not categorically against these patches, but I
> share Marc's view that we've already seen that non-vgic support had been
> broken for multiple versions without anyone complaining,
Oh, did this ever work? Work as in "can actually run a virtual
machine", not as in "kvm doesn't throw an error on initialization".
> and without
> automated testing or substantial interest in the work, the patches
> really are likely to bit-rot.
Well, as far I know the rpi3 is the *only* aarch64 hardware which is
easily available and can (with these patches) run kvm.
More powerful stuff with more ram and sata storage and gbit network
(which I'd love to have to play with arm virt on real hardware) is
announced to be available really soon now since ... half a year at
kvmarm mailing list