On Mon, 2024-02-26 at 00:27 -0800, isaku.yamahata@xxxxxxxxx wrote:
>
> +static void vt_update_cpu_dirty_logging(struct kvm_vcpu *vcpu)
> +{
> + if (KVM_BUG_ON(is_td_vcpu(vcpu), vcpu->kvm))
> + return;
> +
> + vmx_update_cpu_dirty_logging(vcpu);
> +}
Discussed this first part offline, but logging it here. Since
guest_memfd cannot have dirty logging, this is essentially bugging the
VM if somehow they manage anyway. But it should be blocked via the code
in check_memory_region_flags().
On the subject of warnings and KVM_BUG_ON(), my feeling so far is that
this series is quite aggressive about these. Is it due the complexity
of the series? I think maybe we can remove some of the simple ones, but
not sure if there was already some discussion on what level is
appropriate.
