On Mon, 2022-12-12 at 13:13 +0000, Paul Durrant wrote:
> Ok, this tells me that we are intending to handle Hyper-V enlightenments 
> being simultaneously enabled... in which case that MSR above needs to 
> move, along with the cpuid leaves. It should be 0x40000200 in this case.

Ah, yes. That gets entertaining because hyperv_enabled() doesn't work
at the time kvm_arch_init() runs, because kvm_hyperv_expand_features()
hasn't run yet.

But kvm_arch_init() is idempotent so for now I'll just call it *again*
from kvm_arch_init_vcpu() to move the MSR in the hyperv case.

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to