On Thu, Apr 28, 2022, Maciej S. Szmigiero wrote: > On 28.04.2022 11:37, Maxim Levitsky wrote: > > On Sat, 2022-04-23 at 02:14 +0000, Sean Christopherson wrote: > > > @@ -1618,7 +1644,7 @@ static int svm_set_nested_state(struct kvm_vcpu *vcpu, > > > nested_copy_vmcb_control_to_cache(svm, ctl); > > > svm_switch_vmcb(svm, &svm->nested.vmcb02); > > > - nested_vmcb02_prepare_control(svm, save->rip); > > > + nested_vmcb02_prepare_control(svm, svm->vmcb->save.rip); > > > > Is this change intentional? > > It looks to me the final code is correct since "svm->vmcb->save" > contains L2 register save, while "save" has L1 register save. > > It was the patch 1 from this series that was incorrect in > using "save->rip" here instead. Yeah, I botched the fixup.
