Sean Christopherson <seanjc@xxxxxxxxxx> writes:
> On Tue, May 11, 2021, Vitaly Kuznetsov wrote:
>> When nested state migration happens during L1's execution, it
>> is incorrect to modify eVMCS as it is L1 who 'owns' it at the moment.
>> At lease genuine Hyper-v seems to not be very happy when 'clean fields'
>> data changes underneath it.
>>
>> 'Clean fields' data is used in KVM twice: by copy_enlightened_to_vmcs12()
>> and prepare_vmcs02_rare() so we can reset it from prepare_vmcs02() instead.
>>
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@xxxxxxxxxx>
>> ---
>> arch/x86/kvm/vmx/nested.c | 9 ++++++---
>> 1 file changed, 6 insertions(+), 3 deletions(-)
>>
>> diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
>> index 3257a2291693..1661e2e19560 100644
>> --- a/arch/x86/kvm/vmx/nested.c
>> +++ b/arch/x86/kvm/vmx/nested.c
>> @@ -2090,9 +2090,6 @@ void nested_sync_vmcs12_to_shadow(struct kvm_vcpu *vcpu)
>>
>> if (vmx->nested.hv_evmcs) {
>> copy_vmcs12_to_enlightened(vmx);
>> - /* All fields are clean */
>> - vmx->nested.hv_evmcs->hv_clean_fields |=
>> - HV_VMX_ENLIGHTENED_CLEAN_FIELD_ALL;
>> } else {
>> copy_vmcs12_to_shadow(vmx);
>> }
>
> Looks like this makes both paths single lines, i.e. the curly braces can be
> dropped.
>
Yea, I've noticed that but I wasn't exactly sure the benefit of removing
one line here overwhelms the pain of needing an extra step while doing
code archeology with 'git blame' :-) Will drop on v2.
>> @@ -2636,6 +2633,12 @@ static int prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12,
>>
>> kvm_rsp_write(vcpu, vmcs12->guest_rsp);
>> kvm_rip_write(vcpu, vmcs12->guest_rip);
>> +
>> + /* Mark all fields as clean so L1 hypervisor can set what's dirty */
>> + if (hv_evmcs)
>> + vmx->nested.hv_evmcs->hv_clean_fields |=
>> + HV_VMX_ENLIGHTENED_CLEAN_FIELD_ALL;
>> +
>> return 0;
>> }
>>
>> --
>> 2.30.2
>>
>
--
Vitaly
