KVM: x86: Check the right feature bit for MSR_KVM_ASYNC_PF_ACK access

[ Upstream commit 0a31df6 ]

MSR_KVM_ASYNC_PF_ACK MSR is part of interrupt based asynchronous page fault
interface and not the original (deprecated) KVM_FEATURE_ASYNC_PF. This is
stated in Documentation/virt/kvm/msr.rst.

Fixes: 66570e9 ("kvm: x86: only provide PV features if enabled in guest's CPUID")
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Oliver Upton <oupton@google.com>
Message-Id: <20210722123018.260035-1-vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

arch/x86/kvm/x86.c

@@ -3205,7 +3205,7 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 			return 1;
 		break;
 	case MSR_KVM_ASYNC_PF_ACK:
-		if (!guest_pv_has(vcpu, KVM_FEATURE_ASYNC_PF))
+		if (!guest_pv_has(vcpu, KVM_FEATURE_ASYNC_PF_INT))
 			return 1;
 		if (data & 0x1) {
 			vcpu->arch.apf.pageready_pending = false;
@@ -3534,7 +3534,7 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 		msr_info->data = vcpu->arch.apf.msr_int_val;
 		break;
 	case MSR_KVM_ASYNC_PF_ACK:
-		if (!guest_pv_has(vcpu, KVM_FEATURE_ASYNC_PF))
+		if (!guest_pv_has(vcpu, KVM_FEATURE_ASYNC_PF_INT))
 			return 1;
 
 		msr_info->data = 0;