diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 053547fc6f89..eceb97734646 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4580,6 +4580,11 @@ static bool kvm_is_vm_type_supported(unsigned long type)
 	return type < 32 && (kvm_caps.supported_vm_types & BIT(type));
 }
 
+static inline u32 kvm_sync_valid_fields(struct kvm *kvm)
+{
+	return kvm && kvm->arch.has_protected_state ? 0 : KVM_SYNC_X86_VALID_FIELDS;
+}
+
 int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 {
 	int r = 0;
@@ -4688,7 +4693,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
 		break;
 #endif
 	case KVM_CAP_SYNC_REGS:
-		r = KVM_SYNC_X86_VALID_FIELDS;
+		r = kvm_sync_valid_fields(kvm);
 		break;
 	case KVM_CAP_ADJUST_CLOCK:
 		r = KVM_CLOCK_VALID_FLAGS;
@@ -11481,6 +11486,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
 {
 	struct kvm_queued_exception *ex = &vcpu->arch.exception;
 	struct kvm_run *kvm_run = vcpu->run;
+	u32 sync_valid_fields;
 	int r;
 
 	r = kvm_mmu_post_init_vm(vcpu->kvm);
@@ -11526,8 +11532,9 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
 		goto out;
 	}
 
-	if ((kvm_run->kvm_valid_regs & ~KVM_SYNC_X86_VALID_FIELDS) ||
-	    (kvm_run->kvm_dirty_regs & ~KVM_SYNC_X86_VALID_FIELDS)) {
+	sync_valid_fields = kvm_sync_valid_fields(vcpu->kvm);
+	if ((kvm_run->kvm_valid_regs & ~sync_valid_fields) ||
+	    (kvm_run->kvm_dirty_regs & ~sync_valid_fields)) {
 		r = -EINVAL;
 		goto out;
 	}
@@ -11585,7 +11592,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu)
 
 out:
 	kvm_put_guest_fpu(vcpu);
-	if (kvm_run->kvm_valid_regs)
+	if (kvm_run->kvm_valid_regs && likely(!vcpu->arch.guest_state_protected))
 		store_regs(vcpu);
 	post_kvm_run_save(vcpu);
 	kvm_vcpu_srcu_read_unlock(vcpu);