KVM: X86: Don't reset mmu context when X86_CR4_PCIDE 1->0

X86_CR4_PCIDE doesn't participate in kvm_mmu_role, so the mmu context
doesn't need to be reset.  It is only required to flush all the guest
tlb.

Signed-off-by: Lai Jiangshan <laijs@linux.alibaba.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20210919024246.89230-2-jiangshanlai@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

arch/x86/kvm/x86.c

@@ -1022,9 +1022,10 @@ EXPORT_SYMBOL_GPL(kvm_is_valid_cr4);
 
 void kvm_post_set_cr4(struct kvm_vcpu *vcpu, unsigned long old_cr4, unsigned long cr4)
 {
-	if (((cr4 ^ old_cr4) & KVM_MMU_CR4_ROLE_BITS) ||
-	    (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE)))
+	if ((cr4 ^ old_cr4) & KVM_MMU_CR4_ROLE_BITS)
 		kvm_mmu_reset_context(vcpu);
+	else if (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE))
+		kvm_make_request(KVM_REQ_TLB_FLUSH_GUEST, vcpu);
 }
 EXPORT_SYMBOL_GPL(kvm_post_set_cr4);