KVM: x86: Manually flush collapsible SPTEs only when toggling flags

Zapping collapsible sptes, a.k.a. 4k sptes that can be promoted into a
large page, is only necessary when changing only the dirty logging flag
of a memory region.  If the memslot is also being moved, then all sptes
for the memslot are zapped when it is invalidated.  When a memslot is
being created, it is impossible for there to be existing dirty mappings,
e.g. KVM can have MMIO sptes, but not present, and thus dirty, sptes.

Note, the comment and logic are shamelessly borrowed from MIPS's version
of kvm_arch_commit_memory_region().

Fixes: 3ea3b7f ("kvm: mmu: lazy collapse small sptes into large sptes")
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

arch/x86/kvm/x86.c

@@ -9741,8 +9741,13 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 	 * Scan sptes if dirty logging has been stopped, dropping those
 	 * which can be collapsed into a single large-page spte.  Later
 	 * page faults will create the large-page sptes.
+	 *
+	 * There is no need to do this in any of the following cases:
+	 * CREATE:	No dirty mappings will already exist.
+	 * MOVE/DELETE:	The old mappings will already have been cleaned up by
+	 *		kvm_arch_flush_shadow_memslot()
 	 */
-	if ((change != KVM_MR_DELETE) &&
+	if (change == KVM_MR_FLAGS_ONLY &&
 		(old->flags & KVM_MEM_LOG_DIRTY_PAGES) &&
 		!(new->flags & KVM_MEM_LOG_DIRTY_PAGES))
 		kvm_mmu_zap_collapsible_sptes(kvm, new);