fs/proc/task_mmu: reduce scope of lazy mmu region

Update the way arch_[enter|leave]_lazy_mmu_mode() is called in
pagemap_scan_pmd_entry() to follow the normal pattern of holding the ptl
for user space mappings. As a result the scope is reduced to only the pte
table, but that's where most of the performance win is.

While I believe there wasn't technically a bug here, the original scope
made it easier to accidentally nest or, worse, accidentally call something
like kmap() which would expect an immediate mode pte modification but it
would end up deferred.

Link: https://lkml.kernel.org/r/[e-mail]
Signed-off-by: Ryan Roberts <[e-mail]>
Acked-by: David Hildenbrand <[e-mail]>
Acked-by: Juergen Gross <[e-mail]>
Cc: Andreas Larsson <[e-mail]>
Cc: Borislav Betkov <[e-mail]>
Cc: Boris Ostrovsky <[e-mail]>
Cc: Catalin Marinas <[e-mail]>
Cc: Dave Hansen <[e-mail]>
Cc: David S. Miller <[e-mail]>
Cc: "H. Peter Anvin" <[e-mail]>
Cc: Ingo Molnar <[e-mail]>
Cc: Juegren Gross <[e-mail]>
Cc: Matthew Wilcow (Oracle) <[e-mail]>
Cc: Thomas Gleinxer <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>

fs/proc/task_mmu.c

···
 	spinlock_t *ptl;
 	int ret;
 
-	arch_enter_lazy_mmu_mode();
-
 	ret = pagemap_scan_thp_entry(pmd, start, end, walk);
-	if (ret != -ENOENT) {
-		arch_leave_lazy_mmu_mode();
+	if (ret != -ENOENT)
 		return ret;
-	}
 
 	ret = 0;
 	start_pte = pte = pte_offset_map_lock(vma->vm_mm, pmd, start, &ptl);
 	if (!pte) {
-		arch_leave_lazy_mmu_mode();
 		walk->action = ACTION_AGAIN;
 		return 0;
 	}
+
+	arch_enter_lazy_mmu_mode();
 
 	if ((p->arg.flags & PM_SCAN_WP_MATCHING) && !p->vec_out) {
 		/* Fast path for performing exclusive WP */
···
 	if (flush_end)
 		flush_tlb_range(vma, start, addr);
 
-	pte_unmap_unlock(start_pte, ptl);
 	arch_leave_lazy_mmu_mode();
+	pte_unmap_unlock(start_pte, ptl);
 
 	cond_resched();
 	return ret;