mm,ksm: swapoff might need to copy

Before establishing that KSM page migration was the cause of my
WARN_ON_ONCE(page_mapped(page))s, I suspected that they came from the
lack of a ksm_might_need_to_copy() in swapoff's unuse_pte() - which in
many respects is equivalent to faulting in a page.

In fact I've never caught that as the cause: but in theory it does at
least need the KSM_RUN_UNMERGE check in ksm_might_need_to_copy(), to
avoid bringing a KSM page back in when it's not supposed to be.

I intended to copy how it's done in do_swap_page(), but have a strong
aversion to how "swapcache" ends up being used there: rework it with
"page != swapcache".

Signed-off-by: Hugh Dickins <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Petr Holasek <[email protected]>
Cc: Andrea Arcangeli <[email protected]>
Cc: Izik Eidus <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

mm/swapfile.c

@@ -874,11 +874,17 @@ unsigned int count_swap_pages(int type, int free)
 static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 		unsigned long addr, swp_entry_t entry, struct page *page)
 {
+	struct page *swapcache;
 	struct mem_cgroup *memcg;
 	spinlock_t *ptl;
 	pte_t *pte;
 	int ret = 1;
 
+	swapcache = page;
+	page = ksm_might_need_to_copy(page, vma, addr);
+	if (unlikely(!page))
+		return -ENOMEM;
+
 	if (mem_cgroup_try_charge_swapin(vma->vm_mm, page,
 					 GFP_KERNEL, &memcg)) {
 		ret = -ENOMEM;
@@ -897,7 +903,10 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 	get_page(page);
 	set_pte_at(vma->vm_mm, addr, pte,
 		   pte_mkold(mk_pte(page, vma->vm_page_prot)));
-	page_add_anon_rmap(page, vma, addr);
+	if (page == swapcache)
+		page_add_anon_rmap(page, vma, addr);
+	else /* ksm created a completely new copy */
+		page_add_new_anon_rmap(page, vma, addr);
 	mem_cgroup_commit_charge_swapin(page, memcg);
 	swap_free(entry);
 	/*
@@ -908,6 +917,10 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 out:
 	pte_unmap_unlock(pte, ptl);
 out_nolock:
+	if (page != swapcache) {
+		unlock_page(page);
+		put_page(page);
+	}
 	return ret;
 }