[RFC 4/6] mm: mark dirty bit on unuse_pte

[Minchan Kim]

Basically, MADV_FREE relys on the dirty bit in page table entry
to decide whether VM allows to discard the page or not.
IOW, if page table entry includes marked dirty bit, VM shouldn't
discard the page.

However, if swapoff happens, page table entry point out the page
doesn't have marked dirty bit so MADV_FREE might discard the page
wrongly.

To fix the problem, this patch marks page table entry of page
as dirty when swapoff hanppens VM shouldn't discard the page
suddenly under us.

With MADV_FREE point of view, marking dirty unconditionally is
no problem because we dropped swapped page in MADV_FREE sycall
context(ie, Look at madvise_free_pte_range) so every swapping-in
pages are no MADV_FREE hinted pages.

Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
---
 mm/swapfile.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/mm/swapfile.c b/mm/swapfile.c
index a7e72103f23b..cc8b79ab2190 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1118,8 +1118,12 @@ static int unuse_pte(struct vm_area_struct *vma, pmd_t *pmd,
 	dec_mm_counter(vma->vm_mm, MM_SWAPENTS);
 	inc_mm_counter(vma->vm_mm, MM_ANONPAGES);
 	get_page(page);
+	/*
+	 * For preventing sudden freeing by MADV_FREE, pte must have a
+	 * dirty flag.
+	 */
 	set_pte_at(vma->vm_mm, addr, pte,
-		   pte_mkold(mk_pte(page, vma->vm_page_prot)));
+		   pte_mkdirty(pte_mkold(mk_pte(page, vma->vm_page_prot))));
 	if (page == swapcache) {
 		page_add_anon_rmap(page, vma, addr);
 		mem_cgroup_commit_charge(page, memcg, true);
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/