Fix memory ordering bug in page reclaim

As noticed by Nick Piggin, we need to make sure that we check the page
count before we check for PageDirty, since the dirty check is only valid
if the count implies that we're the only possible ones holding the page.

We always did do this, but the code needs a read-memory-barrier to make
sure that the orderign is also honored by the CPU.

(The writer side is ordered due to the atomic decrement and test on the
page count, see the discussion on linux-kernel)

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

diff --git a/mm/vmscan.c b/mm/vmscan.c
index 0ea71e8..64f9570 100644 (file)
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -511,10 +511,11 @@ static int shrink_list(struct list_head *page_list, struct scan_control *sc)
                 * PageDirty _after_ making sure that the page is freeable and
                 * not in use by anybody.       (pagecache + us == 2)
                 */
-               if (page_count(page) != 2 || PageDirty(page)) {
-                       write_unlock_irq(&mapping->tree_lock);
-                       goto keep_locked;
-               }
+               if (unlikely(page_count(page) != 2))
+                       goto cannot_free;
+               smp_rmb();
+               if (unlikely(PageDirty(page)))
+                       goto cannot_free;
 
 #ifdef CONFIG_SWAP
                if (PageSwapCache(page)) {
@@ -538,6 +539,10 @@ free_it:
                        __pagevec_release_nonlru(&freed_pvec);
                continue;
 
+cannot_free:
+               write_unlock_irq(&mapping->tree_lock);
+               goto keep_locked;
+
 activate_locked:
                SetPageActive(page);
                pgactivate++;