Doug / smarc-fsl-linux-kernel

Download as
Browse Code ยป

Commit 21ee9f398be209ccbb62929d35961ca1ed48eec3

Authored by Minchan Kim
Committed by Linus Torvalds
1 parent 2f1da64215
Exists in master and in 6 other branches imx_3.10.17_1.0.1_ga, imx_3.10.53_1.1.0_ga, imx_3.14.28_1.0.0_ga, smarc-imx_3.10.53_1.1.0_ga, smarc-imx_3.14.28_1.0.0_ga, smarc-l5.0.0_1.0.0-ga

vmscan: add barrier to prevent evictable page in unevictable list 

When a race between putback_lru_page() and shmem_lock with lock=0 happens,
progrom execution order is as follows, but clear_bit in processor #1 could
be reordered right before spin_unlock of processor #1.  Then, the page
would be stranded on the unevictable list.

spin_lock
SetPageLRU
spin_unlock
                                clear_bit(AS_UNEVICTABLE)
                                spin_lock
                                if PageLRU()
                                        if !test_bit(AS_UNEVICTABLE)
                                        	move evictable list
smp_mb
if !test_bit(AS_UNEVICTABLE)
        move evictable list
                                spin_unlock

But, pagevec_lookup() in scan_mapping_unevictable_pages() has
rcu_read_[un]lock() so it could protect reordering before reaching
test_bit(AS_UNEVICTABLE) on processor #1 so this problem never happens.
But it's a unexpected side effect and we should solve this problem
properly.

This patch adds a barrier after mapping_clear_unevictable.

I didn't meet this problem but just found during review.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Rik van Riel <riel@redhat.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Showing 2 changed files with 12 additions and 5 deletions Side-by-side Diff

... ... @@ -1068,6 +1068,12 @@
1068 1068 user_shm_unlock(inode->i_size, user);
1069 1069 info->flags &= ~VM_LOCKED;
1070 1070 mapping_clear_unevictable(file->f_mapping);
  1071 + /*
  1072 + * Ensure that a racing putback_lru_page() can see
  1073 + * the pages of this mapping are evictable when we
  1074 + * skip them due to !PageLRU during the scan.
  1075 + */
  1076 + smp_mb__after_clear_bit();
1071 1077 scan_mapping_unevictable_pages(file->f_mapping);
1072 1078 }
1073 1079 retval = 0;
... ... @@ -633,13 +633,14 @@
633 633 lru = LRU_UNEVICTABLE;
634 634 add_page_to_unevictable_list(page);
635 635 /*
636   - * When racing with an mlock clearing (page is
637   - * unlocked), make sure that if the other thread does
638   - * not observe our setting of PG_lru and fails
639   - * isolation, we see PG_mlocked cleared below and move
  636 + * When racing with an mlock or AS_UNEVICTABLE clearing
  637 + * (page is unlocked) make sure that if the other thread
  638 + * does not observe our setting of PG_lru and fails
  639 + * isolation/check_move_unevictable_page,
  640 + * we see PG_mlocked/AS_UNEVICTABLE cleared below and move
640 641 * the page back to the evictable list.
641 642 *
642   - * The other side is TestClearPageMlocked().
  643 + * The other side is TestClearPageMlocked() or shmem_lock().
643 644 */
644 645 smp_mb();
645 646 }