Pages allocated from MIGRATE_RESERVE migratetype pageblocks
are not freed back to MIGRATE_RESERVE migratetype free
lists in free_pcppages_bulk() if we got to that function
through drain_[zone_]pages() or __zone_pcp_update().
The freeing through free_hot_cold_page() is okay because
freepage migratetype is set to pageblock migratetype before
calling free_pcppages_bulk(). If pages of MIGRATE_RESERVE
migratetype end up on the free lists of other migratetype
whole Reserved pageblock may be later changed to the other
migratetype in __rmqueue_fallback() and it will be never
changed back to be a Reserved pageblock. Fix the issue by
preserving freepage migratetype as a pageblock migratetype
(instead of overriding it to the requested migratetype)
for MIGRATE_RESERVE migratetype pages in rmqueue_bulk().
The problem was introduced in v2.6.31 by commit ed0ae21
("page allocator: do not call get_pageblock_migratetype()
more than necessary").
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnier...@samsung.com>
Reported-by: Yong-Taek Lee <ytk....@samsung.com>
Cc: Marek Szyprowski <m.szyprow...@samsung.com>
Cc: Mel Gorman <mgor...@suse.de>
Cc: Hugh Dickins <hu...@google.com>
---
include/linux/mmzone.h | 5 +++++
mm/page_alloc.c | 10 +++++++---
2 files changed, 12 insertions(+), 3 deletions(-)
Index: b/include/linux/mmzone.h
===================================================================
--- a/include/linux/mmzone.h 2014-02-14 18:59:08.177837747 +0100
+++ b/include/linux/mmzone.h 2014-02-14 18:59:09.077837731 +0100
@@ -63,6 +63,11 @@ enum {
MIGRATE_TYPES
};
+static inline bool is_migrate_reserve(int migratetype)
+{
+ return unlikely(migratetype == MIGRATE_RESERVE);
+}
+
#ifdef CONFIG_CMA
# define is_migrate_cma(migratetype) unlikely((migratetype) == MIGRATE_CMA)
#else
Index: b/mm/page_alloc.c
===================================================================
--- a/mm/page_alloc.c 2014-02-14 18:59:08.185837746 +0100
+++ b/mm/page_alloc.c 2014-02-14 18:59:09.077837731 +0100
@@ -1174,7 +1174,7 @@ static int rmqueue_bulk(struct zone *zon
unsigned long count, struct list_head *list,
int migratetype, int cold)
{
- int mt = migratetype, i;
+ int mt, i;
spin_lock(&zone->lock);
for (i = 0; i < count; ++i) {
@@ -1195,9 +1195,13 @@ static int rmqueue_bulk(struct zone *zon
list_add(&page->lru, list);
else
list_add_tail(&page->lru, list);
+ mt = get_pageblock_migratetype(page);
if (IS_ENABLED(CONFIG_CMA)) {
- mt = get_pageblock_migratetype(page);
- if (!is_migrate_cma(mt) && !is_migrate_isolate(mt))
+ if (!is_migrate_cma(mt) && !is_migrate_isolate(mt) &&
+ !is_migrate_reserve(mt))
+ mt = migratetype;
+ } else {
+ if (!is_migrate_reserve(mt))
mt = migratetype;
}
set_freepage_migratetype(page, mt);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
