On 12/03/2015 08:11 AM, Joonsoo Kim wrote:
If compact_order_failed is initialized to 0 and order-9 compaction is continually failed, defer counter will be updated to activate deferring. Although other defer counters will be properly updated, compact_order_failed will not be updated because failed order cannot be lower than compact_order_failed, 0. In this case, low order compaction such as 2, 3 could be deferred due to this wrongly initialized compact_order_failed value. This patch removes this possibility by initializing it to MAX_ORDER. Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Good catch. Acked-by: Vlastimil Babka <vbabka@xxxxxxx>
--- mm/page_alloc.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index d0499ff..7002c66 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -5273,6 +5273,9 @@ static void __paginginit free_area_init_core(struct pglist_data *pgdat) zone_seqlock_init(zone); zone->zone_pgdat = pgdat; zone_pcp_init(zone); +#ifdef CONFIG_COMPACTION + zone->compact_order_failed = MAX_ORDER; +#endif /* For bootup, initialized properly in watermark setup */ mod_zone_page_state(zone, NR_ALLOC_BATCH, zone->managed_pages);
-- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>