After memory hot-added, users could online pages through sysfs, and this
could be done in parallel.
In case two threads online pages in two different empty zones at the
same time, there would be a contention to update the nr_zones.
The patch use pgdat_resize_lock() to protect this critical section.
Signed-off-by: Wei Yang <richard.weiyang@xxxxxxxxx>
---
mm/page_alloc.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index e13987c2e1c4..525a5344a13b 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -5796,9 +5796,12 @@ void __meminit init_currently_empty_zone(struct zone *zone,
{
struct pglist_data *pgdat = zone->zone_pgdat;
int zone_idx = zone_idx(zone) + 1;
+ unsigned long flags;
+ pgdat_resize_lock(pgdat, &flags);
if (zone_idx > pgdat->nr_zones)
pgdat->nr_zones = zone_idx;
+ pgdat_resize_unlock(pgdat, &flags);
zone->zone_start_pfn = zone_start_pfn;
--
2.15.1
