提交 998d39cb 编写于 作者: C Cody P Schafer 提交者: Linus Torvalds

mm/page_alloc: protect pcp->batch accesses with ACCESS_ONCE

pcp->batch could change at any point, avoid relying on it being a stable
value.
Signed-off-by: NCody P Schafer <cody@linux.vnet.ibm.com>
Cc: Gilad Ben-Yossef <gilad@benyossef.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@gmail.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
上级 8d7a8fa9
......@@ -1182,10 +1182,12 @@ void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp)
{
unsigned long flags;
int to_drain;
unsigned long batch;
local_irq_save(flags);
if (pcp->count >= pcp->batch)
to_drain = pcp->batch;
batch = ACCESS_ONCE(pcp->batch);
if (pcp->count >= batch)
to_drain = batch;
else
to_drain = pcp->count;
if (to_drain > 0) {
......@@ -1353,8 +1355,9 @@ void free_hot_cold_page(struct page *page, int cold)
list_add(&page->lru, &pcp->lists[migratetype]);
pcp->count++;
if (pcp->count >= pcp->high) {
free_pcppages_bulk(zone, pcp->batch, pcp);
pcp->count -= pcp->batch;
unsigned long batch = ACCESS_ONCE(pcp->batch);
free_pcppages_bulk(zone, batch, pcp);
pcp->count -= batch;
}
out:
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册