[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20210205062719.74431-1-songmuchun@bytedance.com>
Date: Fri, 5 Feb 2021 14:27:19 +0800
From: Muchun Song <songmuchun@...edance.com>
To: hannes@...xchg.org, mhocko@...nel.org, vdavydov.dev@...il.com,
akpm@...ux-foundation.org
Cc: cgroups@...r.kernel.org, linux-mm@...ck.org,
linux-kernel@...r.kernel.org,
Muchun Song <songmuchun@...edance.com>
Subject: [PATCH] mm: memcontrol: remove rcu_read_lock from get_mem_cgroup_from_page
The get_mem_cgroup_from_page() is called under page lock, so the page
memcg cannot be changed under us. Also, css_get is enough because page
has a reference to the memcg.
If we really want to make the get_mem_cgroup_from_page() suitable for
arbitrary page, we should use page_memcg_rcu() instead of page_memcg()
and call it after rcu_read_lock().
Signed-off-by: Muchun Song <songmuchun@...edance.com>
---
mm/memcontrol.c | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 87f01bc05d1f..6c7f1ea3955e 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1063,16 +1063,15 @@ EXPORT_SYMBOL(get_mem_cgroup_from_mm);
*/
struct mem_cgroup *get_mem_cgroup_from_page(struct page *page)
{
- struct mem_cgroup *memcg = page_memcg(page);
+ struct mem_cgroup *memcg;
if (mem_cgroup_disabled())
return NULL;
- rcu_read_lock();
/* Page should not get uncharged and freed memcg under us. */
- if (!memcg || WARN_ON_ONCE(!css_tryget(&memcg->css)))
- memcg = root_mem_cgroup;
- rcu_read_unlock();
+ memcg = page_memcg(page) ? : root_mem_cgroup;
+ css_get(&memcg->css);
+
return memcg;
}
EXPORT_SYMBOL(get_mem_cgroup_from_page);
--
2.11.0
Powered by blists - more mailing lists