]> git.baikalelectronics.ru Git - kernel.git/commitdiff
mm: memcontrol: use memcg_kmem_enabled in count_objcg_event
authorKairui Song <kasong@tencent.com>
Mon, 19 Sep 2022 18:06:33 +0000 (02:06 +0800)
committerAndrew Morton <akpm@linux-foundation.org>
Mon, 3 Oct 2022 21:03:32 +0000 (14:03 -0700)
Patch series "mm: memcontrol: cleanup and optimize for two accounting
params", v2.

This patch (of 2):

There are currently two helpers for checking if cgroup kmem
accounting is enabled:

- mem_cgroup_kmem_disabled
- memcg_kmem_enabled

mem_cgroup_kmem_disabled is a simple helper that returns true
if cgroup.memory=nokmem is specified, otherwise returns false.

memcg_kmem_enabled is a bit different, it returns true if
cgroup.memory=nokmem is not specified and there was at least one
non-root memory control enabled cgroup ever created. This help improve
performance when kmem accounting was not actually activated. And it's
optimized with static branch.

The usage of mem_cgroup_kmem_disabled is for sub-systems that need to
preallocate data for kmem accounting since they could be initialized
before kmem accounting is activated. But count_objcg_event doesn't
need that, so using memcg_kmem_enabled is better here.

Link: https://lkml.kernel.org/r/20220919180634.45958-1-ryncsn@gmail.com
Link: https://lkml.kernel.org/r/20220919180634.45958-2-ryncsn@gmail.com
Signed-off-by: Kairui Song <kasong@tencent.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Acked-by: Muchun Song <songmuchun@bytedance.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
include/linux/memcontrol.h

index dc7d40e575d5f1f42f9a6ab43945daee1252ac6c..ef479e5542536735d9eade1f5d3f9663a2f6a9ba 100644 (file)
@@ -1778,7 +1778,7 @@ static inline void count_objcg_event(struct obj_cgroup *objcg,
 {
        struct mem_cgroup *memcg;
 
-       if (mem_cgroup_kmem_disabled())
+       if (!memcg_kmem_enabled())
                return;
 
        rcu_read_lock();