memcg: remove unnecessary !mm check from try_get_mem_cgroup_from_mm()
authorJohannes Weiner <hannes@cmpxchg.org>
Mon, 7 Apr 2014 22:37:42 +0000 (15:37 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Mon, 7 Apr 2014 23:35:56 +0000 (16:35 -0700)
Users pass either a mm that has been established under task lock, or use
a verified current->mm, which means the task can't be exiting.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/memcontrol.c

index 8d6cedd..c3b674f 100644 (file)
@@ -1075,13 +1075,6 @@ struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm)
 {
        struct mem_cgroup *memcg = NULL;
 
-       if (!mm)
-               return NULL;
-       /*
-        * Because we have no locks, mm->owner's may be being moved to other
-        * cgroup. We use css_tryget() here even if this looks
-        * pessimistic (rather than adding locks here).
-        */
        rcu_read_lock();
        do {
                memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));