cgroups: change css_set freeing mechanism to be under RCU
Ben Blum [Wed, 23 Sep 2009 22:56:29 +0000 (15:56 -0700)]
Changes css_set freeing mechanism to be under RCU

This is a prepatch for making the procs file writable. In order to free the
old css_sets for each task to be moved as they're being moved, the freeing
mechanism must be RCU-protected, or else we would have to have a call to
synchronize_rcu() for each task before freeing its old css_set.

Signed-off-by: Ben Blum <bblum@google.com>
Signed-off-by: Paul Menage <menage@google.com>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Acked-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Matt Helsley <matthltc@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/cgroup.h
kernel/cgroup.c

index 88e8634..3ac78a2 100644 (file)
@@ -260,6 +260,9 @@ struct css_set {
         * during subsystem registration (at boot time).
         */
        struct cgroup_subsys_state *subsys[CGROUP_SUBSYS_COUNT];
+
+       /* For RCU-protected deletion */
+       struct rcu_head rcu_head;
 };
 
 /*
index 3e356b0..bf8dd1a 100644 (file)
@@ -267,6 +267,12 @@ static struct hlist_head *css_set_hash(struct cgroup_subsys_state *css[])
        return &css_set_table[index];
 }
 
+static void free_css_set_rcu(struct rcu_head *obj)
+{
+       struct css_set *cg = container_of(obj, struct css_set, rcu_head);
+       kfree(cg);
+}
+
 /* We don't maintain the lists running through each css_set to its
  * task until after the first call to cgroup_iter_start(). This
  * reduces the fork()/exit() overhead for people who have cgroups
@@ -310,7 +316,7 @@ static void __put_css_set(struct css_set *cg, int taskexit)
        }
 
        write_unlock(&css_set_lock);
-       kfree(cg);
+       call_rcu(&cg->rcu_head, free_css_set_rcu);
 }
 
 /*