cgroups: remove rcu_read_lock() in cgroupstats_build()

cgroup_iter_* do not need rcu_read_lock().

In cgroup_enable_task_cg_lists(), do_each_thread() and while_each_thread()
are protected by RCU, it's OK, for write_lock(&css_set_lock) implies
rcu_read_lock() in non-RT kernel.

If we need explicit rcu_read_lock(), we should add rcu_read_lock() in
cgroup_enable_task_cg_lists(), not cgroup_iter_*.

Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Acked-by: Paul Menage <menage@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Balbir Singh <balbir@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kernel/cgroup.c

@@ -2055,7 +2055,6 @@ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
 
 	ret = 0;
 	cgrp = dentry->d_fsdata;
-	rcu_read_lock();
 
 	cgroup_iter_start(cgrp, &it);
 	while ((tsk = cgroup_iter_next(cgrp, &it))) {
@@ -2080,7 +2079,6 @@ int cgroupstats_build(struct cgroupstats *stats, struct dentry *dentry)
 	}
 	cgroup_iter_end(cgrp, &it);
 
-	rcu_read_unlock();
 err:
 	return ret;
 }