From 1cae9e06a1e1e68485cff3244626762ffa2bb25e Mon Sep 17 00:00:00 2001
From: Matt Helsley <matthltc@us.ibm.com>
Date: Mon, 10 May 2010 23:18:47 +0200
Subject: [PATCH] --- yaml --- r: 192503 b: refs/heads/master c:
 8f77578cc2debaeb30a4ef6206f4ba10944bdcd8 h: refs/heads/master i:   192501:
 c203d4bc599afcb42e4e460504462ebf66cae698   192499:
 35afa65689bca16fb958ead93b38bbed7e44146a   192495:
 4b35cfa724f39db46fc42b554507c0df2318c89a v: v3

---
 [refs]                        |  2 +-
 trunk/kernel/cgroup_freezer.c | 21 +++++++++++++--------
 2 files changed, 14 insertions(+), 9 deletions(-)

diff --git a/[refs] b/[refs]
index c312f4404bf1..89dfe61d5b99 100644
--- a/[refs]
+++ b/[refs]
@@ -1,2 +1,2 @@
 ---
-refs/heads/master: 543f2503a956601dd490c6cde0ebf6adb4653e50
+refs/heads/master: 8f77578cc2debaeb30a4ef6206f4ba10944bdcd8
diff --git a/trunk/kernel/cgroup_freezer.c b/trunk/kernel/cgroup_freezer.c
index e5c0244962b0..ce71ed53e88f 100644
--- a/trunk/kernel/cgroup_freezer.c
+++ b/trunk/kernel/cgroup_freezer.c
@@ -89,10 +89,10 @@ struct cgroup_subsys freezer_subsys;
 
 /* Locks taken and their ordering
  * ------------------------------
- * css_set_lock
  * cgroup_mutex (AKA cgroup_lock)
- * task->alloc_lock (AKA task_lock)
  * freezer->lock
+ * css_set_lock
+ * task->alloc_lock (AKA task_lock)
  * task->sighand->siglock
  *
  * cgroup code forces css_set_lock to be taken before task->alloc_lock
@@ -100,33 +100,38 @@ struct cgroup_subsys freezer_subsys;
  * freezer_create(), freezer_destroy():
  * cgroup_mutex [ by cgroup core ]
  *
- * can_attach():
- * cgroup_mutex
+ * freezer_can_attach():
+ * cgroup_mutex (held by caller of can_attach)
  *
- * cgroup_frozen():
+ * cgroup_freezing_or_frozen():
  * task->alloc_lock (to get task's cgroup)
  *
  * freezer_fork() (preserving fork() performance means can't take cgroup_mutex):
- * task->alloc_lock (to get task's cgroup)
  * freezer->lock
  *  sighand->siglock (if the cgroup is freezing)
  *
  * freezer_read():
  * cgroup_mutex
  *  freezer->lock
+ *   write_lock css_set_lock (cgroup iterator start)
+ *    task->alloc_lock
  *   read_lock css_set_lock (cgroup iterator start)
  *
  * freezer_write() (freeze):
  * cgroup_mutex
  *  freezer->lock
+ *   write_lock css_set_lock (cgroup iterator start)
+ *    task->alloc_lock
  *   read_lock css_set_lock (cgroup iterator start)
- *    sighand->siglock
+ *    sighand->siglock (fake signal delivery inside freeze_task())
  *
  * freezer_write() (unfreeze):
  * cgroup_mutex
  *  freezer->lock
+ *   write_lock css_set_lock (cgroup iterator start)
+ *    task->alloc_lock
  *   read_lock css_set_lock (cgroup iterator start)
- *    task->alloc_lock (to prevent races with freeze_task())
+ *    task->alloc_lock (inside thaw_process(), prevents race with refrigerator())
  *     sighand->siglock
  */
 static struct cgroup_subsys_state *freezer_create(struct cgroup_subsys *ss,