---

yaml
---
r: 234332
b: refs/heads/master
c: 48228f7
h: refs/heads/master
v: v3

[refs]

@@ -1,2 +1,2 @@
 ---
-refs/heads/master: a3ec4a603faf4244e275bf11b467aad092dfbd8a
+refs/heads/master: 48228f7b470a74b6469a250d2977a13128d8fe96

trunk/kernel/timer.c

@@ -964,13 +964,36 @@ EXPORT_SYMBOL(try_to_del_timer_sync);
  * add_timer_on(). Upon exit the timer is not queued and the handler is
  * not running on any CPU.
  *
+ * Note: You must not hold locks that are held in interrupt context
+ *   while calling this function. Even if the lock has nothing to do
+ *   with the timer in question.  Here's why:
+ *
+ *    CPU0                             CPU1
+ *    ----                             ----
+ *                                   <SOFTIRQ>
+ *                                   call_timer_fn();
+ *                                     base->running_timer = mytimer;
+ *  spin_lock_irq(somelock);
+ *                                     <IRQ>
+ *                                        spin_lock(somelock);
+ *  del_timer_sync(mytimer);
+ *   while (base->running_timer == mytimer);
+ *
+ * Now del_timer_sync() will never return and never release somelock.
+ * The interrupt on the other CPU is waiting to grab somelock but
+ * it has interrupted the softirq that CPU0 is waiting to finish.
+ *
  * The function returns whether it has deactivated a pending timer or not.
  */
 int del_timer_sync(struct timer_list *timer)
 {
 #ifdef CONFIG_LOCKDEP
 	unsigned long flags;
 
+	/*
+	 * If lockdep gives a backtrace here, please reference
+	 * the synchronization rules above.
+	 */
 	local_irq_save(flags);
 	lock_map_acquire(&timer->lockdep_map);
 	lock_map_release(&timer->lockdep_map);