LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH 2/3] hrtimers: simplify lockdep stuff
@ 2008-02-15 18:01 Oleg Nesterov
  0 siblings, 0 replies; only message in thread
From: Oleg Nesterov @ 2008-02-15 18:01 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Arjan van de Ven, Heiko Carstens, Ingo Molnar, Peter Zijlstra,
	Thomas Gleixner, linux-kernel

In order to avoid the false positive from lockdep, each per-cpu base->lock has
the separate lock class and migrate_hrtimers() uses double_spin_lock().

This is overcomplicated: except for migrate_hrtimers() we never take 2 locks at
once, and migrate_hrtimers() can use spin_lock_nested().

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>

--- 25/include/linux/hrtimer.h~3_HRTIMERS	2008-02-15 16:59:17.000000000 +0300
+++ 25/include/linux/hrtimer.h	2008-02-15 19:23:02.000000000 +0300
@@ -173,7 +173,6 @@ struct hrtimer_clock_base {
  * struct hrtimer_cpu_base - the per cpu clock bases
  * @lock:		lock protecting the base and associated clock bases
  *			and timers
- * @lock_key:		the lock_class_key for use with lockdep
  * @clock_base:		array of clock bases for this cpu
  * @curr_timer:		the timer which is executing a callback right now
  * @expires_next:	absolute time of the next event which was scheduled
@@ -189,7 +188,6 @@ struct hrtimer_clock_base {
  */
 struct hrtimer_cpu_base {
 	spinlock_t			lock;
-	struct lock_class_key		lock_key;
 	struct hrtimer_clock_base	clock_base[HRTIMER_MAX_CLOCK_BASES];
 	struct list_head		cb_pending;
 #ifdef CONFIG_HIGH_RES_TIMERS
--- 25/kernel/hrtimer.c~3_HRTIMERS	2008-02-15 16:59:17.000000000 +0300
+++ 25/kernel/hrtimer.c	2008-02-15 19:51:05.000000000 +0300
@@ -1411,7 +1411,6 @@ static void __cpuinit init_hrtimers_cpu(
 	int i;
 
 	spin_lock_init(&cpu_base->lock);
-	lockdep_set_class(&cpu_base->lock, &cpu_base->lock_key);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++)
 		cpu_base->clock_base[i].cpu_base = cpu_base;
@@ -1452,16 +1451,16 @@ static void migrate_hrtimers(int cpu)
 	tick_cancel_sched_timer(cpu);
 
 	local_irq_disable();
-	double_spin_lock(&new_base->lock, &old_base->lock,
-			 smp_processor_id() < cpu);
+	spin_lock(&new_base->lock);
+	spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
 		migrate_hrtimer_list(&old_base->clock_base[i],
 				     &new_base->clock_base[i]);
 	}
 
-	double_spin_unlock(&new_base->lock, &old_base->lock,
-			   smp_processor_id() < cpu);
+	spin_unlock(&old_base->lock);
+	spin_unlock(&new_base->lock);
 	local_irq_enable();
 	put_cpu_var(hrtimer_bases);
 }


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2008-02-15 17:59 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-02-15 18:01 [PATCH 2/3] hrtimers: simplify lockdep stuff Oleg Nesterov

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).