Merge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  hrtimer: splitout peek ahead functionality, fix
  hrtimer: fixup comments
  hrtimer: fix recursion deadlock by re-introducing the softirq
  hrtimer: simplify hotplug migration
  hrtimer: fix HOTPLUG_CPU=n compile warning
  hrtimer: splitout peek ahead functionality

2 files changed, 72 insertions, 73 deletions

diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index af886b26c9d1..9127f6b51a39 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -252,7 +252,8 @@ enum
         BLOCK_SOFTIRQ,
         TASKLET_SOFTIRQ,
         SCHED_SOFTIRQ,
-        RCU_SOFTIRQ,    /* Preferable RCU should always be the last softirq */
+        HRTIMER_SOFTIRQ,
+        RCU_SOFTIRQ,    /* Preferable RCU should always be the last softirq */
 
         NR_SOFTIRQS
 };
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index eb2bfefa6dcc..1455b7651b6b 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -634,7 +634,6 @@ static inline void hrtimer_init_timer_hres(struct hrtimer *timer)
 {
 }
 
-static void __run_hrtimer(struct hrtimer *timer);
 
 /*
  * When High resolution timers are active, try to reprogram. Note, that in case
@@ -646,13 +645,9 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
                                             struct hrtimer_clock_base *base)
 {
         if (base->cpu_base->hres_active && hrtimer_reprogram(timer, base)) {
-                /*
-                 * XXX: recursion check?
-                 * hrtimer_forward() should round up with timer granularity
-                 * so that we never get into inf recursion here,
-                 * it doesn't do that though
-                 */
-                __run_hrtimer(timer);
+                spin_unlock(&base->cpu_base->lock);
+                raise_softirq_irqoff(HRTIMER_SOFTIRQ);
+                spin_lock(&base->cpu_base->lock);
                 return 1;
         }
         return 0;
@@ -705,11 +700,6 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
 }
 static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { }
 static inline void hrtimer_init_timer_hres(struct hrtimer *timer) { }
-static inline int hrtimer_reprogram(struct hrtimer *timer,
-                                    struct hrtimer_clock_base *base)
-{
-        return 0;
-}
 
 #endif /* CONFIG_HIGH_RES_TIMERS */
 
@@ -780,9 +770,11 @@ EXPORT_SYMBOL_GPL(hrtimer_forward);
  *
  * The timer is inserted in expiry order. Insertion into the
  * red black tree is O(log(n)). Must hold the base lock.
+ *
+ * Returns 1 when the new timer is the leftmost timer in the tree.
  */
-static void enqueue_hrtimer(struct hrtimer *timer,
-                            struct hrtimer_clock_base *base, int reprogram)
+static int enqueue_hrtimer(struct hrtimer *timer,
+                           struct hrtimer_clock_base *base)
 {
         struct rb_node **link = &base->active.rb_node;
         struct rb_node *parent = NULL;
@@ -814,20 +806,8 @@ static void enqueue_hrtimer(struct hrtimer *timer,
          * Insert the timer to the rbtree and check whether it
          * replaces the first pending timer
          */
-        if (leftmost) {
-                /*
-                 * Reprogram the clock event device. When the timer is already
-                 * expired hrtimer_enqueue_reprogram has either called the
-                 * callback or added it to the pending list and raised the
-                 * softirq.
-                 *
-                 * This is a NOP for !HIGHRES
-                 */
-                if (reprogram && hrtimer_enqueue_reprogram(timer, base))
-                        return;
-
+        if (leftmost)
                 base->first = &timer->node;
-        }
 
         rb_link_node(&timer->node, parent, link);
         rb_insert_color(&timer->node, &base->active);
@@ -836,6 +816,8 @@ static void enqueue_hrtimer(struct hrtimer *timer,
          * state of a possibly running callback.
          */
         timer->state |= HRTIMER_STATE_ENQUEUED;
+
+        return leftmost;
 }
 
 /*
@@ -912,7 +894,7 @@ hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, unsigned long delta_n
 {
         struct hrtimer_clock_base *base, *new_base;
         unsigned long flags;
-        int ret;
+        int ret, leftmost;
 
         base = lock_hrtimer_base(timer, &flags);
 
@@ -940,12 +922,16 @@ hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim, unsigned long delta_n
 
         timer_stats_hrtimer_set_start_info(timer);
 
+        leftmost = enqueue_hrtimer(timer, new_base);
+
         /*
          * Only allow reprogramming if the new base is on this CPU.
          * (it might still be on another CPU if the timer was pending)
+         *
+         * XXX send_remote_softirq() ?
          */
-        enqueue_hrtimer(timer, new_base,
-                        new_base->cpu_base == &__get_cpu_var(hrtimer_bases));
+        if (leftmost && new_base->cpu_base == &__get_cpu_var(hrtimer_bases))
+                hrtimer_enqueue_reprogram(timer, new_base);
 
         unlock_hrtimer_base(timer, &flags);
 
@@ -1157,13 +1143,13 @@ static void __run_hrtimer(struct hrtimer *timer)
         spin_lock(&cpu_base->lock);
 
         /*
-         * Note: We clear the CALLBACK bit after enqueue_hrtimer to avoid
-         * reprogramming of the event hardware. This happens at the end of this
-         * function anyway.
+         * Note: We clear the CALLBACK bit after enqueue_hrtimer and
+         * we do not reprogramm the event hardware. Happens either in
+         * hrtimer_start_range_ns() or in hrtimer_interrupt()
          */
         if (restart != HRTIMER_NORESTART) {
                 BUG_ON(timer->state != HRTIMER_STATE_CALLBACK);
-                enqueue_hrtimer(timer, base, 0);
+                enqueue_hrtimer(timer, base);
         }
         timer->state &= ~HRTIMER_STATE_CALLBACK;
 }
@@ -1243,6 +1229,22 @@ void hrtimer_interrupt(struct clock_event_device *dev)
         }
 }
 
+/*
+ * local version of hrtimer_peek_ahead_timers() called with interrupts
+ * disabled.
+ */
+static void __hrtimer_peek_ahead_timers(void)
+{
+        struct tick_device *td;
+
+        if (!hrtimer_hres_active())
+                return;
+
+        td = &__get_cpu_var(tick_cpu_device);
+        if (td && td->evtdev)
+                hrtimer_interrupt(td->evtdev);
+}
+
 /**
  * hrtimer_peek_ahead_timers -- run soft-expired timers now
  *
@@ -1254,20 +1256,23 @@ void hrtimer_interrupt(struct clock_event_device *dev)
  */