---

yaml
---
r: 48591
b: refs/heads/master
c: 79bf2bb
h: refs/heads/master
i:
  48589: e6d4de3
  48587: 77ba173
  48583: ce30775
  48575: 949a613
v: v3

[refs]

@@ -1,2 +1,2 @@
 ---
-refs/heads/master: f8381cba04ba8173fd5a2b8e5cd8b3290ee13a98
+refs/heads/master: 79bf2bb335b85db25d27421c798595a2fa2a0e82

trunk/Documentation/kernel-parameters.txt

@@ -1078,6 +1078,10 @@ and is between 256 and 4096 characters. It is defined in the file
 			in certain environments such as networked servers or
 			real-time systems.
 
+	nohz=		[KNL] Boottime enable/disable dynamic ticks
+			Valid arguments: on, off
+			Default: on
+
 	noirqbalance	[IA-32,SMP,KNL] Disable kernel irq balancing
 
 	noirqdebug	[IA-32] Disables the code which attempts to detect and

trunk/include/linux/hardirq.h

@@ -106,6 +106,16 @@ static inline void account_system_vtime(struct task_struct *tsk)
  * always balanced, so the interrupted value of ->hardirq_context
  * will always be restored.
  */
+#define __irq_enter()					\
+	do {						\
+		account_system_vtime(current);		\
+		add_preempt_count(HARDIRQ_OFFSET);	\
+		trace_hardirq_enter();			\
+	} while (0)
+
+/*
+ * Enter irq context (on NO_HZ, update jiffies):
+ */
 extern void irq_enter(void);
 
 /*
@@ -123,7 +133,7 @@ extern void irq_enter(void);
  */
 extern void irq_exit(void);
 
-#define nmi_enter()		do { lockdep_off(); irq_enter(); } while (0)
+#define nmi_enter()		do { lockdep_off(); __irq_enter(); } while (0)
 #define nmi_exit()		do { __irq_exit(); lockdep_on(); } while (0)
 
 #endif /* LINUX_HARDIRQ_H */

trunk/include/linux/hrtimer.h

@@ -201,4 +201,10 @@ extern void hrtimer_run_queues(void);
 /* Bootup initialization: */
 extern void __init hrtimers_init(void);
 
+#if BITS_PER_LONG < 64
+extern unsigned long ktime_divns(const ktime_t kt, s64 div);
+#else /* BITS_PER_LONG < 64 */
+# define ktime_divns(kt, div)		(unsigned long)((kt).tv64 / (div))
+#endif
+
 #endif

trunk/include/linux/tick.h

@@ -20,12 +20,79 @@ struct tick_device {
 	enum tick_device_mode mode;
 };
 
+enum tick_nohz_mode {
+	NOHZ_MODE_INACTIVE,
+	NOHZ_MODE_LOWRES,
+	NOHZ_MODE_HIGHRES,
+};
+
+/**
+ * struct tick_sched - sched tick emulation and no idle tick control/stats
+ * @sched_timer:	hrtimer to schedule the periodic tick in high
+ *			resolution mode
+ * @idle_tick:		Store the last idle tick expiry time when the tick
+ *			timer is modified for idle sleeps. This is necessary
+ *			to resume the tick timer operation in the timeline
+ *			when the CPU returns from idle
+ * @tick_stopped:	Indicator that the idle tick has been stopped
+ * @idle_jiffies:	jiffies at the entry to idle for idle time accounting
+ * @idle_calls:		Total number of idle calls
+ * @idle_sleeps:	Number of idle calls, where the sched tick was stopped
+ * @idle_entrytime:	Time when the idle call was entered
+ * @idle_sleeptime:	Sum of the time slept in idle with sched tick stopped
+ */
+struct tick_sched {
+	struct hrtimer			sched_timer;
+	unsigned long			check_clocks;
+	enum tick_nohz_mode		nohz_mode;
+	ktime_t				idle_tick;
+	int				tick_stopped;
+	unsigned long			idle_jiffies;
+	unsigned long			idle_calls;
+	unsigned long			idle_sleeps;
+	ktime_t				idle_entrytime;
+	ktime_t				idle_sleeptime;
+	unsigned long			last_jiffies;
+	unsigned long			next_jiffies;
+	ktime_t				idle_expires;
+};
+
 extern void __init tick_init(void);
+extern int tick_is_oneshot_available(void);
+
+# ifdef CONFIG_HIGH_RES_TIMERS
+extern int tick_init_highres(void);
+extern int tick_program_event(ktime_t expires, int force);
+extern void tick_setup_sched_timer(void);
+extern void tick_cancel_sched_timer(int cpu);
+# else
+static inline void tick_cancel_sched_timer(int cpu) { }
+# endif /* HIGHRES */
 
-#else
+# ifdef CONFIG_TICK_ONESHOT
+extern void tick_clock_notify(void);
+extern int tick_check_oneshot_change(int allow_nohz);
+extern struct tick_sched *tick_get_tick_sched(int cpu);
+# else
+static inline void tick_clock_notify(void) { }
+static inline int tick_check_oneshot_change(int allow_nohz) { return 0; }
+# endif
 
+#else /* CONFIG_GENERIC_CLOCKEVENTS */
 static inline void tick_init(void) { }
+static inline void tick_cancel_sched_timer(int cpu) { }
+static inline void tick_clock_notify(void) { }
+static inline int tick_check_oneshot_change(int allow_nohz) { return 0; }
+#endif /* !CONFIG_GENERIC_CLOCKEVENTS */
 
-#endif
+# ifdef CONFIG_NO_HZ
+extern void tick_nohz_stop_sched_tick(void);
+extern void tick_nohz_restart_sched_tick(void);
+extern void tick_nohz_update_jiffies(void);
+# else
+static inline void tick_nohz_stop_sched_tick(void) { }
+static inline void tick_nohz_restart_sched_tick(void) { }
+static inline void tick_nohz_update_jiffies(void) { }
+# endif /* !NO_HZ */
 
 #endif

trunk/kernel/hrtimer.c

@@ -2,8 +2,8 @@
  *  linux/kernel/hrtimer.c
  *
  *  Copyright(C) 2005-2006, Thomas Gleixner <tglx@linutronix.de>
- *  Copyright(C) 2005-2006, Red Hat, Inc., Ingo Molnar
- *  Copyright(C) 2006	    Timesys Corp., Thomas Gleixner <tglx@timesys.com>
+ *  Copyright(C) 2005-2007, Red Hat, Inc., Ingo Molnar
+ *  Copyright(C) 2006-2007  Timesys Corp., Thomas Gleixner <tglx@timesys.com>
  *
  *  High-resolution kernel timers
  *
@@ -38,6 +38,7 @@
 #include <linux/notifier.h>
 #include <linux/syscalls.h>
 #include <linux/interrupt.h>
+#include <linux/tick.h>
 
 #include <asm/uaccess.h>
 
@@ -288,7 +289,7 @@ ktime_t ktime_add_ns(const ktime_t kt, u64 nsec)
 /*
  * Divide a ktime value by a nanosecond value
  */
-static unsigned long ktime_divns(const ktime_t kt, s64 div)
+unsigned long ktime_divns(const ktime_t kt, s64 div)
 {
 	u64 dclc, inc, dns;
 	int sft = 0;
@@ -305,9 +306,6 @@ static unsigned long ktime_divns(const ktime_t kt, s64 div)
 
 	return (unsigned long) dclc;
 }
-
-#else /* BITS_PER_LONG < 64 */
-# define ktime_divns(kt, div)		(unsigned long)((kt).tv64 / (div))
 #endif /* BITS_PER_LONG >= 64 */
 
 /*
@@ -682,6 +680,16 @@ void hrtimer_run_queues(void)
 	struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
 	int i;
 
+	/*
+	 * This _is_ ugly: We have to check in the softirq context,
+	 * whether we can switch to highres and / or nohz mode. The
+	 * clocksource switch happens in the timer interrupt with
+	 * xtime_lock held. Notification from there only sets the
+	 * check bit in the tick_oneshot code, otherwise we might
+	 * deadlock vs. xtime_lock.
+	 */
+	tick_check_oneshot_change(1);
+
 	hrtimer_get_softirq_time(cpu_base);
 
 	for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++)

trunk/kernel/softirq.c

@@ -17,6 +17,7 @@
 #include <linux/kthread.h>
 #include <linux/rcupdate.h>
 #include <linux/smp.h>
+#include <linux/tick.h>
 
 #include <asm/irq.h>
 /*
@@ -278,9 +279,11 @@ EXPORT_SYMBOL(do_softirq);
  */
 void irq_enter(void)
 {
-	account_system_vtime(current);
-	add_preempt_count(HARDIRQ_OFFSET);
-	trace_hardirq_enter();
+	__irq_enter();
+#ifdef CONFIG_NO_HZ
+	if (idle_cpu(smp_processor_id()))
+		tick_nohz_update_jiffies();
+#endif
 }
 
 #ifdef __ARCH_IRQ_EXIT_IRQS_DISABLED
@@ -299,6 +302,12 @@ void irq_exit(void)
 	sub_preempt_count(IRQ_EXIT_OFFSET);
 	if (!in_interrupt() && local_softirq_pending())
 		invoke_softirq();
+
+#ifdef CONFIG_NO_HZ
+	/* Make sure that timer wheel updates are propagated */
+	if (!in_interrupt() && idle_cpu(smp_processor_id()) && !need_resched())
+		tick_nohz_stop_sched_tick();
+#endif
 	preempt_enable_no_resched();
 }
 

trunk/kernel/time/Kconfig

@@ -0,0 +1,15 @@
+#
+# Timer subsystem related configuration options
+#
+config TICK_ONESHOT
+	bool
+	default n
+
+config NO_HZ
+	bool "Tickless System (Dynamic Ticks)"
+	depends on GENERIC_TIME && GENERIC_CLOCKEVENTS
+	select TICK_ONESHOT
+	help
+	  This option enables a tickless system: timer interrupts will
+	  only trigger on an as-needed basis both when the system is
+	  busy and when the system is idle.

trunk/kernel/time/Makefile

@@ -3,3 +3,5 @@ obj-y += ntp.o clocksource.o jiffies.o
 obj-$(CONFIG_GENERIC_CLOCKEVENTS)		+= clockevents.o
 obj-$(CONFIG_GENERIC_CLOCKEVENTS)		+= tick-common.o
 obj-$(CONFIG_GENERIC_CLOCKEVENTS_BROADCAST)	+= tick-broadcast.o
+obj-$(CONFIG_TICK_ONESHOT)			+= tick-oneshot.o
+obj-$(CONFIG_TICK_ONESHOT)			+= tick-sched.o

trunk/kernel/time/clocksource.c

@@ -29,6 +29,7 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/sched.h> /* for spin_unlock_irq() using preempt_count() m68k */
+#include <linux/tick.h>
 
 /* XXX - Would like a better way for initializing curr_clocksource */
 extern struct clocksource clocksource_jiffies;
@@ -109,6 +110,13 @@ static void clocksource_watchdog(unsigned long data)
 			if ((cs->flags & CLOCK_SOURCE_IS_CONTINUOUS) &&
 			    (watchdog->flags & CLOCK_SOURCE_IS_CONTINUOUS)) {
 				cs->flags |= CLOCK_SOURCE_VALID_FOR_HRES;
+				/*
+				 * We just marked the clocksource as
+				 * highres-capable, notify the rest of the
+				 * system as well so that we transition
+				 * into high-res mode:
+				 */
+				tick_clock_notify();
 			}
 			cs->flags |= CLOCK_SOURCE_WATCHDOG;
 			cs->wd_last = csnow;