1 files changed, 184 insertions, 360 deletions

diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index 56c8687b29b..8c34363d6f1 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -1,68 +1,83 @@
 /*
- *  arch/s390/kernel/vtime.c
  *    Virtual cpu timer based timer functions.
  *
- *  S390 version
- *    Copyright (C) 2004 IBM Deutschland Entwicklung GmbH, IBM Corporation
+ *    Copyright IBM Corp. 2004, 2012
  *    Author(s): Jan Glauber <jan.glauber@de.ibm.com>
  */
 
-#include <linux/module.h>
+#include <linux/kernel_stat.h>
+#include <linux/notifier.h>
+#include <linux/kprobes.h>
+#include <linux/export.h>
 #include <linux/kernel.h>
+#include <linux/timex.h>
+#include <linux/types.h>
 #include <linux/time.h>
-#include <linux/delay.h>
-#include <linux/init.h>
+#include <linux/cpu.h>
 #include <linux/smp.h>
-#include <linux/types.h>
-#include <linux/timex.h>
-#include <linux/notifier.h>
-#include <linux/kernel_stat.h>
-#include <linux/rcupdate.h>
-#include <linux/posix-timers.h>
 
-#include <asm/s390_ext.h>
-#include <asm/timer.h>
 #include <asm/irq_regs.h>
 #include <asm/cputime.h>
+#include <asm/vtimer.h>
+#include <asm/vtime.h>
+#include <asm/irq.h>
+#include "entry.h"
 
-static DEFINE_PER_CPU(struct vtimer_queue, virt_cpu_timer);
+static void virt_timer_expire(void);
 
 DEFINE_PER_CPU(struct s390_idle_data, s390_idle);
 
-static inline __u64 get_vtimer(void)
+static LIST_HEAD(virt_timer_list);
+static DEFINE_SPINLOCK(virt_timer_lock);
+static atomic64_t virt_timer_current;
+static atomic64_t virt_timer_elapsed;
+
+static inline u64 get_vtimer(void)
 {
-	__u64 timer;
+	u64 timer;
 
-	asm volatile("STPT %0" : "=m" (timer));
+	asm volatile("stpt %0" : "=m" (timer));
 	return timer;
 }
 
-static inline void set_vtimer(__u64 expires)
+static inline void set_vtimer(u64 expires)
 {
-	__u64 timer;
+	u64 timer;
 
-	asm volatile ("  STPT %0\n"  /* Store current cpu timer value */
-		      "  SPT %1"     /* Set new value immediatly afterwards */
-		      : "=m" (timer) : "m" (expires) );
+	asm volatile(
+		"	stpt	%0\n"	/* Store current cpu timer value */
+		"	spt	%1"	/* Set new value imm. afterwards */
+		: "=m" (timer) : "m" (expires));
 	S390_lowcore.system_timer += S390_lowcore.last_update_timer - timer;
 	S390_lowcore.last_update_timer = expires;
 }
 
+static inline int virt_timer_forward(u64 elapsed)
+{
+	BUG_ON(!irqs_disabled());
+
+	if (list_empty(&virt_timer_list))
+		return 0;
+	elapsed = atomic64_add_return(elapsed, &virt_timer_elapsed);
+	return elapsed >= atomic64_read(&virt_timer_current);
+}
+
 /*
  * Update process times based on virtual cpu times stored by entry.S
  * to the lowcore fields user_timer, system_timer & steal_clock.
  */
-static void do_account_vtime(struct task_struct *tsk, int hardirq_offset)
+static int do_account_vtime(struct task_struct *tsk, int hardirq_offset)
 {
 	struct thread_info *ti = task_thread_info(tsk);
-	__u64 timer, clock, user, system, steal;
+	u64 timer, clock, user, system, steal;
 
 	timer = S390_lowcore.last_update_timer;
 	clock = S390_lowcore.last_update_clock;
-	asm volatile ("  STPT %0\n"    /* Store current cpu timer value */
-		      "  STCK %1"      /* Store current tod clock value */
-		      : "=m" (S390_lowcore.last_update_timer),
-		        "=m" (S390_lowcore.last_update_clock) );
+	asm volatile(
+		"	stpt	%0\n"	/* Store current cpu timer value */
+		"	stck	%1"	/* Store current tod clock value */
+		: "=m" (S390_lowcore.last_update_timer),
+		  "=m" (S390_lowcore.last_update_clock));
 	S390_lowcore.system_timer += timer - S390_lowcore.last_update_timer;
 	S390_lowcore.steal_timer += S390_lowcore.last_update_clock - clock;
 
@@ -81,9 +96,11 @@ static void do_account_vtime(struct task_struct *tsk, int hardirq_offset)
 		S390_lowcore.steal_timer = 0;
 		account_steal_time(steal);
 	}
+
+	return virt_timer_forward(user + system);
 }
 
-void account_vtime(struct task_struct *prev, struct task_struct *next)
+void vtime_task_switch(struct task_struct *prev)
 {
 	struct thread_info *ti;
 
@@ -91,24 +108,32 @@ void account_vtime(struct task_struct *prev, struct task_struct *next)
 	ti = task_thread_info(prev);
 	ti->user_timer = S390_lowcore.user_timer;
 	ti->system_timer = S390_lowcore.system_timer;
-	ti = task_thread_info(next);
+	ti = task_thread_info(current);
 	S390_lowcore.user_timer = ti->user_timer;
 	S390_lowcore.system_timer = ti->system_timer;
 }
 
-void account_process_tick(struct task_struct *tsk, int user_tick)
+/*
+ * In s390, accounting pending user time also implies
+ * accounting system time in order to correctly compute
+ * the stolen time accounting.
+ */
+void vtime_account_user(struct task_struct *tsk)
 {
-	do_account_vtime(tsk, HARDIRQ_OFFSET);
+	if (do_account_vtime(tsk, HARDIRQ_OFFSET))
+		virt_timer_expire();
 }
 
 /*
  * Update process times based on virtual cpu times stored by entry.S
  * to the lowcore fields user_timer, system_timer & steal_clock.
  */
-void account_system_vtime(struct task_struct *tsk)
+void vtime_account_irq_enter(struct task_struct *tsk)
 {
 	struct thread_info *ti = task_thread_info(tsk);
-	__u64 timer, system;
+	u64 timer, system;
+
+	WARN_ON_ONCE(!irqs_disabled());
 
 	timer = S390_lowcore.last_update_timer;
 	S390_lowcore.last_update_timer = get_vtimer();
@@ -118,153 +143,56 @@ void account_system_vtime(struct task_struct *tsk)
 	S390_lowcore.steal_timer -= system;
 	ti->system_timer = S390_lowcore.system_timer;
 	account_system_time(tsk, 0, system, system);
+
+	virt_timer_forward(system);
 }
-EXPORT_SYMBOL_GPL(account_system_vtime);
+EXPORT_SYMBOL_GPL(vtime_account_irq_enter);
 
-void vtime_start_cpu(__u64 int_clock, __u64 enter_timer)
+void vtime_account_system(struct task_struct *tsk)
+__attribute__((alias("vtime_account_irq_enter")));
+EXPORT_SYMBOL_GPL(vtime_account_system);
+
+void __kprobes vtime_stop_cpu(void)
 {
 	struct s390_idle_data *idle = &__get_cpu_var(s390_idle);
-	struct vtimer_queue *vq = &__get_cpu_var(virt_cpu_timer);
-	__u64 idle_time, expires;
+	unsigned long long idle_time;
+	unsigned long psw_mask;
 
-	if (idle->idle_enter == 0ULL)
-		return;
+	trace_hardirqs_on();
 
-	/* Account time spent with enabled wait psw loaded as idle time. */
-	idle_time = int_clock - idle->idle_enter;
-	account_idle_time(idle_time);
-	S390_lowcore.steal_timer +=
-		idle->idle_enter - S390_lowcore.last_update_clock;
-	S390_lowcore.last_update_clock = int_clock;
-
-	/* Account system time spent going idle. */
-	S390_lowcore.system_timer += S390_lowcore.last_update_timer - vq->idle;
-	S390_lowcore.last_update_timer = enter_timer;
-
-	/* Restart vtime CPU timer */
-	if (vq->do_spt) {
-		/* Program old expire value but first save progress. */
-		expires = vq->idle - enter_timer;
-		expires += get_vtimer();
-		set_vtimer(expires);
-	} else {
-		/* Don't account the CPU timer delta while the cpu was idle. */
-		vq->elapsed -= vq->idle - enter_timer;
-	}
+	/* Wait for external, I/O or machine check interrupt. */
+	psw_mask = PSW_KERNEL_BITS | PSW_MASK_WAIT | PSW_MASK_DAT |
+		PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK;
+	idle->nohz_delay = 0;
+
+	/* Call the assembler magic in entry.S */
+	psw_idle(idle, psw_mask);
 
+	/* Account time spent with enabled wait psw loaded as idle time. */
 	idle->sequence++;
 	smp_wmb();
+	idle_time = idle->clock_idle_exit - idle->clock_idle_enter;
+	idle->clock_idle_enter = idle->clock_idle_exit = 0ULL;
 	idle->idle_time += idle_time;
-	idle->idle_enter = 0ULL;
 	idle->idle_count++;
+	account_idle_time(idle_time);
 	smp_wmb();
 	idle->sequence++;
 }
 
-void vtime_stop_cpu(void)
-{
-	struct s390_idle_data *idle = &__get_cpu_var(s390_idle);
-	struct vtimer_queue *vq = &__get_cpu_var(virt_cpu_timer);
-	psw_t psw;
-
-	/* Wait for external, I/O or machine check interrupt. */
-	psw.mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_IO | PSW_MASK_EXT;
-
-	idle->nohz_delay = 0;
-
-	/* Check if the CPU timer needs to be reprogrammed. */
-	if (vq->do_spt) {
-		__u64 vmax = VTIMER_MAX_SLICE;
-		/*
-		 * The inline assembly is equivalent to
-		 *	vq->idle = get_cpu_timer();
-		 *	set_cpu_timer(VTIMER_MAX_SLICE);
-		 *	idle->idle_enter = get_clock();
-		 *	__load_psw_mask(psw_kernel_bits | PSW_MASK_WAIT |
-		 *			   PSW_MASK_IO | PSW_MASK_EXT);
-		 * The difference is that the inline assembly makes sure that
-		 * the last three instruction are stpt, stck and lpsw in that
-		 * order. This is done to increase the precision.
-		 */
-		asm volatile(
-#ifndef CONFIG_64BIT
-			"	basr	1,0\n"
-			"0:	ahi	1,1f-0b\n"
-			"	st	1,4(%2)\n"
-#else /* CONFIG_64BIT */
-			"	larl	1,1f\n"
-			"	stg	1,8(%2)\n"
-#endif /* CONFIG_64BIT */
-			"	stpt	0(%4)\n"
-			"	spt	0(%5)\n"
-			"	stck	0(%3)\n"
-#ifndef CONFIG_64BIT
-			"	lpsw	0(%2)\n"
-#else /* CONFIG_64BIT */
-			"	lpswe	0(%2)\n"
-#endif /* CONFIG_64BIT */
-			"1:"
-			: "=m" (idle->idle_enter), "=m" (vq->idle)
-			: "a" (&psw), "a" (&idle->idle_enter),
-			  "a" (&vq->idle), "a" (&vmax), "m" (vmax), "m" (psw)
-			: "memory", "cc", "1");
-	} else {
-		/*
-		 * The inline assembly is equivalent to
-		 *	vq->idle = get_cpu_timer();
-		 *	idle->idle_enter = get_clock();
-		 *	__load_psw_mask(psw_kernel_bits | PSW_MASK_WAIT |
-		 *			   PSW_MASK_IO | PSW_MASK_EXT);
-		 * The difference is that the inline assembly makes sure that
-		 * the last three instruction are stpt, stck and lpsw in that
-		 * order. This is done to increase the precision.
-		 */
-		asm volatile(
-#ifndef CONFIG_64BIT
-			"	basr	1,0\n"
-			"0:	ahi	1,1f-0b\n"
-			"	st	1,4(%2)\n"
-#else /* CONFIG_64BIT */
-			"	larl	1,1f\n"
-			"	stg	1,8(%2)\n"
-#endif /* CONFIG_64BIT */
-			"	stpt	0(%4)\n"
-			"	stck	0(%3)\n"
-#ifndef CONFIG_64BIT
-			"	lpsw	0(%2)\n"
-#else /* CONFIG_64BIT */
-			"	lpswe	0(%2)\n"
-#endif /* CONFIG_64BIT */
-			"1:"
-			: "=m" (idle->idle_enter), "=m" (vq->idle)
-			: "a" (&psw), "a" (&idle->idle_enter),
-			  "a" (&vq->idle), "m" (psw)
-			: "memory", "cc", "1");
-	}
-}
-
 cputime64_t s390_get_idle_time(int cpu)
 {
-	struct s390_idle_data *idle;
-	unsigned long long now, idle_time, idle_enter;
+	struct s390_idle_data *idle = &per_cpu(s390_idle, cpu);
+	unsigned long long now, idle_enter, idle_exit;
 	unsigned int sequence;
 
-	idle = &per_cpu(s390_idle, cpu);
-
-	now = get_clock();
-repeat:
-	sequence = idle->sequence;
-	smp_rmb();
-	if (sequence & 1)
-		goto repeat;
-	idle_time = 0;
-	idle_enter = idle->idle_enter;
-	if (idle_enter != 0ULL && idle_enter < now)
-		idle_time = now - idle_enter;
-	smp_rmb();
-	if (idle->sequence != sequence)
-		goto repeat;
-	return idle_time;
+	do {
+		now = get_tod_clock();
+		sequence = ACCESS_ONCE(idle->sequence);
+		idle_enter = ACCESS_ONCE(idle->clock_idle_enter);
+		idle_exit = ACCESS_ONCE(idle->clock_idle_exit);
+	} while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence));
+	return idle_enter ? ((idle_exit ?: now) - idle_enter) : 0;
 }
 
 /*
@@ -273,11 +201,11 @@ repeat:
  */
 static void list_add_sorted(struct vtimer_list *timer, struct list_head *head)
 {
-	struct vtimer_list *event;
+	struct vtimer_list *tmp;
 
-	list_for_each_entry(event, head, entry) {
-		if (event->expires > timer->expires) {
-			list_add_tail(&timer->entry, &event->entry);
+	list_for_each_entry(tmp, head, entry) {
+		if (tmp->expires > timer->expires) {
+			list_add_tail(&timer->entry, &tmp->entry);
 			return;
 		}
 	}
@@ -285,83 +213,45 @@ static void list_add_sorted(struct vtimer_list *timer, struct list_head *head)
 }
 
 /*
- * Do the callback functions of expired vtimer events.
- * Called from within the interrupt handler.
- */
-static void do_callbacks(struct list_head *cb_list)
-{
-	struct vtimer_queue *vq;
-	struct vtimer_list *event, *tmp;
-
-	if (list_empty(cb_list))
-		return;
-
-	vq = &__get_cpu_var(virt_cpu_timer);
-
-	list_for_each_entry_safe(event, tmp, cb_list, entry) {
-		list_del_init(&event->entry);
-		(event->function)(event->data);
-		if (event->interval) {
-			/* Recharge interval timer */
-			event->expires = event->interval + vq->elapsed;
-			spin_lock(&vq->lock);
-			list_add_sorted(event, &vq->list);
-			spin_unlock(&vq->lock);
-		}
-	}
-}
-
-/*
- * Handler for the virtual CPU timer.
+ * Handler for expired virtual CPU timer.
  */
-static void do_cpu_timer_interrupt(unsigned int ext_int_code,
-				   unsigned int param32, unsigned long param64)
+static void virt_timer_expire(void)
 {
-	struct vtimer_queue *vq;
-	struct vtimer_list *event, *tmp;
-	struct list_head cb_list;	/* the callback queue */
-	__u64 elapsed, next;
-
-	INIT_LIST_HEAD(&cb_list);
-	vq = &__get_cpu_var(virt_cpu_timer);
-
-	/* walk timer list, fire all expired events */
-	spin_lock(&vq->lock);
-
-	elapsed = vq->elapsed + (vq->timer - S390_lowcore.async_enter_timer);
-	BUG_ON((s64) elapsed < 0);
-	vq->elapsed = 0;
-	list_for_each_entry_safe(event, tmp, &vq->list, entry) {
-		if (event->expires < elapsed)
+	struct vtimer_list *timer, *tmp;
+	unsigned long elapsed;
+	LIST_HEAD(cb_list);
+
+	/* walk timer list, fire all expired timers */
+	spin_lock(&virt_timer_lock);
+	elapsed = atomic64_read(&virt_timer_elapsed);
+	list_for_each_entry_safe(timer, tmp, &virt_timer_list, entry) {
+		if (timer->expires < elapsed)
 			/* move expired timer to the callback queue */
-			list_move_tail(&event->entry, &cb_list);
+			list_move_tail(&timer->entry, &cb_list);
 		else
-			event->expires -= elapsed;
+			timer->expires -= elapsed;
+	}
+	if (!list_empty(&virt_timer_list)) {
+		timer = list_first_entry(&virt_timer_list,
+					 struct vtimer_list, entry);
+		atomic64_set(&virt_timer_current, timer->expires);
+	}
+	atomic64_sub(elapsed, &virt_timer_elapsed);
+	spin_unlock(&virt_timer_lock);
+
+	/* Do callbacks and recharge periodic timers */
+	list_for_each_entry_safe(timer, tmp, &cb_list, entry) {
+		list_del_init(&timer->entry);
+		timer->function(timer->data);
+		if (timer->interval) {
+			/* Recharge interval timer */
+			timer->expires = timer->interval +
+				atomic64_read(&virt_timer_elapsed);
+			spin_lock(&virt_timer_lock);
+			list_add_sorted(timer, &virt_timer_list);
+			spin_unlock(&virt_timer_lock);
+		}
 	}
-	spin_unlock(&vq->lock);
-
-	vq->do_spt = list_empty(&cb_list);
-	do_callbacks(&cb_list);
-
-	/* next event is first in list */
-	next = VTIMER_MAX_SLICE;
-	spin_lock(&vq->lock);
-	if (!list_empty(&vq->list)) {
-		event = list_first_entry(&vq->list, struct vtimer_list, entry);
-		next = event->expires;
-	} else
-		vq->do_spt = 0;
-	spin_unlock(&vq->lock);
-	/*
-	 * To improve precision add the time spent by the
-	 * interrupt handler to the elapsed time.
-	 * Note: CPU timer counts down and we got an interrupt,
-	 *	 the current content is negative
-	 */
-	elapsed = S390_lowcore.async_enter_timer - get_vtimer();
-	set_vtimer(next - elapsed);
-	vq->timer = next - elapsed;
-	vq->elapsed = elapsed;
 }
 
 void init_virt_timer(struct vtimer_list *timer)
@@ -373,179 +263,108 @@ EXPORT_SYMBOL(init_virt_timer);
 
 static inline int vtimer_pending(struct vtimer_list *timer)
 {
-	return (!list_empty(&timer->entry));
+	return !list_empty(&timer->entry);
 }
 
-/*
- * this function should only run on the specified CPU
- */
 static void internal_add_vtimer(struct vtimer_list *timer)
 {
-	struct vtimer_queue *vq;
-	unsigned long flags;
-	__u64 left, expires;
-
-	vq = &per_cpu(virt_cpu_timer, timer->cpu);
-	spin_lock_irqsave(&vq->lock, flags);
-
-	BUG_ON(timer->cpu != smp_processor_id());
-
-	if (list_empty(&vq->list)) {
-		/* First timer on this cpu, just program it. */
-		list_add(&timer->entry, &vq->list);
-		set_vtimer(timer->expires);
-		vq->timer = timer->expires;
-		vq->elapsed = 0;
+	if (list_empty(&virt_timer_list)) {
+		/* First timer, just program it. */
+		atomic64_set(&virt_timer_current, timer->expires);
+		atomic64_set(&virt_timer_elapsed, 0);
+		list_add(&timer->entry, &virt_timer_list);
 	} else {
-		/* Check progress of old timers. */
-		expires = timer->expires;
-		left = get_vtimer();
-		if (likely((s64) expires < (s64) left)) {
+		/* Update timer against current base. */
+		timer->expires += atomic64_read(&virt_timer_elapsed);
+		if (likely((s64) timer->expires <
+			   (s64) atomic64_read(&virt_timer_current)))
 			/* The new timer expires before the current timer. */
-			set_vtimer(expires);
-			vq->elapsed += vq->timer - left;
-			vq->timer = expires;
-		} else {
-			vq->elapsed += vq->timer - left;
-			vq->timer = left;
-		}
-		/* Insert new timer into per cpu list. */
-		timer->expires += vq->elapsed;
-		list_add_sorted(timer, &vq->list);
+			atomic64_set(&virt_timer_current, timer->expires);
+		/* Insert new timer into the list. */
+		list_add_sorted(timer, &virt_timer_list);
 	}
-
-	spin_unlock_irqrestore(&vq->lock, flags);
-	/* release CPU acquired in prepare_vtimer or mod_virt_timer() */
-	put_cpu();
 }
 
-static inline void prepare_vtimer(struct vtimer_list *timer)
+static void __add_vtimer(struct vtimer_list *timer, int periodic)
 {
-	BUG_ON(!timer->function);
-	BUG_ON(!timer->expires || timer->expires > VTIMER_MAX_SLICE);
-	BUG_ON(vtimer_pending(timer));
-	timer->cpu = get_cpu();
+	unsigned long flags;
+
+	timer->interval = periodic ? timer->expires : 0;
+	spin_lock_irqsave(&virt_timer_lock, flags);
+	internal_add_vtimer(timer);
+	spin_unlock_irqrestore(&virt_timer_lock, flags);
 }
 
 /*
  * add_virt_timer - add an oneshot virtual CPU timer
  */
-void add_virt_timer(void *new)
+void add_virt_timer(struct vtimer_list *timer)
 {
-	struct vtimer_list *timer;
-
-	timer = (struct vtimer_list *)new;
-	prepare_vtimer(timer);
-	timer->interval = 0;
-	internal_add_vtimer(timer);
+	__add_vtimer(timer, 0);
 }
 EXPORT_SYMBOL(add_virt_timer);
 
 /*
  * add_virt_timer_int - add an interval virtual CPU timer
  */
-void add_virt_timer_periodic(void *new)
+void add_virt_timer_periodic(struct vtimer_list *timer)
 {
-	struct vtimer_list *timer;
-
-	timer = (struct vtimer_list *)new;
-	prepare_vtimer(timer);
-	timer->interval = timer->expires;
-	internal_add_vtimer(timer);
+	__add_vtimer(timer, 1);
 }
 EXPORT_SYMBOL(add_virt_timer_periodic);
 
-int __mod_vtimer(struct vtimer_list *timer, __u64 expires, int periodic)
+static int __mod_vtimer(struct vtimer_list *timer, u64 expires, int periodic)
 {
-	struct vtimer_queue *vq;
 	unsigned long flags;
-	int cpu;
+	int rc;
 
 	BUG_ON(!timer->function);
-	BUG_ON(!expires || expires > VTIMER_MAX_SLICE);
 
 	if (timer->expires == expires && vtimer_pending(timer))
 		return 1;
-
-	cpu = get_cpu();
-	vq = &per_cpu(virt_cpu_timer, cpu);
-
-	/* disable interrupts before test if timer is pending */
-	spin_lock_irqsave(&vq->lock, flags);
-
-	/* if timer isn't pending add it on the current CPU */
-	if (!vtimer_pending(timer)) {
-		spin_unlock_irqrestore(&vq->lock, flags);
-
-		if (periodic)
-			timer->interval = expires;
-		else
-			timer->interval = 0;
-		timer->expires = expires;
-		timer->cpu = cpu;
-		internal_add_vtimer(timer);
-		return 0;
-	}
-
-	/* check if we run on the right CPU */
-	BUG_ON(timer->cpu != cpu);
-
-	list_del_init(&timer->entry);
+	spin_lock_irqsave(&virt_timer_lock, flags);
+	rc = vtimer_pending(timer);
+	if (rc)
+		list_del_init(&timer->entry);
+	timer->interval = periodic ? expires : 0;
 	timer->expires = expires;
-	if (periodic)
-		timer->interval = expires;
-
-	/* the timer can't expire anymore so we can release the lock */
-	spin_unlock_irqrestore(&vq->lock, flags);
 	internal_add_vtimer(timer);
-	return 1;
+	spin_unlock_irqrestore(&virt_timer_lock, flags);
+	return rc;
 }
 
 /*
- * If we change a pending timer the function must be called on the CPU
- * where the timer is running on.
- *
  * returns whether it has modified a pending timer (1) or not (0)
  */
-int mod_virt_timer(struct vtimer_list *timer, __u64 expires)
+int mod_virt_timer(struct vtimer_list *timer, u64 expires)
 {
 	return __mod_vtimer(timer, expires, 0);
 }
 EXPORT_SYMBOL(mod_virt_timer);
 
 /*
- * If we change a pending timer the function must be called on the CPU
- * where the timer is running on.
- *
  * returns whether it has modified a pending timer (1) or not (0)
  */
-int mod_virt_timer_periodic(struct vtimer_list *timer, __u64 expires)
+int mod_virt_timer_periodic(struct vtimer_list *timer, u64 expires)
 {
 	return __mod_vtimer(timer, expires, 1);
 }
 EXPORT_SYMBOL(mod_virt_timer_periodic);
 
 /*
- * delete a virtual timer
+ * Delete a virtual timer.
  *
  * returns whether the deleted timer was pending (1) or not (0)
  */
 int del_virt_timer(struct vtimer_list *timer)
 {
 	unsigned long flags;
-	struct vtimer_queue *vq;
 
-	/* check if timer is pending */
 	if (!vtimer_pending(timer))
 		return 0;
-
-	vq = &per_cpu(virt_cpu_timer, timer->cpu);
-	spin_lock_irqsave(&vq->lock, flags);
-
-	/* we don't interrupt a running timer, just let it expire! */
+	spin_lock_irqsave(&virt_timer_lock, flags);
 	list_del_init(&timer->entry);
-
-	spin_unlock_irqrestore(&vq->lock, flags);
+	spin_unlock_irqrestore(&virt_timer_lock, flags);
 	return 1;
 }
 EXPORT_SYMBOL(del_virt_timer);
@@ -555,24 +374,29 @@ EXPORT_SYMBOL(del_virt_timer);
  */
 void init_cpu_vtimer(void)
 {
-	struct vtimer_queue *vq;
+	/* set initial cpu timer */
+	set_vtimer(VTIMER_MAX_SLICE);
+}
 
-	/* initialize per cpu vtimer structure */
-	vq = &__get_cpu_var(virt_cpu_timer);
-	INIT_LIST_HEAD(&vq->list);
-	spin_lock_init(&vq->lock);
+static int s390_nohz_notify(struct notifier_block *self, unsigned long action,
+			    void *hcpu)
+{
+	struct s390_idle_data *idle;
+	long cpu = (long) hcpu;
 
-	/* enable cpu timer interrupts */
-	__ctl_set_bit(0,10);
+	idle = &per_cpu(s390_idle, cpu);
+	switch (action & ~CPU_TASKS_FROZEN) {
+	case CPU_DYING:
+		idle->nohz_delay = 0;
+	default:
+		break;
+	}
+	return NOTIFY_OK;
 }
 
 void __init vtime_init(void)
 {
-	/* request the cpu timer external interrupt */
-	if (register_external_interrupt(0x1005, do_cpu_timer_interrupt))
-		panic("Couldn't request external interrupt 0x1005");
-
 	/* Enable cpu timer interrupts on the boot cpu. */
 	init_cpu_vtimer();
+	cpu_notifier(s390_nohz_notify, 0);
 }
-