TPM: fix suspend and resume failure
[safe/jmp/linux-2.6] / kernel / hrtimer.c
index 476cb0c..e65dd0b 100644 (file)
@@ -59,6 +59,7 @@ ktime_t ktime_get(void)
 
        return timespec_to_ktime(now);
 }
+EXPORT_SYMBOL_GPL(ktime_get);
 
 /**
  * ktime_get_real - get the real (wall-) time in ktime_t format
@@ -135,20 +136,17 @@ EXPORT_SYMBOL_GPL(ktime_get_ts);
 static void hrtimer_get_softirq_time(struct hrtimer_cpu_base *base)
 {
        ktime_t xtim, tomono;
-       struct timespec xts;
+       struct timespec xts, tom;
        unsigned long seq;
 
        do {
                seq = read_seqbegin(&xtime_lock);
-#ifdef CONFIG_NO_HZ
-               getnstimeofday(&xts);
-#else
-               xts = xtime;
-#endif
+               xts = current_kernel_time();
+               tom = wall_to_monotonic;
        } while (read_seqretry(&xtime_lock, seq));
 
        xtim = timespec_to_ktime(xts);
-       tomono = timespec_to_ktime(wall_to_monotonic);
+       tomono = timespec_to_ktime(tom);
        base->clock_base[CLOCK_REALTIME].softirq_time = xtim;
        base->clock_base[CLOCK_MONOTONIC].softirq_time =
                ktime_add(xtim, tomono);
@@ -277,6 +275,32 @@ ktime_t ktime_add_ns(const ktime_t kt, u64 nsec)
 
        return ktime_add(kt, tmp);
 }
+
+EXPORT_SYMBOL_GPL(ktime_add_ns);
+
+/**
+ * ktime_sub_ns - Subtract a scalar nanoseconds value from a ktime_t variable
+ * @kt:                minuend
+ * @nsec:      the scalar nsec value to subtract
+ *
+ * Returns the subtraction of @nsec from @kt in ktime_t format
+ */
+ktime_t ktime_sub_ns(const ktime_t kt, u64 nsec)
+{
+       ktime_t tmp;
+
+       if (likely(nsec < NSEC_PER_SEC)) {
+               tmp.tv64 = nsec;
+       } else {
+               unsigned long rem = do_div(nsec, NSEC_PER_SEC);
+
+               tmp = ktime_set((long)nsec, rem);
+       }
+
+       return ktime_sub(kt, tmp);
+}
+
+EXPORT_SYMBOL_GPL(ktime_sub_ns);
 # endif /* !CONFIG_KTIME_SCALAR */
 
 /*
@@ -388,7 +412,7 @@ static int hrtimer_reprogram(struct hrtimer *timer,
        /*
         * When the callback is running, we do not reprogram the clock event
         * device. The timer callback is either running on a different CPU or
-        * the callback is executed in the hrtimer_interupt context. The
+        * the callback is executed in the hrtimer_interrupt context. The
         * reprogramming is handled either by the softirq, which called the
         * callback or at the end of the hrtimer_interrupt.
         */
@@ -458,6 +482,18 @@ void clock_was_set(void)
 }
 
 /*
+ * During resume we might have to reprogram the high resolution timer
+ * interrupt (on the local CPU):
+ */
+void hres_timers_resume(void)
+{
+       WARN_ON_ONCE(num_online_cpus() > 1);
+
+       /* Retrigger the CPU local events: */
+       retrigger_next_event(NULL);
+}
+
+/*
  * Check, whether the timer is on the callback pending list
  */
 static inline int hrtimer_cb_pending(const struct hrtimer *timer)
@@ -540,19 +576,22 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
 /*
  * Switch to high resolution mode
  */
-static void hrtimer_switch_to_hres(void)
+static int hrtimer_switch_to_hres(void)
 {
-       struct hrtimer_cpu_base *base = &__get_cpu_var(hrtimer_bases);
+       int cpu = smp_processor_id();
+       struct hrtimer_cpu_base *base = &per_cpu(hrtimer_bases, cpu);
        unsigned long flags;
 
        if (base->hres_active)
-               return;
+               return 1;
 
        local_irq_save(flags);
 
        if (tick_init_highres()) {
                local_irq_restore(flags);
-               return;
+               printk(KERN_WARNING "Could not switch to high resolution "
+                                   "mode on CPU %d\n", cpu);
+               return 0;
        }
        base->hres_active = 1;
        base->clock_base[CLOCK_REALTIME].resolution = KTIME_HIGH_RES;
@@ -563,15 +602,16 @@ static void hrtimer_switch_to_hres(void)
        /* "Retrigger" the interrupt to get things going */
        retrigger_next_event(NULL);
        local_irq_restore(flags);
-       printk(KERN_INFO "Switched to high resolution mode on CPU %d\n",
+       printk(KERN_DEBUG "Switched to high resolution mode on CPU %d\n",
               smp_processor_id());
+       return 1;
 }
 
 #else
 
 static inline int hrtimer_hres_active(void) { return 0; }
 static inline int hrtimer_is_hres_enabled(void) { return 0; }
-static inline void hrtimer_switch_to_hres(void) { }
+static inline int hrtimer_switch_to_hres(void) { return 0; }
 static inline void hrtimer_force_reprogram(struct hrtimer_cpu_base *base) { }
 static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
                                            struct hrtimer_clock_base *base)
@@ -598,7 +638,7 @@ void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer, void *addr)
 #endif
 
 /*
- * Counterpart to lock_timer_base above:
+ * Counterpart to lock_hrtimer_base above:
  */
 static inline
 void unlock_hrtimer_base(const struct hrtimer *timer, unsigned long *flags)
@@ -643,9 +683,16 @@ hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval)
                orun++;
        }
        timer->expires = ktime_add(timer->expires, interval);
+       /*
+        * Make sure, that the result did not wrap with a very large
+        * interval.
+        */
+       if (timer->expires.tv64 < 0)
+               timer->expires = ktime_set(KTIME_SEC_MAX, 0);
 
        return orun;
 }
+EXPORT_SYMBOL_GPL(hrtimer_forward);
 
 /*
  * enqueue_hrtimer - internal function to (re)start a timer
@@ -659,6 +706,7 @@ static void enqueue_hrtimer(struct hrtimer *timer,
        struct rb_node **link = &base->active.rb_node;
        struct rb_node *parent = NULL;
        struct hrtimer *entry;
+       int leftmost = 1;
 
        /*
         * Find the right place in the rbtree:
@@ -670,18 +718,19 @@ static void enqueue_hrtimer(struct hrtimer *timer,
                 * We dont care about collisions. Nodes with
                 * the same expiry time stay together.
                 */
-               if (timer->expires.tv64 < entry->expires.tv64)
+               if (timer->expires.tv64 < entry->expires.tv64) {
                        link = &(*link)->rb_left;
-               else
+               } else {
                        link = &(*link)->rb_right;
+                       leftmost = 0;
+               }
        }
 
        /*
         * Insert the timer to the rbtree and check whether it
         * replaces the first pending timer
         */
-       if (!base->first || timer->expires.tv64 <
-           rb_entry(base->first, struct hrtimer, node)->expires.tv64) {
+       if (leftmost) {
                /*
                 * Reprogram the clock event device. When the timer is already
                 * expired hrtimer_enqueue_reprogram has either called the
@@ -801,12 +850,25 @@ hrtimer_start(struct hrtimer *timer, ktime_t tim, const enum hrtimer_mode mode)
 #ifdef CONFIG_TIME_LOW_RES
                tim = ktime_add(tim, base->resolution);
 #endif
+               /*
+                * Careful here: User space might have asked for a
+                * very long sleep, so the add above might result in a
+                * negative number, which enqueues the timer in front
+                * of the queue.
+                */
+               if (tim.tv64 < 0)
+                       tim.tv64 = KTIME_MAX;
        }
        timer->expires = tim;
 
        timer_stats_hrtimer_set_start_info(timer);
 
-       enqueue_hrtimer(timer, new_base, base == 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)
+        */
+       enqueue_hrtimer(timer, new_base,
+                       new_base->cpu_base == &__get_cpu_var(hrtimer_bases));
 
        unlock_hrtimer_base(timer, &flags);
 
@@ -1130,6 +1192,9 @@ static inline void run_hrtimer_queue(struct hrtimer_cpu_base *cpu_base,
                if (base->softirq_time.tv64 <= timer->expires.tv64)
                        break;
 
+#ifdef CONFIG_HIGH_RES_TIMERS
+               WARN_ON_ONCE(timer->cb_mode == HRTIMER_CB_IRQSAFE_NO_SOFTIRQ);
+#endif
                timer_stats_account_hrtimer(timer);
 
                fn = timer->function;
@@ -1173,7 +1238,8 @@ void hrtimer_run_queues(void)
         * deadlock vs. xtime_lock.
         */
        if (tick_check_oneshot_change(!hrtimer_is_hres_enabled()))
-               hrtimer_switch_to_hres();
+               if (hrtimer_switch_to_hres())
+                       return;
 
        hrtimer_get_softirq_time(cpu_base);
 
@@ -1228,8 +1294,7 @@ static int __sched do_nanosleep(struct hrtimer_sleeper *t, enum hrtimer_mode mod
 long __sched hrtimer_nanosleep_restart(struct restart_block *restart)
 {
        struct hrtimer_sleeper t;
-       struct timespec __user *rmtp;
-       struct timespec tu;
+       struct timespec *rmtp;
        ktime_t time;
 
        restart->fn = do_no_restart_syscall;
@@ -1240,14 +1305,12 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart)
        if (do_nanosleep(&t, HRTIMER_MODE_ABS))
                return 0;
 
-       rmtp = (struct timespec __user *) restart->arg1;
+       rmtp = (struct timespec *)restart->arg1;
        if (rmtp) {
                time = ktime_sub(t.timer.expires, t.timer.base->get_time());
                if (time.tv64 <= 0)
                        return 0;
-               tu = ktime_to_timespec(time);
-               if (copy_to_user(rmtp, &tu, sizeof(tu)))
-                       return -EFAULT;
+               *rmtp = ktime_to_timespec(time);
        }
 
        restart->fn = hrtimer_nanosleep_restart;
@@ -1256,12 +1319,11 @@ long __sched hrtimer_nanosleep_restart(struct restart_block *restart)
        return -ERESTART_RESTARTBLOCK;
 }
 
-long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp,
+long hrtimer_nanosleep(struct timespec *rqtp, struct timespec *rmtp,
                       const enum hrtimer_mode mode, const clockid_t clockid)
 {
        struct restart_block *restart;
        struct hrtimer_sleeper t;
-       struct timespec tu;
        ktime_t rem;
 
        hrtimer_init(&t.timer, clockid, mode);
@@ -1277,9 +1339,7 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp,
                rem = ktime_sub(t.timer.expires, t.timer.base->get_time());
                if (rem.tv64 <= 0)
                        return 0;
-               tu = ktime_to_timespec(rem);
-               if (copy_to_user(rmtp, &tu, sizeof(tu)))
-                       return -EFAULT;
+               *rmtp = ktime_to_timespec(rem);
        }
 
        restart = &current_thread_info()->restart_block;
@@ -1295,7 +1355,8 @@ long hrtimer_nanosleep(struct timespec *rqtp, struct timespec __user *rmtp,
 asmlinkage long
 sys_nanosleep(struct timespec __user *rqtp, struct timespec __user *rmtp)
 {
-       struct timespec tu;
+       struct timespec tu, rmt;
+       int ret;
 
        if (copy_from_user(&tu, rqtp, sizeof(tu)))
                return -EFAULT;
@@ -1303,7 +1364,15 @@ sys_nanosleep(struct timespec __user *rqtp, struct timespec __user *rmtp)
        if (!timespec_valid(&tu))
                return -EINVAL;
 
-       return hrtimer_nanosleep(&tu, rmtp, HRTIMER_MODE_REL, CLOCK_MONOTONIC);
+       ret = hrtimer_nanosleep(&tu, rmtp ? &rmt : NULL, HRTIMER_MODE_REL,
+                               CLOCK_MONOTONIC);
+
+       if (ret && rmtp) {
+               if (copy_to_user(rmtp, &rmt, sizeof(*rmtp)))
+                       return -EFAULT;
+       }
+
+       return ret;
 }
 
 /*
@@ -1355,17 +1424,16 @@ static void migrate_hrtimers(int cpu)
        tick_cancel_sched_timer(cpu);
 
        local_irq_disable();
-
-       spin_lock(&new_base->lock);
-       spin_lock(&old_base->lock);
+       double_spin_lock(&new_base->lock, &old_base->lock,
+                        smp_processor_id() < cpu);
 
        for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
                migrate_hrtimer_list(&old_base->clock_base[i],
                                     &new_base->clock_base[i]);
        }
-       spin_unlock(&old_base->lock);
-       spin_unlock(&new_base->lock);
 
+       double_spin_unlock(&new_base->lock, &old_base->lock,
+                          smp_processor_id() < cpu);
        local_irq_enable();
        put_cpu_var(hrtimer_bases);
 }
@@ -1374,16 +1442,18 @@ static void migrate_hrtimers(int cpu)
 static int __cpuinit hrtimer_cpu_notify(struct notifier_block *self,
                                        unsigned long action, void *hcpu)
 {
-       long cpu = (long)hcpu;
+       unsigned int cpu = (long)hcpu;
 
        switch (action) {
 
        case CPU_UP_PREPARE:
+       case CPU_UP_PREPARE_FROZEN:
                init_hrtimers_cpu(cpu);
                break;
 
 #ifdef CONFIG_HOTPLUG_CPU
        case CPU_DEAD:
+       case CPU_DEAD_FROZEN:
                clockevents_notify(CLOCK_EVT_NOTIFY_CPU_DEAD, &cpu);
                migrate_hrtimers(cpu);
                break;