hrtimers: Prepare full preemption

Make cancellation of a running callback in softirq context safe
against preemption.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

7 files changed, 88 insertions, 7 deletions

diff --git a/fs/timerfd.c b/fs/timerfd.c
index d69ad801eb80..82d0f52414a6 100644
--- a/fs/timerfd.c
+++ b/fs/timerfd.c
@@ -471,7 +471,10 @@ static int do_timerfd_settime(int ufd, int flags,
 				break;
 		}
 		spin_unlock_irq(&ctx->wqh.lock);
-		cpu_relax();
+		if (isalarm(ctx))
+			hrtimer_wait_for_timer(&ctx->t.alarm.timer);
+		else
+			hrtimer_wait_for_timer(&ctx->t.tmr);
 	}
 
 	/*
diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
index b8bbaabd5aff..73ad7309436a 100644
--- a/include/linux/hrtimer.h
+++ b/include/linux/hrtimer.h
@@ -22,6 +22,7 @@
 #include <linux/percpu.h>
 #include <linux/timer.h>
 #include <linux/timerqueue.h>
+#include <linux/wait.h>
 
 struct hrtimer_clock_base;
 struct hrtimer_cpu_base;
@@ -216,6 +217,9 @@ struct hrtimer_cpu_base {
 	ktime_t				expires_next;
 	struct hrtimer			*next_timer;
 	ktime_t				softirq_expires_next;
+#ifdef CONFIG_PREEMPT_RT_BASE
+	wait_queue_head_t		wait;
+#endif
 	struct hrtimer			*softirq_next_timer;
 	struct hrtimer_clock_base	clock_base[HRTIMER_MAX_CLOCK_BASES];
 } ____cacheline_aligned;
@@ -433,6 +437,13 @@ static inline void hrtimer_restart(struct hrtimer *timer)
 	hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
 }
 
+/* Softirq preemption could deadlock timer removal */
+#ifdef CONFIG_PREEMPT_RT_BASE
+  extern void hrtimer_wait_for_timer(const struct hrtimer *timer);
+#else
+# define hrtimer_wait_for_timer(timer)	do { cpu_relax(); } while (0)
+#endif
+
 /* Query timers: */
 extern ktime_t __hrtimer_get_remaining(const struct hrtimer *timer, bool adjust);
 
@@ -458,7 +469,7 @@ static inline int hrtimer_is_queued(struct hrtimer *timer)
  * Helper function to check, whether the timer is running the callback
  * function
  */
-static inline int hrtimer_callback_running(struct hrtimer *timer)
+static inline int hrtimer_callback_running(const struct hrtimer *timer)
 {
 	return timer->base->running == timer;
 }
diff --git a/include/linux/posix-timers.h b/include/linux/posix-timers.h
index ee7e987ea1b4..0571b498db73 100644
--- a/include/linux/posix-timers.h
+++ b/include/linux/posix-timers.h
@@ -114,8 +114,8 @@ struct k_itimer {
 		struct {
 			struct alarm	alarmtimer;
 		} alarm;
-		struct rcu_head		rcu;
 	} it;
+	struct rcu_head		rcu;
 };
 
 void run_posix_cpu_timers(struct task_struct *task);
diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c
index fa5de5e8de61..6020ee66e517 100644
--- a/kernel/time/alarmtimer.c
+++ b/kernel/time/alarmtimer.c
@@ -436,7 +436,7 @@ int alarm_cancel(struct alarm *alarm)
 		int ret = alarm_try_to_cancel(alarm);
 		if (ret >= 0)
 			return ret;
-		cpu_relax();
+		hrtimer_wait_for_timer(&alarm->timer);
 	}
 }
 EXPORT_SYMBOL_GPL(alarm_cancel);
diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c
index 4f43ece42f3b..923a650e5c35 100644
--- a/kernel/time/hrtimer.c
+++ b/kernel/time/hrtimer.c
@@ -939,6 +939,33 @@ u64 hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval)
 }
 EXPORT_SYMBOL_GPL(hrtimer_forward);
 
+#ifdef CONFIG_PREEMPT_RT_BASE
+# define wake_up_timer_waiters(b)	wake_up(&(b)->wait)
+
+/**
+ * hrtimer_wait_for_timer - Wait for a running timer
+ *
+ * @timer:	timer to wait for
+ *
+ * The function waits in case the timers callback function is
+ * currently executed on the waitqueue of the timer base. The
+ * waitqueue is woken up after the timer callback function has
+ * finished execution.
+ */
+void hrtimer_wait_for_timer(const struct hrtimer *timer)
+{
+	struct hrtimer_clock_base *base = timer->base;
+
+	if (base && base->cpu_base &&
+	    base->index >= HRTIMER_BASE_MONOTONIC_SOFT)
+		wait_event(base->cpu_base->wait,
+				!(hrtimer_callback_running(timer)));
+}
+
+#else
+# define wake_up_timer_waiters(b)	do { } while (0)
+#endif
+
 /*
  * enqueue_hrtimer - internal function to (re)start a timer
  *
@@ -1171,7 +1198,7 @@ int hrtimer_cancel(struct hrtimer *timer)
 
 		if (ret >= 0)
 			return ret;
-		cpu_relax();
+		hrtimer_wait_for_timer(timer);
 	}
 }
 EXPORT_SYMBOL_GPL(hrtimer_cancel);
@@ -1477,6 +1504,7 @@ static __latent_entropy void hrtimer_run_softirq(struct softirq_action *h)
 	hrtimer_update_softirq_timer(cpu_base, true);
 
 	raw_spin_unlock_irqrestore(&cpu_base->lock, flags);
+	wake_up_timer_waiters(cpu_base);
 }
 
 #ifdef CONFIG_HIGH_RES_TIMERS
@@ -1846,6 +1874,9 @@ int hrtimers_prepare_cpu(unsigned int cpu)
 	cpu_base->softirq_next_timer = NULL;
 	cpu_base->expires_next = KTIME_MAX;
 	cpu_base->softirq_expires_next = KTIME_MAX;
+#ifdef CONFIG_PREEMPT_RT_BASE
+	init_waitqueue_head(&cpu_base->wait);
+#endif
 	return 0;
 }
 
diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c
index 9a65713c8309..55b0e58368bf 100644
--- a/kernel/time/itimer.c
+++ b/kernel/time/itimer.c
@@ -215,6 +215,7 @@ again:
 		/* We are sharing ->siglock with it_real_fn() */
 		if (hrtimer_try_to_cancel(timer) < 0) {
 			spin_unlock_irq(&tsk->sighand->siglock);
+			hrtimer_wait_for_timer(&tsk->signal->real_timer);
 			goto again;
 		}
 		expires = timeval_to_ktime(value->it_value);
diff --git a/kernel/time/posix-timers.c b/kernel/time/posix-timers.c
index 4b9127e95430..8cb187bdc4de 100644
--- a/kernel/time/posix-timers.c
+++ b/kernel/time/posix-timers.c
@@ -466,7 +466,7 @@ static struct k_itimer * alloc_posix_timer(void)
 
 static void k_itimer_rcu_free(struct rcu_head *head)
 {
-	struct k_itimer *tmr = container_of(head, struct k_itimer, it.rcu);
+	struct k_itimer *tmr = container_of(head, struct k_itimer, rcu);
 
 	kmem_cache_free(posix_timers_cache, tmr);
 }
@@ -483,7 +483,7 @@ static void release_posix_timer(struct k_itimer *tmr, int it_id_set)
 	}
 	put_pid(tmr->it_pid);
 	sigqueue_free(tmr->sigq);
-	call_rcu(&tmr->it.rcu, k_itimer_rcu_free);
+	call_rcu(&tmr->rcu, k_itimer_rcu_free);
 }
 
 static int common_timer_create(struct k_itimer *new_timer)
@@ -824,6 +824,22 @@ static void common_hrtimer_arm(struct k_itimer *timr, ktime_t expires,
 		hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
 }
 
+/*
+ * Protected by RCU!
+ */
+static void timer_wait_for_callback(const struct k_clock *kc, struct k_itimer *timr)
+{
+#ifdef CONFIG_PREEMPT_RT_FULL
+	if (kc->timer_arm == common_hrtimer_arm)
+		hrtimer_wait_for_timer(&timr->it.real.timer);
+	else if (kc == &alarm_clock)
+		hrtimer_wait_for_timer(&timr->it.alarm.alarmtimer.timer);
+	else
+		/* FIXME: Whacky hack for posix-cpu-timers */
+		schedule_timeout(1);
+#endif
+}
+
 static int common_hrtimer_try_to_cancel(struct k_itimer *timr)
 {
 	return hrtimer_try_to_cancel(&timr->it.real.timer);
@@ -888,6 +904,7 @@ retry:
 	if (!timr)
 		return -EINVAL;
 
+	rcu_read_lock();
 	kc = timr->kclock;
 	if (WARN_ON_ONCE(!kc || !kc->timer_set))
 		error = -EINVAL;
@@ -896,9 +913,12 @@ retry:
 
 	unlock_timer(timr, flag);
 	if (error == TIMER_RETRY) {
+		timer_wait_for_callback(kc, timr);
 		old_spec64 = NULL;	// We already got the old time...
+		rcu_read_unlock();
 		goto retry;
 	}
+	rcu_read_unlock();
 
 	return error;
 }
@@ -980,10 +1000,15 @@ retry_delete:
 	if (!timer)
 		return -EINVAL;
 
+	rcu_read_lock();
 	if (timer_delete_hook(timer) == TIMER_RETRY) {
 		unlock_timer(timer, flags);
+		timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
+					timer);
+		rcu_read_unlock();
 		goto retry_delete;
 	}
+	rcu_read_unlock();
 
 	spin_lock(&current->sighand->siglock);
 	list_del(&timer->list);
@@ -1009,8 +1034,18 @@ static void itimer_delete(struct k_itimer *timer)
 retry_delete:
 	spin_lock_irqsave(&timer->it_lock, flags);
 
+	/* On RT we can race with a deletion */
+	if (!timer->it_signal) {
+		unlock_timer(timer, flags);
+		return;
+	}
+
 	if (timer_delete_hook(timer) == TIMER_RETRY) {
+		rcu_read_lock();
 		unlock_timer(timer, flags);
+		timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
+					timer);
+		rcu_read_unlock();
 		goto retry_delete;
 	}
 	list_del(&timer->list);