[GIT PULL] timers fixes

[GIT PULL] timers fixes

[Frederic Weisbecker]

Thomas, Ingo,

Please pull the timers/urgent branch that can be found at:

git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks.git
	timers/urgent

HEAD: aebacb7f6ca1926918734faae14d1f0b6fae5cb7

Thanks,
	Frederic
---

Frederic Weisbecker (1):
      posix-cpu-timers: Fix rearm racing against process tick

Nicolas Saenz Julienne (1):
      timers: Fix get_next_timer_interrupt() with no timers pending


 kernel/time/posix-cpu-timers.c | 10 +++++-----
 kernel/time/timer.c            |  8 +++++---
 2 files changed, 10 insertions(+), 8 deletions(-)

[PATCH 1/2] posix-cpu-timers: Fix rearm racing against process tick

[Frederic Weisbecker]

Since the process wide cputime counter is started locklessly from
posix_cpu_timer_rearm(), it can be concurrently stopped by operations
on other timers from the same thread group, such as in the following
unlucky scenario:

         CPU 0                                CPU 1
         -----                                -----
                                           timer_settime(TIMER B)
   posix_cpu_timer_rearm(TIMER A)
       cpu_clock_sample_group()
           (pct->timers_active already true)

                                           handle_posix_cpu_timers()
                                               check_process_timers()
                                                   stop_process_timers()
                                                       pct->timers_active = false
       arm_timer(TIMER A)

   tick -> run_posix_cpu_timers()
       // sees !pct->timers_active, ignore
       // our TIMER A

Fix this with simply locking process wide cputime counting start and
timer arm in the same block.

Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Fixes: 60f2ceaa8111 ("posix-cpu-timers: Remove unnecessary locking around cpu_clock_sample_group")
Cc: stable@vger.kernel.org
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Eric W. Biederman <ebiederm@xmission.com>
---
 kernel/time/posix-cpu-timers.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c
index 29a5e54e6e10..517be7fd175e 100644
--- a/kernel/time/posix-cpu-timers.c
+++ b/kernel/time/posix-cpu-timers.c
@@ -991,6 +991,11 @@ static void posix_cpu_timer_rearm(struct k_itimer *timer)
 	if (!p)
 		goto out;
 
+	/* Protect timer list r/w in arm_timer() */
+	sighand = lock_task_sighand(p, &flags);
+	if (unlikely(sighand == NULL))
+		goto out;
+
 	/*
 	 * Fetch the current sample and update the timer's expiry time.
 	 */
@@ -1001,11 +1006,6 @@ static void posix_cpu_timer_rearm(struct k_itimer *timer)
 
 	bump_cpu_timer(timer, now);
 
-	/* Protect timer list r/w in arm_timer() */
-	sighand = lock_task_sighand(p, &flags);
-	if (unlikely(sighand == NULL))
-		goto out;
-
 	/*
 	 * Now re-arm for the new expiry time.
 	 */
-- 
2.25.1

[PATCH 2/2] timers: Fix get_next_timer_interrupt() with no timers pending

[Frederic Weisbecker]

From: Nicolas Saenz Julienne <nsaenzju@redhat.com>

31cd0e119d50 ("timers: Recalculate next timer interrupt only when
necessary") subtly altered get_next_timer_interrupt()'s behaviour. The
function no longer consistently returns KTIME_MAX with no timers
pending.

In order to decide if there are any timers pending we check whether the
next expiry will happen NEXT_TIMER_MAX_DELTA jiffies from now.
Unfortunately, the next expiry time and the timer base clock are no
longer updated in unison. The former changes upon certain timer
operations (enqueue, expire, detach), whereas the latter keeps track of
jiffies as they move forward. Ultimately breaking the logic above.

A simplified example:

- Upon entering get_next_timer_interrupt() with:

	jiffies = 1
	base->clk = 0;
	base->next_expiry = NEXT_TIMER_MAX_DELTA;

  'base->next_expiry == base->clk + NEXT_TIMER_MAX_DELTA', the function
  returns KTIME_MAX.

- 'base->clk' is updated to the jiffies value.

- The next time we enter get_next_timer_interrupt(), taking into account
  no timer operations happened:

	base->clk = 1;
	base->next_expiry = NEXT_TIMER_MAX_DELTA;

  'base->next_expiry != base->clk + NEXT_TIMER_MAX_DELTA', the function
  returns a valid expire time, which is incorrect.

This ultimately might unnecessarily rearm sched's timer on nohz_full
setups, and add latency to the system[1].

So, introduce 'base->timers_pending'[2], update it every time
'base->next_expiry' changes, and use it in get_next_timer_interrupt().

[1] See tick_nohz_stop_tick().
[2] A quick pahole check on x86_64 and arm64 shows it doesn't make
    'struct timer_base' any bigger.

Fixes: 31cd0e119d50 ("timers: Recalculate next timer interrupt only when necessary")
Signed-off-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
---
 kernel/time/timer.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/kernel/time/timer.c b/kernel/time/timer.c
index 3fadb58fc9d7..9eb11c2209e5 100644
--- a/kernel/time/timer.c
+++ b/kernel/time/timer.c
@@ -207,6 +207,7 @@ struct timer_base {
 	unsigned int		cpu;
 	bool			next_expiry_recalc;
 	bool			is_idle;
+	bool			timers_pending;
 	DECLARE_BITMAP(pending_map, WHEEL_SIZE);
 	struct hlist_head	vectors[WHEEL_SIZE];
 } ____cacheline_aligned;
@@ -595,6 +596,7 @@ static void enqueue_timer(struct timer_base *base, struct timer_list *timer,
 		 * can reevaluate the wheel:
 		 */
 		base->next_expiry = bucket_expiry;
+		base->timers_pending = true;
 		base->next_expiry_recalc = false;
 		trigger_dyntick_cpu(base, timer);
 	}
@@ -1582,6 +1584,7 @@ static unsigned long __next_timer_interrupt(struct timer_base *base)
 	}
 
 	base->next_expiry_recalc = false;
+	base->timers_pending = !(next == base->clk + NEXT_TIMER_MAX_DELTA);
 
 	return next;
 }
@@ -1633,7 +1636,6 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem)
 	struct timer_base *base = this_cpu_ptr(&timer_bases[BASE_STD]);
 	u64 expires = KTIME_MAX;
 	unsigned long nextevt;
-	bool is_max_delta;
 
 	/*
 	 * Pretend that there is no timer pending if the cpu is offline.
@@ -1646,7 +1648,6 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem)
 	if (base->next_expiry_recalc)
 		base->next_expiry = __next_timer_interrupt(base);
 	nextevt = base->next_expiry;
-	is_max_delta = (nextevt == base->clk + NEXT_TIMER_MAX_DELTA);
 
 	/*
 	 * We have a fresh next event. Check whether we can forward the
@@ -1664,7 +1665,7 @@ u64 get_next_timer_interrupt(unsigned long basej, u64 basem)
 		expires = basem;
 		base->is_idle = false;
 	} else {
-		if (!is_max_delta)
+		if (base->timers_pending)
 			expires = basem + (u64)(nextevt - basej) * TICK_NSEC;
 		/*
 		 * If we expect to sleep more than a tick, mark the base idle.
@@ -1947,6 +1948,7 @@ int timers_prepare_cpu(unsigned int cpu)
 		base = per_cpu_ptr(&timer_bases[b], cpu);
 		base->clk = jiffies;
 		base->next_expiry = base->clk + NEXT_TIMER_MAX_DELTA;
+		base->timers_pending = false;
 		base->is_idle = false;
 	}
 	return 0;
-- 
2.25.1

[tip: timers/urgent] Merge branch 'timers/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks into timers/urgent

[tip-bot2 for Thomas Gleixner]

The following commit has been merged into the timers/urgent branch of tip:

Commit-ID:     ff5a6a3550cef4a272fee19520a13699343b6a47
Gitweb:        https://git.kernel.org/tip/ff5a6a3550cef4a272fee19520a13699343b6a47
Author:        Thomas Gleixner <tglx@linutronix.de>
AuthorDate:    Tue, 20 Jul 2021 12:51:23 +02:00
Committer:     Thomas Gleixner <tglx@linutronix.de>
CommitterDate: Tue, 20 Jul 2021 12:51:23 +02:00

Merge branch 'timers/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks into timers/urgent

Pull dyntick fixes from Frederic Weisbecker:

  - Fix a rearm race in the posix cpu timer code
  - Handle get_next_timer_interrupt() correctly when no timers are pending

Link: https://lore.kernel.org/r/20210715104218.81276-1-frederic@kernel.org
---

[GIT PULL] timers fixes

[Ingo Molnar]

Linus,

Please pull the latest timers-urgent-for-linus git tree from:

   git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus

   # HEAD: de95a991bb72e009f47e0c4bbc90fc5f594588d5 tick/sched: Annotate lockless access to last_jiffies_update

Three fixes: fix link failure on Alpha, fix a Sparse warning and 
annotate/robustify a lockless access in the NOHZ code.

 Thanks,

	Ingo

------------------>
Arnd Bergmann (1):
      time/posix-stubs: Provide compat itimer supoprt for alpha

Eric Dumazet (1):
      tick/sched: Annotate lockless access to last_jiffies_update

Vincenzo Frascino (1):
      lib/vdso: Make __cvdso_clock_getres() static


 kernel/time/posix-stubs.c |  3 +++
 kernel/time/tick-sched.c  | 14 +++++++++-----
 lib/vdso/gettimeofday.c   |  1 +
 3 files changed, 13 insertions(+), 5 deletions(-)

diff --git a/kernel/time/posix-stubs.c b/kernel/time/posix-stubs.c
index 67df65f887ac..20c65a7d4e3a 100644
--- a/kernel/time/posix-stubs.c
+++ b/kernel/time/posix-stubs.c
@@ -151,6 +151,9 @@ SYSCALL_DEFINE4(clock_nanosleep, const clockid_t, which_clock, int, flags,
 
 #ifdef CONFIG_COMPAT
 COMPAT_SYS_NI(timer_create);
+#endif
+
+#if defined(CONFIG_COMPAT) || defined(CONFIG_ALPHA)
 COMPAT_SYS_NI(getitimer);
 COMPAT_SYS_NI(setitimer);
 #endif
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 8b192e67aabc..a792d21cac64 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -58,8 +58,9 @@ static void tick_do_update_jiffies64(ktime_t now)
 
 	/*
 	 * Do a quick check without holding jiffies_lock:
+	 * The READ_ONCE() pairs with two updates done later in this function.
 	 */
-	delta = ktime_sub(now, last_jiffies_update);
+	delta = ktime_sub(now, READ_ONCE(last_jiffies_update));
 	if (delta < tick_period)
 		return;
 
@@ -70,8 +71,9 @@ static void tick_do_update_jiffies64(ktime_t now)
 	if (delta >= tick_period) {
 
 		delta = ktime_sub(delta, tick_period);
-		last_jiffies_update = ktime_add(last_jiffies_update,
-						tick_period);
+		/* Pairs with the lockless read in this function. */
+		WRITE_ONCE(last_jiffies_update,
+			   ktime_add(last_jiffies_update, tick_period));
 
 		/* Slow path for long timeouts */
 		if (unlikely(delta >= tick_period)) {
@@ -79,8 +81,10 @@ static void tick_do_update_jiffies64(ktime_t now)
 
 			ticks = ktime_divns(delta, incr);
 
-			last_jiffies_update = ktime_add_ns(last_jiffies_update,
-							   incr * ticks);
+			/* Pairs with the lockless read in this function. */
+			WRITE_ONCE(last_jiffies_update,
+				   ktime_add_ns(last_jiffies_update,
+						incr * ticks));
 		}
 		do_timer(++ticks);
 
diff --git a/lib/vdso/gettimeofday.c b/lib/vdso/gettimeofday.c
index 9ecfd3b547ba..42bd8ab955fa 100644
--- a/lib/vdso/gettimeofday.c
+++ b/lib/vdso/gettimeofday.c
@@ -221,6 +221,7 @@ int __cvdso_clock_getres_common(clockid_t clock, struct __kernel_timespec *res)
 	return 0;
 }
 
+static __maybe_unused
 int __cvdso_clock_getres(clockid_t clock, struct __kernel_timespec *res)
 {
 	int ret = __cvdso_clock_getres_common(clock, res);

Re: [GIT PULL] timers fixes

[pr-tracker-bot]

The pull request you sent on Sat, 18 Jan 2020 19:47:46 +0100:

> git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git timers-urgent-for-linus

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/7ff15cd0458c288afa954de843bb1903c723e5b3

Thank you!

-- 
Deet-doot-dot, I am a bot.
https://korg.wiki.kernel.org/userdoc/prtracker

[GIT PULL] timers fixes

[Ingo Molnar]

Linus,

Please pull the latest timers-fixes-for-linus git tree from:

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

 Thanks,

	Ingo

------------------>
H Hartley Sweeten (1):
      timekeeping: Make local variables static

Nicolas Pitre (1):
      time: Rename misnamed minsec argument of clocks_calc_mult_shift()


 kernel/time/clocksource.c |    8 ++++----
 kernel/time/timekeeping.c |    4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
index c18d7ef..8588abc 100644
--- a/kernel/time/clocksource.c
+++ b/kernel/time/clocksource.c
@@ -113,7 +113,7 @@ EXPORT_SYMBOL_GPL(timecounter_cyc2time);
  * @shift:	pointer to shift variable
  * @from:	frequency to convert from
  * @to:		frequency to convert to
- * @minsec:	guaranteed runtime conversion range in seconds
+ * @maxsec:	guaranteed runtime conversion range in seconds
  *
  * The function evaluates the shift/mult pair for the scaled math
  * operations of clocksources and clockevents.
@@ -122,7 +122,7 @@ EXPORT_SYMBOL_GPL(timecounter_cyc2time);
  * NSEC_PER_SEC == 1GHz and @from is the counter frequency. For clock
  * event @to is the counter frequency and @from is NSEC_PER_SEC.
  *
- * The @minsec conversion range argument controls the time frame in
+ * The @maxsec conversion range argument controls the time frame in
  * seconds which must be covered by the runtime conversion with the
  * calculated mult and shift factors. This guarantees that no 64bit
  * overflow happens when the input value of the conversion is
@@ -131,7 +131,7 @@ EXPORT_SYMBOL_GPL(timecounter_cyc2time);
  * factors.
  */
 void
-clocks_calc_mult_shift(u32 *mult, u32 *shift, u32 from, u32 to, u32 minsec)
+clocks_calc_mult_shift(u32 *mult, u32 *shift, u32 from, u32 to, u32 maxsec)
 {
 	u64 tmp;
 	u32 sft, sftacc= 32;
@@ -140,7 +140,7 @@ clocks_calc_mult_shift(u32 *mult, u32 *shift, u32 from, u32 to, u32 minsec)
 	 * Calculate the shift factor which is limiting the conversion
 	 * range:
 	 */
-	tmp = ((u64)minsec * from) >> 32;
+	tmp = ((u64)maxsec * from) >> 32;
 	while (tmp) {
 		tmp >>=1;
 		sftacc--;
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 5bb86da..eef7452 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -49,7 +49,7 @@ struct timekeeper {
 	u32	mult;
 };
 
-struct timekeeper timekeeper;
+static struct timekeeper timekeeper;
 
 /**
  * timekeeper_setup_internals - Set up internals to use clocksource clock.
@@ -164,7 +164,7 @@ static struct timespec total_sleep_time;
 /*
  * The raw monotonic time for the CLOCK_MONOTONIC_RAW posix clock.
  */
-struct timespec raw_time;
+static struct timespec raw_time;
 
 /* flag for if timekeeping is suspended */
 int __read_mostly timekeeping_suspended;

[GIT PULL] timers fixes

[Ingo Molnar]

Linus,

Please pull the latest timers-fixes-for-linus git tree from:

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

 Thanks,

	Ingo

------------------>
Heiko Carstens (1):
      hrtimer: move timer stats helper functions to hrtimer.c

Thomas Gleixner (2):
      hrtimer: Tune hrtimer_interrupt hang logic
      itimer: Fix the itimer trace print format


 include/linux/hrtimer.h      |   56 +++----------------
 include/trace/events/timer.h |    8 ++--
 kernel/hrtimer.c             |  121 ++++++++++++++++++++++++++---------------
 kernel/time/timer_list.c     |    5 ++-
 4 files changed, 94 insertions(+), 96 deletions(-)

diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
index 9bace4b..af634e9 100644
--- a/include/linux/hrtimer.h
+++ b/include/linux/hrtimer.h
@@ -162,10 +162,11 @@ struct hrtimer_clock_base {
  * @expires_next:	absolute time of the next event which was scheduled
  *			via clock_set_next_event()
  * @hres_active:	State of high resolution mode
- * @check_clocks:	Indictator, when set evaluate time source and clock
- *			event devices whether high resolution mode can be
- *			activated.
- * @nr_events:		Total number of timer interrupt events
+ * @hang_detected:	The last hrtimer interrupt detected a hang
+ * @nr_events:		Total number of hrtimer interrupt events
+ * @nr_retries:		Total number of hrtimer interrupt retries
+ * @nr_hangs:		Total number of hrtimer interrupt hangs
+ * @max_hang_time:	Maximum time spent in hrtimer_interrupt
  */
 struct hrtimer_cpu_base {
 	spinlock_t			lock;
@@ -173,7 +174,11 @@ struct hrtimer_cpu_base {
 #ifdef CONFIG_HIGH_RES_TIMERS
 	ktime_t				expires_next;
 	int				hres_active;
+	int				hang_detected;
 	unsigned long			nr_events;
+	unsigned long			nr_retries;
+	unsigned long			nr_hangs;
+	ktime_t				max_hang_time;
 #endif
 };
 
@@ -435,47 +440,4 @@ extern u64 ktime_divns(const ktime_t kt, s64 div);
 /* Show pending timers: */
 extern void sysrq_timer_list_show(void);
 
-/*
- * Timer-statistics info:
- */
-#ifdef CONFIG_TIMER_STATS
-
-extern void timer_stats_update_stats(void *timer, pid_t pid, void *startf,
-				     void *timerf, char *comm,
-				     unsigned int timer_flag);
-
-static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
-{
-	if (likely(!timer_stats_active))
-		return;
-	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
-				 timer->function, timer->start_comm, 0);
-}
-
-extern void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer,
-						 void *addr);
-
-static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
-{
-	__timer_stats_hrtimer_set_start_info(timer, __builtin_return_address(0));
-}
-
-static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
-{
-	timer->start_site = NULL;
-}
-#else
-static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
-{
-}
-
-static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
-{
-}
-
-static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
-{
-}
-#endif
-
 #endif
diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h
index e5ce87a..9496b96 100644
--- a/include/trace/events/timer.h
+++ b/include/trace/events/timer.h
@@ -301,8 +301,8 @@ TRACE_EVENT(itimer_state,
 		__entry->interval_usec	= value->it_interval.tv_usec;
 	),
 
-	TP_printk("which=%d expires=%lu it_value=%lu.%lu it_interval=%lu.%lu",
-		  __entry->which, __entry->expires,
+	TP_printk("which=%d expires=%llu it_value=%ld.%ld it_interval=%ld.%ld",
+		  __entry->which, (unsigned long long)__entry->expires,
 		  __entry->value_sec, __entry->value_usec,
 		  __entry->interval_sec, __entry->interval_usec)
 );
@@ -331,8 +331,8 @@ TRACE_EVENT(itimer_expire,
 		__entry->pid	= pid_nr(pid);
 	),
 
-	    TP_printk("which=%d pid=%d now=%lu", __entry->which,
-		      (int) __entry->pid, __entry->now)
+	TP_printk("which=%d pid=%d now=%llu", __entry->which,
+		  (int) __entry->pid, (unsigned long long)__entry->now)
 );
 
 #endif /*  _TRACE_TIMER_H */
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
index ede5277..d2f9239 100644
--- a/kernel/hrtimer.c
+++ b/kernel/hrtimer.c
@@ -557,7 +557,7 @@ hrtimer_force_reprogram(struct hrtimer_cpu_base *cpu_base, int skip_equal)
 static int hrtimer_reprogram(struct hrtimer *timer,
 			     struct hrtimer_clock_base *base)
 {
-	ktime_t *expires_next = &__get_cpu_var(hrtimer_bases).expires_next;
+	struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
 	ktime_t expires = ktime_sub(hrtimer_get_expires(timer), base->offset);
 	int res;
 
@@ -582,7 +582,16 @@ static int hrtimer_reprogram(struct hrtimer *timer,
 	if (expires.tv64 < 0)
 		return -ETIME;
 
-	if (expires.tv64 >= expires_next->tv64)
+	if (expires.tv64 >= cpu_base->expires_next.tv64)
+		return 0;
+
+	/*
+	 * If a hang was detected in the last timer interrupt then we
+	 * do not schedule a timer which is earlier than the expiry
+	 * which we enforced in the hang detection. We want the system
+	 * to make progress.
+	 */
+	if (cpu_base->hang_detected)
 		return 0;
 
 	/*
@@ -590,7 +599,7 @@ static int hrtimer_reprogram(struct hrtimer *timer,
 	 */
 	res = tick_program_event(expires, 0);
 	if (!IS_ERR_VALUE(res))
-		*expires_next = expires;
+		cpu_base->expires_next = expires;
 	return res;
 }
 
@@ -747,17 +756,33 @@ static inline void hrtimer_init_timer_hres(struct hrtimer *timer) { }
 
 #endif /* CONFIG_HIGH_RES_TIMERS */
 
-#ifdef CONFIG_TIMER_STATS
-void __timer_stats_hrtimer_set_start_info(struct hrtimer *timer, void *addr)
+static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
 {
+#ifdef CONFIG_TIMER_STATS
 	if (timer->start_site)
 		return;
-
-	timer->start_site = addr;
+	timer->start_site = __builtin_return_address(0);
 	memcpy(timer->start_comm, current->comm, TASK_COMM_LEN);
 	timer->start_pid = current->pid;
+#endif
 }
+
+static inline void timer_stats_hrtimer_clear_start_info(struct hrtimer *timer)
+{
+#ifdef CONFIG_TIMER_STATS
+	timer->start_site = NULL;
+#endif
+}
+
+static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
+{
+#ifdef CONFIG_TIMER_STATS
+	if (likely(!timer_stats_active))
+		return;
+	timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
+				 timer->function, timer->start_comm, 0);
 #endif
+}
 
 /*
  * Counterpart to lock_hrtimer_base above:
@@ -1217,30 +1242,6 @@ static void __run_hrtimer(struct hrtimer *timer, ktime_t *now)
 
 #ifdef CONFIG_HIGH_RES_TIMERS
 
-static int force_clock_reprogram;
-
-/*
- * After 5 iteration's attempts, we consider that hrtimer_interrupt()
- * is hanging, which could happen with something that slows the interrupt
- * such as the tracing. Then we force the clock reprogramming for each future
- * hrtimer interrupts to avoid infinite loops and use the min_delta_ns
- * threshold that we will overwrite.
- * The next tick event will be scheduled to 3 times we currently spend on
- * hrtimer_interrupt(). This gives a good compromise, the cpus will spend
- * 1/4 of their time to process the hrtimer interrupts. This is enough to
- * let it running without serious starvation.
- */
-
-static inline void
-hrtimer_interrupt_hanging(struct clock_event_device *dev,
-			ktime_t try_time)
-{
-	force_clock_reprogram = 1;
-	dev->min_delta_ns = (unsigned long)try_time.tv64 * 3;
-	printk(KERN_WARNING "hrtimer: interrupt too slow, "
-	       "forcing clock min delta to %llu ns\n",
-	       (unsigned long long) dev->min_delta_ns);
-}
 /*
  * High resolution timer interrupt
  * Called with interrupts disabled
@@ -1249,21 +1250,15 @@ void hrtimer_interrupt(struct clock_event_device *dev)
 {
 	struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
 	struct hrtimer_clock_base *base;
-	ktime_t expires_next, now;
-	int nr_retries = 0;
-	int i;
+	ktime_t expires_next, now, entry_time, delta;
+	int i, retries = 0;
 
 	BUG_ON(!cpu_base->hres_active);
 	cpu_base->nr_events++;
 	dev->next_event.tv64 = KTIME_MAX;
 
- retry:
-	/* 5 retries is enough to notice a hang */
-	if (!(++nr_retries % 5))
-		hrtimer_interrupt_hanging(dev, ktime_sub(ktime_get(), now));
-
-	now = ktime_get();
-
+	entry_time = now = ktime_get();
+retry:
 	expires_next.tv64 = KTIME_MAX;
 
 	spin_lock(&cpu_base->lock);
@@ -1325,10 +1320,48 @@ void hrtimer_interrupt(struct clock_event_device *dev)
 	spin_unlock(&cpu_base->lock);
 
 	/* Reprogramming necessary ? */
-	if (expires_next.tv64 != KTIME_MAX) {
-		if (tick_program_event(expires_next, force_clock_reprogram))
-			goto retry;
+	if (expires_next.tv64 == KTIME_MAX ||
+	    !tick_program_event(expires_next, 0)) {
+		cpu_base->hang_detected = 0;
+		return;
 	}
+
+	/*
+	 * The next timer was already expired due to:
+	 * - tracing
+	 * - long lasting callbacks
+	 * - being scheduled away when running in a VM
+	 *
+	 * We need to prevent that we loop forever in the hrtimer
+	 * interrupt routine. We give it 3 attempts to avoid
+	 * overreacting on some spurious event.
+	 */
+	now = ktime_get();
+	cpu_base->nr_retries++;
+	if (++retries < 3)
+		goto retry;
+	/*
+	 * Give the system a chance to do something else than looping
+	 * here. We stored the entry time, so we know exactly how long
+	 * we spent here. We schedule the next event this amount of
+	 * time away.
+	 */
+	cpu_base->nr_hangs++;
+	cpu_base->hang_detected = 1;
+	delta = ktime_sub(now, entry_time);
+	if (delta.tv64 > cpu_base->max_hang_time.tv64)
+		cpu_base->max_hang_time = delta;
+	/*
+	 * Limit it to a sensible value as we enforce a longer
+	 * delay. Give the CPU at least 100ms to catch up.
+	 */
+	if (delta.tv64 > 100 * NSEC_PER_MSEC)
+		expires_next = ktime_add_ns(now, 100 * NSEC_PER_MSEC);
+	else
+		expires_next = ktime_add(now, delta);
+	tick_program_event(expires_next, 1);
+	printk_once(KERN_WARNING "hrtimer: interrupt took %llu ns\n",
+		    ktime_to_ns(delta));
 }
 
 /*
diff --git a/kernel/time/timer_list.c b/kernel/time/timer_list.c
index 665c76e..9d80db4 100644
--- a/kernel/time/timer_list.c
+++ b/kernel/time/timer_list.c
@@ -150,6 +150,9 @@ static void print_cpu(struct seq_file *m, int cpu, u64 now)
 	P_ns(expires_next);
 	P(hres_active);
 	P(nr_events);
+	P(nr_retries);
+	P(nr_hangs);
+	P_ns(max_hang_time);
 #endif
 #undef P
 #undef P_ns
@@ -254,7 +257,7 @@ static int timer_list_show(struct seq_file *m, void *v)
 	u64 now = ktime_to_ns(ktime_get());
 	int cpu;
 
-	SEQ_printf(m, "Timer List Version: v0.4\n");
+	SEQ_printf(m, "Timer List Version: v0.5\n");
 	SEQ_printf(m, "HRTIMER_MAX_CLOCK_BASES: %d\n", HRTIMER_MAX_CLOCK_BASES);
 	SEQ_printf(m, "now at %Ld nsecs\n", (unsigned long long)now);
 

[git pull] timers fixes

[Ingo Molnar]

Linus,

Please pull the latest timers-fixes-for-linus git tree from:

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

 Thanks,

	Ingo

------------------>
Thomas Gleixner (1):
      clockevents: check broadcast tick device not the clock events device


 kernel/time/tick-broadcast.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c
index bd70345..cb01cd8 100644
--- a/kernel/time/tick-broadcast.c
+++ b/kernel/time/tick-broadcast.c
@@ -235,7 +235,8 @@ static void tick_do_broadcast_on_off(void *why)
 	case CLOCK_EVT_NOTIFY_BROADCAST_FORCE:
 		if (!cpu_isset(cpu, tick_broadcast_mask)) {
 			cpu_set(cpu, tick_broadcast_mask);
-			if (bc->mode == TICKDEV_MODE_PERIODIC)
+			if (tick_broadcast_device.mode ==
+			    TICKDEV_MODE_PERIODIC)
 				clockevents_shutdown(dev);
 		}
 		if (*reason == CLOCK_EVT_NOTIFY_BROADCAST_FORCE)
@@ -245,7 +246,8 @@ static void tick_do_broadcast_on_off(void *why)
 		if (!tick_broadcast_force &&
 		    cpu_isset(cpu, tick_broadcast_mask)) {
 			cpu_clear(cpu, tick_broadcast_mask);
-			if (bc->mode == TICKDEV_MODE_PERIODIC)
+			if (tick_broadcast_device.mode ==
+			    TICKDEV_MODE_PERIODIC)
 				tick_setup_periodic(dev, 0);
 		}
 		break;

[PATCH] x86: Add missing idle.h include to process_32.c

[Rafael J. Wysocki]

Hi,

This fixes a very recent compilation breakage on 32-bit x86.

Please apply.

Thanks,
Rafael

---
From: Rafael J. Wysocki <rjw@sisk.pl>

x86: Add missing idle.h include to process_32.c

Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
---
 arch/x86/kernel/process_32.c |    1 +
 1 file changed, 1 insertion(+)

Index: linux-2.6/arch/x86/kernel/process_32.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/process_32.c
+++ linux-2.6/arch/x86/kernel/process_32.c
@@ -55,6 +55,7 @@
 #include <asm/tlbflush.h>
 #include <asm/cpu.h>
 #include <asm/kdebug.h>
+#include <asm/idle.h>
 
 asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
 

Re: [PATCH] x86: Add missing idle.h include to process_32.c

[Ingo Molnar]

* Rafael J. Wysocki <rjw@sisk.pl> wrote:

> Hi,
> 
> This fixes a very recent compilation breakage on 32-bit x86.

the fix is queued up in tip/timers/urgent - sorry about that.

	Ingo

[git pull] timers fixes

[Ingo Molnar]

Linus,

Please pull the latest timers-fixes-for-linus git tree from:

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

Thanks,

	Ingo

------------------>
Marc Dionne (1):
      x86: prevent stale state of c1e_mask across CPU offline/online, fix


 arch/x86/kernel/process_32.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
index 4b3cfdf..31f40b2 100644
--- a/arch/x86/kernel/process_32.c
+++ b/arch/x86/kernel/process_32.c
@@ -55,6 +55,7 @@
 #include <asm/tlbflush.h>
 #include <asm/cpu.h>
 #include <asm/kdebug.h>
+#include <asm/idle.h>
 
 asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
 

Re: [git pull] timers fixes

[Benjamin Herrenschmidt]

On Wed, 2008-09-24 at 09:42 +0200, Ingo Molnar wrote:
> Linus,
> 
> Please pull the latest timers-fixes-for-linus git tree from:
> 
>    git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git timers-fixes-for-linus
> 
> Thanks,
> 
> 	Ingo
> 
> ------------------>
> Marc Dionne (1):
>       x86: prevent stale state of c1e_mask across CPU offline/online, fix

Is that all the changeset comment ? Am I the only one to find that
totally useless ? Or is that the mail doesn't contain the actual
comment, only the title followed by the patch ?

Cheers,
Ben.

> 
>  arch/x86/kernel/process_32.c |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
> index 4b3cfdf..31f40b2 100644
> --- a/arch/x86/kernel/process_32.c
> +++ b/arch/x86/kernel/process_32.c
> @@ -55,6 +55,7 @@
>  #include <asm/tlbflush.h>
>  #include <asm/cpu.h>
>  #include <asm/kdebug.h>
> +#include <asm/idle.h>
>  
>  asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
>  
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

Re: [git pull] timers fixes

[Benjamin Herrenschmidt]

On Wed, 2008-09-24 at 18:49 +1000, Benjamin Herrenschmidt wrote:
> On Wed, 2008-09-24 at 09:42 +0200, Ingo Molnar wrote:
> > Linus,
> > 
> > Please pull the latest timers-fixes-for-linus git tree from:
> > 
> >    git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git timers-fixes-for-linus
> > 
> > Thanks,
> > 
> > 	Ingo
> > 
> > ------------------>
> > Marc Dionne (1):
> >       x86: prevent stale state of c1e_mask across CPU offline/online, fix
> 
> Is that all the changeset comment ? Am I the only one to find that
> totally useless ? Or is that the mail doesn't contain the actual
> comment, only the title followed by the patch ?

Allright, so in fact, it's more like the log doesn't match
the patch in the email..

Ben.

> Cheers,
> Ben.
> 
> > 
> >  arch/x86/kernel/process_32.c |    1 +
> >  1 files changed, 1 insertions(+), 0 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
> > index 4b3cfdf..31f40b2 100644
> > --- a/arch/x86/kernel/process_32.c
> > +++ b/arch/x86/kernel/process_32.c
> > @@ -55,6 +55,7 @@
> >  #include <asm/tlbflush.h>
> >  #include <asm/cpu.h>
> >  #include <asm/kdebug.h>
> > +#include <asm/idle.h>
> >  
> >  asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
> >  
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/

Re: [git pull] timers fixes

[Ingo Molnar]

* Benjamin Herrenschmidt <benh@kernel.crashing.org> wrote:

> On Wed, 2008-09-24 at 18:49 +1000, Benjamin Herrenschmidt wrote:
> > On Wed, 2008-09-24 at 09:42 +0200, Ingo Molnar wrote:
> > > Linus,
> > > 
> > > Please pull the latest timers-fixes-for-linus git tree from:
> > > 
> > >    git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git timers-fixes-for-linus
> > > 
> > > Thanks,
> > > 
> > > 	Ingo
> > > 
> > > ------------------>
> > > Marc Dionne (1):
> > >       x86: prevent stale state of c1e_mask across CPU offline/online, fix
> > 
> > Is that all the changeset comment ? Am I the only one to find that
> > totally useless ? Or is that the mail doesn't contain the actual
> > comment, only the title followed by the patch ?
> 
> Allright, so in fact, it's more like the log doesn't match
> the patch in the email..

only because i already picked up a different submission at that time, 
and posted in this thread.

i picked up this one from lkml:

  Subject: [PATCH] Fix build error introduced by commit 4faac97d44ac27bdbb010a9c3597401a8f89341f

see the full commit log below, it has all the details.

	Ingo

------------>
>From 1eda81495a49a4ee91d8863b0a441a624375efea Mon Sep 17 00:00:00 2001
From: Marc Dionne <marc.c.dionne@gmail.com>
Date: Tue, 23 Sep 2008 22:40:02 -0400
Subject: [PATCH] x86: prevent stale state of c1e_mask across CPU offline/online, fix

Fix build error introduced by commit 4faac97d44ac27 ("x86: prevent stale
state of c1e_mask across CPU offline/online").

process_32.c needs to include idle.h to get the prototype for
c1e_remove_cpu()

Signed-off-by: Marc Dionne <marc.c.dionne@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/kernel/process_32.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
index 4b3cfdf..31f40b2 100644
--- a/arch/x86/kernel/process_32.c
+++ b/arch/x86/kernel/process_32.c
@@ -55,6 +55,7 @@
 #include <asm/tlbflush.h>
 #include <asm/cpu.h>
 #include <asm/kdebug.h>
+#include <asm/idle.h>
 
 asmlinkage void ret_from_fork(void) __asm__("ret_from_fork");
 

Re: [git pull] timers fixes

[Ingo Molnar]

* Benjamin Herrenschmidt <benh@kernel.crashing.org> wrote:

> On Wed, 2008-09-24 at 09:42 +0200, Ingo Molnar wrote:
> > Linus,
> > 
> > Please pull the latest timers-fixes-for-linus git tree from:
> > 
> >    git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git timers-fixes-for-linus
> > 
> > Thanks,
> > 
> > 	Ingo
> > 
> > ------------------>
> > Marc Dionne (1):
> >       x86: prevent stale state of c1e_mask across CPU offline/online, fix
> 
> Is that all the changeset comment ? Am I the only one to find that 
> totally useless ? Or is that the mail doesn't contain the actual 
> comment, only the title followed by the patch ?

yeah. The build error is something like:

 arch/x86/kernel/process_32.c:78: warning: data definition has no type or storage class
 arch/x86/kernel/process_32.c:78: warning: type defaults to 'int' in declaration of 'c1e_remove_cpu'
 arch/x86/kernel/process_32.c:78: warning: parameter names (without types) in function declaration

	Ingo