From 80a05b9ffa7dc13f6693902dd8999a2b61a3a0d7 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 12 Mar 2010 17:34:14 +0100 Subject: clockevents: Sanitize min_delta_ns adjustment and prevent overflows The current logic which handles clock events programming failures can increase min_delta_ns unlimited and even can cause overflows. Sanitize it by: - prevent zero increase when min_delta_ns == 1 - limiting min_delta_ns to a jiffie - bail out if the jiffie limit is hit - add retries stats for /proc/timer_list so we can gather data Reported-by: Uwe Kleine-Koenig Signed-off-by: Thomas Gleixner --- kernel/time/tick-oneshot.c | 52 +++++++++++++++++++++++++++++++++++----------- kernel/time/timer_list.c | 3 ++- 2 files changed, 42 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/time/tick-oneshot.c b/kernel/time/tick-oneshot.c index 0a8a213016f0..aada0e52680a 100644 --- a/kernel/time/tick-oneshot.c +++ b/kernel/time/tick-oneshot.c @@ -22,6 +22,29 @@ #include "tick-internal.h" +/* Limit min_delta to a jiffie */ +#define MIN_DELTA_LIMIT (NSEC_PER_SEC / HZ) + +static int tick_increase_min_delta(struct clock_event_device *dev) +{ + /* Nothing to do if we already reached the limit */ + if (dev->min_delta_ns >= MIN_DELTA_LIMIT) + return -ETIME; + + if (dev->min_delta_ns < 5000) + dev->min_delta_ns = 5000; + else + dev->min_delta_ns += dev->min_delta_ns >> 1; + + if (dev->min_delta_ns > MIN_DELTA_LIMIT) + dev->min_delta_ns = MIN_DELTA_LIMIT; + + printk(KERN_WARNING "CE: %s increased min_delta_ns to %llu nsec\n", + dev->name ? dev->name : "?", + (unsigned long long) dev->min_delta_ns); + return 0; +} + /** * tick_program_event internal worker function */ @@ -37,23 +60,28 @@ int tick_dev_program_event(struct clock_event_device *dev, ktime_t expires, if (!ret || !force) return ret; + dev->retries++; /* - * We tried 2 times to program the device with the given - * min_delta_ns. If that's not working then we double it + * We tried 3 times to program the device with the given + * min_delta_ns. If that's not working then we increase it * and emit a warning. */ if (++i > 2) { /* Increase the min. delta and try again */ - if (!dev->min_delta_ns) - dev->min_delta_ns = 5000; - else - dev->min_delta_ns += dev->min_delta_ns >> 1; - - printk(KERN_WARNING - "CE: %s increasing min_delta_ns to %llu nsec\n", - dev->name ? dev->name : "?", - (unsigned long long) dev->min_delta_ns << 1); - + if (tick_increase_min_delta(dev)) { + /* + * Get out of the loop if min_delta_ns + * hit the limit already. That's + * better than staying here forever. + * + * We clear next_event so we have a + * chance that the box survives. + */ + printk(KERN_WARNING + "CE: Reprogramming failure. Giving up\n"); + dev->next_event.tv64 = KTIME_MAX; + return -ETIME; + } i = 0; } diff --git a/kernel/time/timer_list.c b/kernel/time/timer_list.c index bdfb8dd1050c..1a4a7dd78777 100644 --- a/kernel/time/timer_list.c +++ b/kernel/time/timer_list.c @@ -228,6 +228,7 @@ print_tickdevice(struct seq_file *m, struct tick_device *td, int cpu) SEQ_printf(m, " event_handler: "); print_name_offset(m, dev->event_handler); SEQ_printf(m, "\n"); + SEQ_printf(m, " retries: %lu\n", dev->retries); } static void timer_list_show_tickdevices(struct seq_file *m) @@ -257,7 +258,7 @@ 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: v0.5\n"); + SEQ_printf(m, "Timer List Version: v0.6\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); -- cgit v1.2.3 From 829b6c1ef488856c6a46a2f705f5068062d5f34c Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 11 Mar 2010 14:04:30 -0800 Subject: timer stats: Fix del_timer_sync() and try_to_del_timer_sync() These functions forgot to run timer_stats_timer_clear_start_info(). It's unobvious what effect this has and whether it matters much - we won't be printing it out anyway if the timer's detached. Untested, just an Ingo trollpatch. [ Nevertheless correct - tglx ] Signed-off-by: Andrew Morton Cc: Ingo Molnar Cc: johnstul@us.ibm.com Signed-off-by: Thomas Gleixner --- kernel/timer.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel') diff --git a/kernel/timer.c b/kernel/timer.c index c61a7949387f..fc965eae0e87 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -880,6 +880,7 @@ int try_to_del_timer_sync(struct timer_list *timer) if (base->running_timer == timer) goto out; + timer_stats_timer_clear_start_info(timer); ret = 0; if (timer_pending(timer)) { detach_timer(timer, 1); -- cgit v1.2.3 From 15365c108ea27598e265f8c13e7051d99ca5b0b9 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:31 -0800 Subject: posix-cpu-timers: Reset expire cache when no timer is running When a process deletes cpu timer or a timer expires we do not clear the expiration cache sig->cputimer_expires. As a result the fastpath_timer_check() which prevents us to loop over all threads in case no timer is active is not working and we run the slow path needlessly on every tick. Zero sig->cputimer_expires in stop_process_timers(). Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Spencer Candland Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner --- kernel/posix-cpu-timers.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 438ff4523513..edec25a68e9e 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -1060,9 +1060,9 @@ static void check_thread_timers(struct task_struct *tsk, } } -static void stop_process_timers(struct task_struct *tsk) +static void stop_process_timers(struct signal_struct *sig) { - struct thread_group_cputimer *cputimer = &tsk->signal->cputimer; + struct thread_group_cputimer *cputimer = &sig->cputimer; unsigned long flags; if (!cputimer->running) @@ -1071,6 +1071,10 @@ static void stop_process_timers(struct task_struct *tsk) spin_lock_irqsave(&cputimer->lock, flags); cputimer->running = 0; spin_unlock_irqrestore(&cputimer->lock, flags); + + sig->cputime_expires.prof_exp = cputime_zero; + sig->cputime_expires.virt_exp = cputime_zero; + sig->cputime_expires.sched_exp = 0; } static u32 onecputick; @@ -1131,7 +1135,7 @@ static void check_process_timers(struct task_struct *tsk, list_empty(&timers[CPUCLOCK_VIRT]) && cputime_eq(sig->it[CPUCLOCK_VIRT].expires, cputime_zero) && list_empty(&timers[CPUCLOCK_SCHED])) { - stop_process_timers(tsk); + stop_process_timers(sig); return; } -- cgit v1.2.3 From 830ec0458c390f29c6c99e1ff7feab9e36368d12 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 18 Mar 2010 14:47:30 -0700 Subject: time: Fix accumulation bug triggered by long delay. The logarithmic accumulation done in the timekeeping has some overflow protection that limits the max shift value. That means it will take more then shift loops to accumulate all of the cycles. This causes the shift decrement to underflow, which causes the loop to never exit. The simplest fix would be simply to do a: if (shift) shift--; However that is not optimal, as we know the cycle offset is larger then the interval << shift, the above would make shift drop to zero, then we would be spinning for quite awhile accumulating at interval chunks at a time. Instead, this patch only decreases shift if the offset is smaller then cycle_interval << shift. This makes sure we accumulate using the largest chunks possible without overflowing tick_length, and limits the number of iterations through the loop. This issue was found and reported by Sonic Zhang, who also tested the fix. Many thanks your explanation and testing! Reported-by: Sonic Zhang Signed-off-by: John Stultz Tested-by: Sonic Zhang LKML-Reference: <1268948850-5225-1-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner --- kernel/time/timekeeping.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 16736379a9ca..39f6177fafac 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -818,7 +818,8 @@ void update_wall_time(void) shift = min(shift, maxshift); while (offset >= timekeeper.cycle_interval) { offset = logarithmic_accumulation(offset, shift); - shift--; + if(offset < timekeeper.cycle_interval<