X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=kernel%2Ftimer.c;h=db3377bb236d203ec553c62f85dba9a7fa085158;hb=refs%2Fheads%2Fvserver;hp=c64ba1995955fad550147f6260c3c068e7ed5c80;hpb=76828883507a47dae78837ab5dec5a5b4513c667;p=linux-2.6.git diff --git a/kernel/timer.c b/kernel/timer.c index c64ba1995..db3377bb2 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -34,7 +34,9 @@ #include #include #include +#include #include +#include #include #include @@ -43,12 +45,6 @@ #include #include -#ifdef CONFIG_TIME_INTERPOLATION -static void time_interpolator_update(long delta_nsec); -#else -#define time_interpolator_update(x) -#endif - u64 jiffies_64 __cacheline_aligned_in_smp = INITIAL_JIFFIES; EXPORT_SYMBOL(jiffies_64); @@ -56,7 +52,6 @@ EXPORT_SYMBOL(jiffies_64); /* * per-CPU timer vector definitions: */ - #define TVN_BITS (CONFIG_BASE_SMALL ? 4 : 6) #define TVR_BITS (CONFIG_BASE_SMALL ? 6 : 8) #define TVN_SIZE (1 << TVN_BITS) @@ -64,11 +59,6 @@ EXPORT_SYMBOL(jiffies_64); #define TVN_MASK (TVN_SIZE - 1) #define TVR_MASK (TVR_SIZE - 1) -struct timer_base_s { - spinlock_t lock; - struct timer_list *running_timer; -}; - typedef struct tvec_s { struct list_head vec[TVN_SIZE]; } tvec_t; @@ -78,7 +68,8 @@ typedef struct tvec_root_s { } tvec_root_t; struct tvec_t_base_s { - struct timer_base_s t_base; + spinlock_t lock; + struct timer_list *running_timer; unsigned long timer_jiffies; tvec_root_t tv1; tvec_t tv2; @@ -88,13 +79,148 @@ struct tvec_t_base_s { } ____cacheline_aligned_in_smp; typedef struct tvec_t_base_s tvec_base_t; -static DEFINE_PER_CPU(tvec_base_t, tvec_bases); + +tvec_base_t boot_tvec_bases; +EXPORT_SYMBOL(boot_tvec_bases); +static DEFINE_PER_CPU(tvec_base_t *, tvec_bases) = &boot_tvec_bases; + +/** + * __round_jiffies - function to round jiffies to a full second + * @j: the time in (absolute) jiffies that should be rounded + * @cpu: the processor number on which the timeout will happen + * + * __round_jiffies rounds an absolute time in the future (in jiffies) + * up or down to (approximately) full seconds. This is useful for timers + * for which the exact time they fire does not matter too much, as long as + * they fire approximately every X seconds. + * + * By rounding these timers to whole seconds, all such timers will fire + * at the same time, rather than at various times spread out. The goal + * of this is to have the CPU wake up less, which saves power. + * + * The exact rounding is skewed for each processor to avoid all + * processors firing at the exact same time, which could lead + * to lock contention or spurious cache line bouncing. + * + * The return value is the rounded version of the "j" parameter. + */ +unsigned long __round_jiffies(unsigned long j, int cpu) +{ + int rem; + unsigned long original = j; + + /* + * We don't want all cpus firing their timers at once hitting the + * same lock or cachelines, so we skew each extra cpu with an extra + * 3 jiffies. This 3 jiffies came originally from the mm/ code which + * already did this. + * The skew is done by adding 3*cpunr, then round, then subtract this + * extra offset again. + */ + j += cpu * 3; + + rem = j % HZ; + + /* + * If the target jiffie is just after a whole second (which can happen + * due to delays of the timer irq, long irq off times etc etc) then + * we should round down to the whole second, not up. Use 1/4th second + * as cutoff for this rounding as an extreme upper bound for this. + */ + if (rem < HZ/4) /* round down */ + j = j - rem; + else /* round up */ + j = j - rem + HZ; + + /* now that we have rounded, subtract the extra skew again */ + j -= cpu * 3; + + if (j <= jiffies) /* rounding ate our timeout entirely; */ + return original; + return j; +} +EXPORT_SYMBOL_GPL(__round_jiffies); + +/** + * __round_jiffies_relative - function to round jiffies to a full second + * @j: the time in (relative) jiffies that should be rounded + * @cpu: the processor number on which the timeout will happen + * + * __round_jiffies_relative rounds a time delta in the future (in jiffies) + * up or down to (approximately) full seconds. This is useful for timers + * for which the exact time they fire does not matter too much, as long as + * they fire approximately every X seconds. + * + * By rounding these timers to whole seconds, all such timers will fire + * at the same time, rather than at various times spread out. The goal + * of this is to have the CPU wake up less, which saves power. + * + * The exact rounding is skewed for each processor to avoid all + * processors firing at the exact same time, which could lead + * to lock contention or spurious cache line bouncing. + * + * The return value is the rounded version of the "j" parameter. + */ +unsigned long __round_jiffies_relative(unsigned long j, int cpu) +{ + /* + * In theory the following code can skip a jiffy in case jiffies + * increments right between the addition and the later subtraction. + * However since the entire point of this function is to use approximate + * timeouts, it's entirely ok to not handle that. + */ + return __round_jiffies(j + jiffies, cpu) - jiffies; +} +EXPORT_SYMBOL_GPL(__round_jiffies_relative); + +/** + * round_jiffies - function to round jiffies to a full second + * @j: the time in (absolute) jiffies that should be rounded + * + * round_jiffies rounds an absolute time in the future (in jiffies) + * up or down to (approximately) full seconds. This is useful for timers + * for which the exact time they fire does not matter too much, as long as + * they fire approximately every X seconds. + * + * By rounding these timers to whole seconds, all such timers will fire + * at the same time, rather than at various times spread out. The goal + * of this is to have the CPU wake up less, which saves power. + * + * The return value is the rounded version of the "j" parameter. + */ +unsigned long round_jiffies(unsigned long j) +{ + return __round_jiffies(j, raw_smp_processor_id()); +} +EXPORT_SYMBOL_GPL(round_jiffies); + +/** + * round_jiffies_relative - function to round jiffies to a full second + * @j: the time in (relative) jiffies that should be rounded + * + * round_jiffies_relative rounds a time delta in the future (in jiffies) + * up or down to (approximately) full seconds. This is useful for timers + * for which the exact time they fire does not matter too much, as long as + * they fire approximately every X seconds. + * + * By rounding these timers to whole seconds, all such timers will fire + * at the same time, rather than at various times spread out. The goal + * of this is to have the CPU wake up less, which saves power. + * + * The return value is the rounded version of the "j" parameter. + */ +unsigned long round_jiffies_relative(unsigned long j) +{ + return __round_jiffies_relative(j, raw_smp_processor_id()); +} +EXPORT_SYMBOL_GPL(round_jiffies_relative); + static inline void set_running_timer(tvec_base_t *base, struct timer_list *timer) { #ifdef CONFIG_SMP - base->t_base.running_timer = timer; + base->running_timer = timer; #endif } @@ -140,16 +266,7 @@ static void internal_add_timer(tvec_base_t *base, struct timer_list *timer) list_add_tail(&timer->entry, vec); } -typedef struct timer_base_s timer_base_t; -/* - * Used by TIMER_INITIALIZER, we can't use per_cpu(tvec_bases) - * at compile time, and we need timer->base to lock the timer. - */ -timer_base_t __init_timer_base - ____cacheline_aligned_in_smp = { .lock = SPIN_LOCK_UNLOCKED }; -EXPORT_SYMBOL(__init_timer_base); - -/*** +/** * init_timer - initialize a timer. * @timer: the timer to be initialized * @@ -159,7 +276,7 @@ EXPORT_SYMBOL(__init_timer_base); void fastcall init_timer(struct timer_list *timer) { timer->entry.next = NULL; - timer->base = &per_cpu(tvec_bases, raw_smp_processor_id()).t_base; + timer->base = __raw_get_cpu_var(tvec_bases); } EXPORT_SYMBOL(init_timer); @@ -175,7 +292,7 @@ static inline void detach_timer(struct timer_list *timer, } /* - * We are using hashed locking: holding per_cpu(tvec_bases).t_base.lock + * We are using hashed locking: holding per_cpu(tvec_bases).lock * means that all timers which are tied to this base via timer->base are * locked, and the base itself is locked too. * @@ -186,10 +303,11 @@ static inline void detach_timer(struct timer_list *timer, * possible to set timer->base = NULL and drop the lock: the timer remains * locked. */ -static timer_base_t *lock_timer_base(struct timer_list *timer, +static tvec_base_t *lock_timer_base(struct timer_list *timer, unsigned long *flags) + __acquires(timer->base->lock) { - timer_base_t *base; + tvec_base_t *base; for (;;) { base = timer->base; @@ -206,8 +324,7 @@ static timer_base_t *lock_timer_base(struct timer_list *timer, int __mod_timer(struct timer_list *timer, unsigned long expires) { - timer_base_t *base; - tvec_base_t *new_base; + tvec_base_t *base, *new_base; unsigned long flags; int ret = 0; @@ -220,9 +337,9 @@ int __mod_timer(struct timer_list *timer, unsigned long expires) ret = 1; } - new_base = &__get_cpu_var(tvec_bases); + new_base = __get_cpu_var(tvec_bases); - if (base != &new_base->t_base) { + if (base != new_base) { /* * We are trying to schedule the timer on the local CPU. * However we can't change timer's base while it is running, @@ -230,28 +347,26 @@ int __mod_timer(struct timer_list *timer, unsigned long expires) * handler yet has not finished. This also guarantees that * the timer is serialized wrt itself. */ - if (unlikely(base->running_timer == timer)) { - /* The timer remains on a former base */ - new_base = container_of(base, tvec_base_t, t_base); - } else { + if (likely(base->running_timer != timer)) { /* See the comment in lock_timer_base() */ timer->base = NULL; spin_unlock(&base->lock); - spin_lock(&new_base->t_base.lock); - timer->base = &new_base->t_base; + base = new_base; + spin_lock(&base->lock); + timer->base = base; } } timer->expires = expires; - internal_add_timer(new_base, timer); - spin_unlock_irqrestore(&new_base->t_base.lock, flags); + internal_add_timer(base, timer); + spin_unlock_irqrestore(&base->lock, flags); return ret; } EXPORT_SYMBOL(__mod_timer); -/*** +/** * add_timer_on - start a timer on a particular CPU * @timer: the timer to be added * @cpu: the CPU to start it on @@ -260,20 +375,21 @@ EXPORT_SYMBOL(__mod_timer); */ void add_timer_on(struct timer_list *timer, int cpu) { - tvec_base_t *base = &per_cpu(tvec_bases, cpu); + tvec_base_t *base = per_cpu(tvec_bases, cpu); unsigned long flags; BUG_ON(timer_pending(timer) || !timer->function); - spin_lock_irqsave(&base->t_base.lock, flags); - timer->base = &base->t_base; + spin_lock_irqsave(&base->lock, flags); + timer->base = base; internal_add_timer(base, timer); - spin_unlock_irqrestore(&base->t_base.lock, flags); + spin_unlock_irqrestore(&base->lock, flags); } -/*** +/** * mod_timer - modify a timer's timeout * @timer: the timer to be modified + * @expires: new timeout in jiffies * * mod_timer is a more efficient way to update the expire field of an * active timer (if the timer is inactive it will be activated) @@ -307,7 +423,7 @@ int mod_timer(struct timer_list *timer, unsigned long expires) EXPORT_SYMBOL(mod_timer); -/*** +/** * del_timer - deactive a timer. * @timer: the timer to be deactivated * @@ -320,7 +436,7 @@ EXPORT_SYMBOL(mod_timer); */ int del_timer(struct timer_list *timer) { - timer_base_t *base; + tvec_base_t *base; unsigned long flags; int ret = 0; @@ -339,7 +455,10 @@ int del_timer(struct timer_list *timer) EXPORT_SYMBOL(del_timer); #ifdef CONFIG_SMP -/* +/** + * try_to_del_timer_sync - Try to deactivate a timer + * @timer: timer do del + * * This function tries to deactivate a timer. Upon successful (ret >= 0) * exit the timer is not queued and the handler is not running on any CPU. * @@ -347,7 +466,7 @@ EXPORT_SYMBOL(del_timer); */ int try_to_del_timer_sync(struct timer_list *timer) { - timer_base_t *base; + tvec_base_t *base; unsigned long flags; int ret = -1; @@ -367,7 +486,7 @@ out: return ret; } -/*** +/** * del_timer_sync - deactivate a timer and wait for the handler to finish. * @timer: the timer to be deactivated * @@ -390,6 +509,7 @@ int del_timer_sync(struct timer_list *timer) int ret = try_to_del_timer_sync(timer); if (ret >= 0) return ret; + cpu_relax(); } } @@ -399,46 +519,42 @@ EXPORT_SYMBOL(del_timer_sync); static int cascade(tvec_base_t *base, tvec_t *tv, int index) { /* cascade all the timers from tv up one level */ - struct list_head *head, *curr; + struct timer_list *timer, *tmp; + struct list_head tv_list; + + list_replace_init(tv->vec + index, &tv_list); - head = tv->vec + index; - curr = head->next; /* - * We are removing _all_ timers from the list, so we don't have to - * detach them individually, just clear the list afterwards. + * We are removing _all_ timers from the list, so we + * don't have to detach them individually. */ - while (curr != head) { - struct timer_list *tmp; - - tmp = list_entry(curr, struct timer_list, entry); - BUG_ON(tmp->base != &base->t_base); - curr = curr->next; - internal_add_timer(base, tmp); + list_for_each_entry_safe(timer, tmp, &tv_list, entry) { + BUG_ON(timer->base != base); + internal_add_timer(base, timer); } - INIT_LIST_HEAD(head); return index; } -/*** +#define INDEX(N) ((base->timer_jiffies >> (TVR_BITS + (N) * TVN_BITS)) & TVN_MASK) + +/** * __run_timers - run all expired timers (if any) on this CPU. * @base: the timer vector to be processed. * * This function cascades all vectors and executes all expired timer * vectors. */ -#define INDEX(N) (base->timer_jiffies >> (TVR_BITS + N * TVN_BITS)) & TVN_MASK - static inline void __run_timers(tvec_base_t *base) { struct timer_list *timer; - spin_lock_irq(&base->t_base.lock); + spin_lock_irq(&base->lock); while (time_after_eq(jiffies, base->timer_jiffies)) { - struct list_head work_list = LIST_HEAD_INIT(work_list); + struct list_head work_list; struct list_head *head = &work_list; int index = base->timer_jiffies & TVR_MASK; - + /* * Cascade timers: */ @@ -447,8 +563,8 @@ static inline void __run_timers(tvec_base_t *base) (!cascade(base, &base->tv3, INDEX(1))) && !cascade(base, &base->tv4, INDEX(2))) cascade(base, &base->tv5, INDEX(3)); - ++base->timer_jiffies; - list_splice_init(base->tv1.vec + index, &work_list); + ++base->timer_jiffies; + list_replace_init(base->tv1.vec + index, &work_list); while (!list_empty(head)) { void (*fn)(unsigned long); unsigned long data; @@ -459,7 +575,7 @@ static inline void __run_timers(tvec_base_t *base) set_running_timer(base, timer); detach_timer(timer, 1); - spin_unlock_irq(&base->t_base.lock); + spin_unlock_irq(&base->lock); { int preempt_count = preempt_count(); fn(data); @@ -472,11 +588,11 @@ static inline void __run_timers(tvec_base_t *base) BUG(); } } - spin_lock_irq(&base->t_base.lock); + spin_lock_irq(&base->lock); } } set_running_timer(base, NULL); - spin_unlock_irq(&base->t_base.lock); + spin_unlock_irq(&base->lock); } #ifdef CONFIG_NO_IDLE_HZ @@ -506,8 +622,8 @@ unsigned long next_timer_interrupt(void) } hr_expires += jiffies; - base = &__get_cpu_var(tvec_bases); - spin_lock(&base->t_base.lock); + base = __get_cpu_var(tvec_bases); + spin_lock(&base->lock); expires = base->timer_jiffies + (LONG_MAX >> 1); list = NULL; @@ -555,7 +671,23 @@ found: expires = nte->expires; } } - spin_unlock(&base->t_base.lock); + spin_unlock(&base->lock); + + /* + * It can happen that other CPUs service timer IRQs and increment + * jiffies, but we have not yet got a local timer tick to process + * the timer wheels. In that case, the expiry time can be before + * jiffies, but since the high-resolution timer here is relative to + * jiffies, the default expression when high-resolution timers are + * not active, + * + * time_before(MAX_JIFFY_OFFSET + jiffies, expires) + * + * would falsely evaluate to true. If that is the case, just + * return jiffies so that we can immediately fire the local timer + */ + if (time_before(expires, jiffies)) + return jiffies; if (time_before(hr_expires, expires)) return hr_expires; @@ -566,12 +698,6 @@ found: /******************************************************************/ -/* - * Timekeeping variables - */ -unsigned long tick_usec = TICK_USEC; /* USER_HZ period (usec) */ -unsigned long tick_nsec = TICK_NSEC; /* ACTHZ period (nsec) */ - /* * The current time * wall_to_monotonic is what we need to add to xtime (or xtime corrected @@ -585,244 +711,393 @@ struct timespec wall_to_monotonic __attribute__ ((aligned (16))); EXPORT_SYMBOL(xtime); -/* Don't completely fail for HZ > 500. */ -int tickadj = 500/HZ ? : 1; /* microsecs */ +/* XXX - all of this timekeeping code should be later moved to time.c */ +#include +static struct clocksource *clock; /* pointer to current clocksource */ -/* - * phase-lock loop variables +#ifdef CONFIG_GENERIC_TIME +/** + * __get_nsec_offset - Returns nanoseconds since last call to periodic_hook + * + * private function, must hold xtime_lock lock when being + * called. Returns the number of nanoseconds since the + * last call to update_wall_time() (adjusted by NTP scaling) */ -/* TIME_ERROR prevents overwriting the CMOS clock */ -int time_state = TIME_OK; /* clock synchronization status */ -int time_status = STA_UNSYNC; /* clock status bits */ -long time_offset; /* time adjustment (us) */ -long time_constant = 2; /* pll time constant */ -long time_tolerance = MAXFREQ; /* frequency tolerance (ppm) */ -long time_precision = 1; /* clock precision (us) */ -long time_maxerror = NTP_PHASE_LIMIT; /* maximum error (us) */ -long time_esterror = NTP_PHASE_LIMIT; /* estimated error (us) */ -static long time_phase; /* phase offset (scaled us) */ -long time_freq = (((NSEC_PER_SEC + HZ/2) % HZ - HZ/2) << SHIFT_USEC) / NSEC_PER_USEC; - /* frequency offset (scaled ppm)*/ -static long time_adj; /* tick adjust (scaled 1 / HZ) */ -long time_reftime; /* time at last adjustment (s) */ -long time_adjust; -long time_next_adjust; +static inline s64 __get_nsec_offset(void) +{ + cycle_t cycle_now, cycle_delta; + s64 ns_offset; -/* - * this routine handles the overflow of the microsecond field + /* read clocksource: */ + cycle_now = clocksource_read(clock); + + /* calculate the delta since the last update_wall_time: */ + cycle_delta = (cycle_now - clock->cycle_last) & clock->mask; + + /* convert to nanoseconds: */ + ns_offset = cyc2ns(clock, cycle_delta); + + return ns_offset; +} + +/** + * __get_realtime_clock_ts - Returns the time of day in a timespec + * @ts: pointer to the timespec to be set * - * The tricky bits of code to handle the accurate clock support - * were provided by Dave Mills (Mills@UDEL.EDU) of NTP fame. - * They were originally developed for SUN and DEC kernels. - * All the kudos should go to Dave for this stuff. + * Returns the time of day in a timespec. Used by + * do_gettimeofday() and get_realtime_clock_ts(). + */ +static inline void __get_realtime_clock_ts(struct timespec *ts) +{ + unsigned long seq; + s64 nsecs; + + do { + seq = read_seqbegin(&xtime_lock); + + *ts = xtime; + nsecs = __get_nsec_offset(); + + } while (read_seqretry(&xtime_lock, seq)); + + timespec_add_ns(ts, nsecs); +} + +/** + * getnstimeofday - Returns the time of day in a timespec + * @ts: pointer to the timespec to be set * + * Returns the time of day in a timespec. */ -static void second_overflow(void) +void getnstimeofday(struct timespec *ts) { - long ltemp; + __get_realtime_clock_ts(ts); +} - /* Bump the maxerror field */ - time_maxerror += time_tolerance >> SHIFT_USEC; - if (time_maxerror > NTP_PHASE_LIMIT) { - time_maxerror = NTP_PHASE_LIMIT; - time_status |= STA_UNSYNC; - } +EXPORT_SYMBOL(getnstimeofday); - /* - * Leap second processing. If in leap-insert state at the end of the - * day, the system clock is set back one second; if in leap-delete - * state, the system clock is set ahead one second. The microtime() - * routine or external clock driver will insure that reported time is - * always monotonic. The ugly divides should be replaced. - */ - switch (time_state) { - case TIME_OK: - if (time_status & STA_INS) - time_state = TIME_INS; - else if (time_status & STA_DEL) - time_state = TIME_DEL; - break; - case TIME_INS: - if (xtime.tv_sec % 86400 == 0) { - xtime.tv_sec--; - wall_to_monotonic.tv_sec++; - /* - * The timer interpolator will make time change - * gradually instead of an immediate jump by one second - */ - time_interpolator_update(-NSEC_PER_SEC); - time_state = TIME_OOP; - clock_was_set(); - printk(KERN_NOTICE "Clock: inserting leap second " - "23:59:60 UTC\n"); - } - break; - case TIME_DEL: - if ((xtime.tv_sec + 1) % 86400 == 0) { - xtime.tv_sec++; - wall_to_monotonic.tv_sec--; - /* - * Use of time interpolator for a gradual change of - * time - */ - time_interpolator_update(NSEC_PER_SEC); - time_state = TIME_WAIT; - clock_was_set(); - printk(KERN_NOTICE "Clock: deleting leap second " - "23:59:59 UTC\n"); - } - break; - case TIME_OOP: - time_state = TIME_WAIT; - break; - case TIME_WAIT: - if (!(time_status & (STA_INS | STA_DEL))) - time_state = TIME_OK; - } +/** + * do_gettimeofday - Returns the time of day in a timeval + * @tv: pointer to the timeval to be set + * + * NOTE: Users should be converted to using get_realtime_clock_ts() + */ +void do_gettimeofday(struct timeval *tv) +{ + struct timespec now; - /* - * Compute the phase adjustment for the next second. In PLL mode, the - * offset is reduced by a fixed factor times the time constant. In FLL - * mode the offset is used directly. In either mode, the maximum phase - * adjustment for each second is clamped so as to spread the adjustment - * over not more than the number of seconds between updates. - */ - ltemp = time_offset; - if (!(time_status & STA_FLL)) - ltemp = shift_right(ltemp, SHIFT_KG + time_constant); - ltemp = min(ltemp, (MAXPHASE / MINSEC) << SHIFT_UPDATE); - ltemp = max(ltemp, -(MAXPHASE / MINSEC) << SHIFT_UPDATE); - time_offset -= ltemp; - time_adj = ltemp << (SHIFT_SCALE - SHIFT_HZ - SHIFT_UPDATE); + __get_realtime_clock_ts(&now); + tv->tv_sec = now.tv_sec; + tv->tv_usec = now.tv_nsec/1000; +} - /* - * Compute the frequency estimate and additional phase adjustment due - * to frequency error for the next second. When the PPS signal is - * engaged, gnaw on the watchdog counter and update the frequency - * computed by the pll and the PPS signal. - */ - pps_valid++; - if (pps_valid == PPS_VALID) { /* PPS signal lost */ - pps_jitter = MAXTIME; - pps_stabil = MAXFREQ; - time_status &= ~(STA_PPSSIGNAL | STA_PPSJITTER | - STA_PPSWANDER | STA_PPSERROR); - } - ltemp = time_freq + pps_freq; - time_adj += shift_right(ltemp,(SHIFT_USEC + SHIFT_HZ - SHIFT_SCALE)); +EXPORT_SYMBOL(do_gettimeofday); +/** + * do_settimeofday - Sets the time of day + * @tv: pointer to the timespec variable containing the new time + * + * Sets the time of day to the new time and update NTP and notify hrtimers + */ +int do_settimeofday(struct timespec *tv) +{ + unsigned long flags; + time_t wtm_sec, sec = tv->tv_sec; + long wtm_nsec, nsec = tv->tv_nsec; -#if HZ == 100 - /* - * Compensate for (HZ==100) != (1 << SHIFT_HZ). Add 25% and 3.125% to - * get 128.125; => only 0.125% error (p. 14) - */ - time_adj += shift_right(time_adj, 2) + shift_right(time_adj, 5); -#endif -#if HZ == 250 - /* - * Compensate for (HZ==250) != (1 << SHIFT_HZ). Add 1.5625% and - * 0.78125% to get 255.85938; => only 0.05% error (p. 14) - */ - time_adj += shift_right(time_adj, 6) + shift_right(time_adj, 7); -#endif -#if HZ == 1000 - /* - * Compensate for (HZ==1000) != (1 << SHIFT_HZ). Add 1.5625% and - * 0.78125% to get 1023.4375; => only 0.05% error (p. 14) - */ - time_adj += shift_right(time_adj, 6) + shift_right(time_adj, 7); + if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC) + return -EINVAL; + + write_seqlock_irqsave(&xtime_lock, flags); + + nsec -= __get_nsec_offset(); + + wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - sec); + wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - nsec); + + set_normalized_timespec(&xtime, sec, nsec); + set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec); + + clock->error = 0; + ntp_clear(); + + write_sequnlock_irqrestore(&xtime_lock, flags); + + /* signal hrtimers about time change */ + clock_was_set(); + + return 0; +} + +EXPORT_SYMBOL(do_settimeofday); + +/** + * change_clocksource - Swaps clocksources if a new one is available + * + * Accumulates current time interval and initializes new clocksource + */ +static int change_clocksource(void) +{ + struct clocksource *new; + cycle_t now; + u64 nsec; + new = clocksource_get_next(); + if (clock != new) { + now = clocksource_read(new); + nsec = __get_nsec_offset(); + timespec_add_ns(&xtime, nsec); + + clock = new; + clock->cycle_last = now; + printk(KERN_INFO "Time: %s clocksource has been installed.\n", + clock->name); + return 1; + } else if (clock->update_callback) { + return clock->update_callback(); + } + return 0; +} +#else +static inline int change_clocksource(void) +{ + return 0; +} #endif + +/** + * timeofday_is_continuous - check to see if timekeeping is free running + */ +int timekeeping_is_continuous(void) +{ + unsigned long seq; + int ret; + + do { + seq = read_seqbegin(&xtime_lock); + + ret = clock->is_continuous; + + } while (read_seqretry(&xtime_lock, seq)); + + return ret; } /* - * Returns how many microseconds we need to add to xtime this tick - * in doing an adjustment requested with adjtime. + * timekeeping_init - Initializes the clocksource and common timekeeping values */ -static long adjtime_adjustment(void) +void __init timekeeping_init(void) { - long time_adjust_step; + unsigned long flags; - time_adjust_step = time_adjust; - if (time_adjust_step) { - /* - * We are doing an adjtime thing. Prepare time_adjust_step to - * be within bounds. Note that a positive time_adjust means we - * want the clock to run faster. - * - * Limit the amount of the step to be in the range - * -tickadj .. +tickadj - */ - time_adjust_step = min(time_adjust_step, (long)tickadj); - time_adjust_step = max(time_adjust_step, (long)-tickadj); - } - return time_adjust_step; + write_seqlock_irqsave(&xtime_lock, flags); + + ntp_clear(); + + clock = clocksource_get_next(); + clocksource_calculate_interval(clock, tick_nsec); + clock->cycle_last = clocksource_read(clock); + + write_sequnlock_irqrestore(&xtime_lock, flags); +} + + +static int timekeeping_suspended; +/** + * timekeeping_resume - Resumes the generic timekeeping subsystem. + * @dev: unused + * + * This is for the generic clocksource timekeeping. + * xtime/wall_to_monotonic/jiffies/etc are + * still managed by arch specific suspend/resume code. + */ +static int timekeeping_resume(struct sys_device *dev) +{ + unsigned long flags; + + write_seqlock_irqsave(&xtime_lock, flags); + /* restart the last cycle value */ + clock->cycle_last = clocksource_read(clock); + clock->error = 0; + timekeeping_suspended = 0; + write_sequnlock_irqrestore(&xtime_lock, flags); + return 0; +} + +static int timekeeping_suspend(struct sys_device *dev, pm_message_t state) +{ + unsigned long flags; + + write_seqlock_irqsave(&xtime_lock, flags); + timekeeping_suspended = 1; + write_sequnlock_irqrestore(&xtime_lock, flags); + return 0; } -/* in the NTP reference this is called "hardclock()" */ -static void update_wall_time_one_tick(void) +/* sysfs resume/suspend bits for timekeeping */ +static struct sysdev_class timekeeping_sysclass = { + .resume = timekeeping_resume, + .suspend = timekeeping_suspend, + set_kset_name("timekeeping"), +}; + +static struct sys_device device_timer = { + .id = 0, + .cls = &timekeeping_sysclass, +}; + +static int __init timekeeping_init_device(void) { - long time_adjust_step, delta_nsec; + int error = sysdev_class_register(&timekeeping_sysclass); + if (!error) + error = sysdev_register(&device_timer); + return error; +} + +device_initcall(timekeeping_init_device); + +/* + * If the error is already larger, we look ahead even further + * to compensate for late or lost adjustments. + */ +static __always_inline int clocksource_bigadjust(s64 error, s64 *interval, + s64 *offset) +{ + s64 tick_error, i; + u32 look_ahead, adj; + s32 error2, mult; - time_adjust_step = adjtime_adjustment(); - if (time_adjust_step) - /* Reduce by this step the amount of time left */ - time_adjust -= time_adjust_step; - delta_nsec = tick_nsec + time_adjust_step * 1000; /* - * Advance the phase, once it gets to one microsecond, then - * advance the tick more. + * Use the current error value to determine how much to look ahead. + * The larger the error the slower we adjust for it to avoid problems + * with losing too many ticks, otherwise we would overadjust and + * produce an even larger error. The smaller the adjustment the + * faster we try to adjust for it, as lost ticks can do less harm + * here. This is tuned so that an error of about 1 msec is adusted + * within about 1 sec (or 2^20 nsec in 2^SHIFT_HZ ticks). */ - time_phase += time_adj; - if ((time_phase >= FINENSEC) || (time_phase <= -FINENSEC)) { - long ltemp = shift_right(time_phase, (SHIFT_SCALE - 10)); - time_phase -= ltemp << (SHIFT_SCALE - 10); - delta_nsec += ltemp; - } - xtime.tv_nsec += delta_nsec; - time_interpolator_update(delta_nsec); + error2 = clock->error >> (TICK_LENGTH_SHIFT + 22 - 2 * SHIFT_HZ); + error2 = abs(error2); + for (look_ahead = 0; error2 > 0; look_ahead++) + error2 >>= 2; - /* Changes by adjtime() do not take effect till next tick. */ - if (time_next_adjust != 0) { - time_adjust = time_next_adjust; - time_next_adjust = 0; + /* + * Now calculate the error in (1 << look_ahead) ticks, but first + * remove the single look ahead already included in the error. + */ + tick_error = current_tick_length() >> + (TICK_LENGTH_SHIFT - clock->shift + 1); + tick_error -= clock->xtime_interval >> 1; + error = ((error - tick_error) >> look_ahead) + tick_error; + + /* Finally calculate the adjustment shift value. */ + i = *interval; + mult = 1; + if (error < 0) { + error = -error; + *interval = -*interval; + *offset = -*offset; + mult = -1; } + for (adj = 0; error > i; adj++) + error >>= 1; + + *interval <<= adj; + *offset <<= adj; + return mult << adj; } /* - * Return how long ticks are at the moment, that is, how much time - * update_wall_time_one_tick will add to xtime next time we call it - * (assuming no calls to do_adjtimex in the meantime). - * The return value is in fixed-point nanoseconds with SHIFT_SCALE-10 - * bits to the right of the binary point. - * This function has no side-effects. + * Adjust the multiplier to reduce the error value, + * this is optimized for the most common adjustments of -1,0,1, + * for other values we can do a bit more work. */ -u64 current_tick_length(void) +static void clocksource_adjust(struct clocksource *clock, s64 offset) { - long delta_nsec; + s64 error, interval = clock->cycle_interval; + int adj; + + error = clock->error >> (TICK_LENGTH_SHIFT - clock->shift - 1); + if (error > interval) { + error >>= 2; + if (likely(error <= interval)) + adj = 1; + else + adj = clocksource_bigadjust(error, &interval, &offset); + } else if (error < -interval) { + error >>= 2; + if (likely(error >= -interval)) { + adj = -1; + interval = -interval; + offset = -offset; + } else + adj = clocksource_bigadjust(error, &interval, &offset); + } else + return; - delta_nsec = tick_nsec + adjtime_adjustment() * 1000; - return ((u64) delta_nsec << (SHIFT_SCALE - 10)) + time_adj; + clock->mult += adj; + clock->xtime_interval += interval; + clock->xtime_nsec -= offset; + clock->error -= (interval - offset) << + (TICK_LENGTH_SHIFT - clock->shift); } -/* - * Using a loop looks inefficient, but "ticks" is - * usually just one (we shouldn't be losing ticks, - * we're doing this this way mainly for interrupt - * latency reasons, not because we think we'll - * have lots of lost timer ticks +/** + * update_wall_time - Uses the current clocksource to increment the wall time + * + * Called from the timer interrupt, must hold a write on xtime_lock. */ -static void update_wall_time(unsigned long ticks) +static void update_wall_time(void) { - do { - ticks--; - update_wall_time_one_tick(); - if (xtime.tv_nsec >= 1000000000) { - xtime.tv_nsec -= 1000000000; + cycle_t offset; + + /* Make sure we're fully resumed: */ + if (unlikely(timekeeping_suspended)) + return; + +#ifdef CONFIG_GENERIC_TIME + offset = (clocksource_read(clock) - clock->cycle_last) & clock->mask; +#else + offset = clock->cycle_interval; +#endif + clock->xtime_nsec += (s64)xtime.tv_nsec << clock->shift; + + /* normally this loop will run just once, however in the + * case of lost or late ticks, it will accumulate correctly. + */ + while (offset >= clock->cycle_interval) { + /* accumulate one interval */ + clock->xtime_nsec += clock->xtime_interval; + clock->cycle_last += clock->cycle_interval; + offset -= clock->cycle_interval; + + if (clock->xtime_nsec >= (u64)NSEC_PER_SEC << clock->shift) { + clock->xtime_nsec -= (u64)NSEC_PER_SEC << clock->shift; xtime.tv_sec++; second_overflow(); } - } while (ticks); + + /* interpolator bits */ + time_interpolator_update(clock->xtime_interval + >> clock->shift); + + /* accumulate error between NTP and clock interval */ + clock->error += current_tick_length(); + clock->error -= clock->xtime_interval << (TICK_LENGTH_SHIFT - clock->shift); + } + + /* correct the clock when NTP error is too big */ + clocksource_adjust(clock, offset); + + /* store full nanoseconds into xtime */ + xtime.tv_nsec = (s64)clock->xtime_nsec >> clock->shift; + clock->xtime_nsec -= (s64)xtime.tv_nsec << clock->shift; + + /* check to see if there is a new clocksource to use */ + if (change_clocksource()) { + clock->error = 0; + clock->xtime_nsec = 0; + clocksource_calculate_interval(clock, tick_nsec); + } } /* @@ -851,7 +1126,7 @@ void update_process_times(int user_tick) */ static unsigned long count_active_tasks(void) { - return (nr_running() + nr_uninterruptible()) * FIXED_1; + return nr_active() * FIXED_1; } /* @@ -876,24 +1151,23 @@ static inline void calc_load(unsigned long ticks) static int count = LOAD_FREQ; count -= ticks; - if (count < 0) { - count += LOAD_FREQ; + if (unlikely(count < 0)) { active_tasks = count_active_tasks(); - CALC_LOAD(avenrun[0], EXP_1, active_tasks); - CALC_LOAD(avenrun[1], EXP_5, active_tasks); - CALC_LOAD(avenrun[2], EXP_15, active_tasks); + do { + CALC_LOAD(avenrun[0], EXP_1, active_tasks); + CALC_LOAD(avenrun[1], EXP_5, active_tasks); + CALC_LOAD(avenrun[2], EXP_15, active_tasks); + count += LOAD_FREQ; + } while (count < 0); } } -/* jiffies at the most recent update of wall time */ -unsigned long wall_jiffies = INITIAL_JIFFIES; - /* * This read-write spinlock protects us from races in SMP while * playing with xtime and avenrun. */ #ifndef ARCH_HAVE_XTIME_LOCK -seqlock_t xtime_lock __cacheline_aligned_in_smp = SEQLOCK_UNLOCKED; +__cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); EXPORT_SYMBOL(xtime_lock); #endif @@ -903,7 +1177,7 @@ EXPORT_SYMBOL(xtime_lock); */ static void run_timer_softirq(struct softirq_action *h) { - tvec_base_t *base = &__get_cpu_var(tvec_bases); + tvec_base_t *base = __get_cpu_var(tvec_bases); hrtimer_run_queues(); if (time_after_eq(jiffies, base->timer_jiffies)) @@ -916,21 +1190,16 @@ static void run_timer_softirq(struct softirq_action *h) void run_local_timers(void) { raise_softirq(TIMER_SOFTIRQ); + softlockup_tick(); } /* * Called by the timer interrupt. xtime_lock must already be taken * by the timer IRQ! */ -static inline void update_times(void) +static inline void update_times(unsigned long ticks) { - unsigned long ticks; - - ticks = jiffies - wall_jiffies; - if (ticks) { - wall_jiffies += ticks; - update_wall_time(ticks); - } + update_wall_time(); calc_load(ticks); } @@ -940,13 +1209,10 @@ static inline void update_times(void) * jiffies is defined in the linker script... */ -void do_timer(struct pt_regs *regs) +void do_timer(unsigned long ticks) { - jiffies_64++; - /* prevent loading jiffies before storing new jiffies_64 value. */ - barrier(); - update_times(); - softlockup_tick(regs); + jiffies_64 += ticks; + update_times(ticks); } #ifdef __ARCH_WANT_SYS_ALARM @@ -957,19 +1223,7 @@ void do_timer(struct pt_regs *regs) */ asmlinkage unsigned long sys_alarm(unsigned int seconds) { - struct itimerval it_new, it_old; - unsigned int oldalarm; - - it_new.it_interval.tv_sec = it_new.it_interval.tv_usec = 0; - it_new.it_value.tv_sec = seconds; - it_new.it_value.tv_usec = 0; - do_setitimer(ITIMER_REAL, &it_new, &it_old); - oldalarm = it_old.it_value.tv_sec; - /* ehhh.. We can't return 0 if we have an alarm pending.. */ - /* And we'd better return too much than too little anyway */ - if ((!oldalarm && it_old.it_value.tv_usec) || it_old.it_value.tv_usec >= 500000) - oldalarm++; - return oldalarm; + return alarm_setitimer(seconds); } #endif @@ -990,9 +1244,9 @@ asmlinkage long sys_getpid(void) } /* - * Accessing ->real_parent is not SMP-safe, it could + * Accessing ->parent is not SMP-safe, it could * change from under us. However, we can use a stale - * value of ->real_parent under rcu_read_lock(), see + * value of ->parent under rcu_read_lock(), see * release_task()->call_rcu(delayed_put_task_struct). */ asmlinkage long sys_getppid(void) @@ -1000,9 +1254,8 @@ asmlinkage long sys_getppid(void) int pid; rcu_read_lock(); - pid = rcu_dereference(current->real_parent)->tgid; + pid = rcu_dereference(current->parent)->tgid; rcu_read_unlock(); - return vx_map_pid(pid); } @@ -1048,7 +1301,7 @@ asmlinkage long sys_getegid(void) static void process_timeout(unsigned long __data) { - wake_up_process((task_t *)__data); + wake_up_process((struct task_struct *)__data); } /** @@ -1102,11 +1355,10 @@ fastcall signed long __sched schedule_timeout(signed long timeout) * should never happens anyway). You just have the printk() * that will tell you if something is gone wrong and where. */ - if (timeout < 0) - { + if (timeout < 0) { printk(KERN_ERR "schedule_timeout: wrong timeout " - "value %lx from %p\n", timeout, - __builtin_return_address(0)); + "value %lx\n", timeout); + dump_stack(); current->state = TASK_RUNNING; goto out; } @@ -1150,8 +1402,9 @@ asmlinkage long sys_gettid(void) return current->pid; } -/* +/** * sys_sysinfo - fill in sysinfo struct + * @info: pointer to buffer to fill */ asmlinkage long sys_sysinfo(struct sysinfo __user *info) { @@ -1241,13 +1494,50 @@ asmlinkage long sys_sysinfo(struct sysinfo __user *info) return 0; } -static void __devinit init_timers_cpu(int cpu) +/* + * lockdep: we want to track each per-CPU base as a separate lock-class, + * but timer-bases are kmalloc()-ed, so we need to attach separate + * keys to them: + */ +static struct lock_class_key base_lock_keys[NR_CPUS]; + +static int __devinit init_timers_cpu(int cpu) { int j; tvec_base_t *base; + static char __devinitdata tvec_base_done[NR_CPUS]; + + if (!tvec_base_done[cpu]) { + static char boot_done; + + if (boot_done) { + /* + * The APs use this path later in boot + */ + base = kmalloc_node(sizeof(*base), GFP_KERNEL, + cpu_to_node(cpu)); + if (!base) + return -ENOMEM; + memset(base, 0, sizeof(*base)); + per_cpu(tvec_bases, cpu) = base; + } else { + /* + * This is for the boot CPU - we use compile-time + * static initialisation because per-cpu memory isn't + * ready yet and because the memory allocators are not + * initialised either. + */ + boot_done = 1; + base = &boot_tvec_bases; + } + tvec_base_done[cpu] = 1; + } else { + base = per_cpu(tvec_bases, cpu); + } + + spin_lock_init(&base->lock); + lockdep_set_class(&base->lock, base_lock_keys + cpu); - base = &per_cpu(tvec_bases, cpu); - spin_lock_init(&base->t_base.lock); for (j = 0; j < TVN_SIZE; j++) { INIT_LIST_HEAD(base->tv5.vec + j); INIT_LIST_HEAD(base->tv4.vec + j); @@ -1258,6 +1548,7 @@ static void __devinit init_timers_cpu(int cpu) INIT_LIST_HEAD(base->tv1.vec + j); base->timer_jiffies = jiffies; + return 0; } #ifdef CONFIG_HOTPLUG_CPU @@ -1268,7 +1559,7 @@ static void migrate_timer_list(tvec_base_t *new_base, struct list_head *head) while (!list_empty(head)) { timer = list_entry(head->next, struct timer_list, entry); detach_timer(timer, 0); - timer->base = &new_base->t_base; + timer->base = new_base; internal_add_timer(new_base, timer); } } @@ -1280,15 +1571,15 @@ static void __devinit migrate_timers(int cpu) int i; BUG_ON(cpu_online(cpu)); - old_base = &per_cpu(tvec_bases, cpu); - new_base = &get_cpu_var(tvec_bases); + old_base = per_cpu(tvec_bases, cpu); + new_base = get_cpu_var(tvec_bases); local_irq_disable(); - spin_lock(&new_base->t_base.lock); - spin_lock(&old_base->t_base.lock); + spin_lock(&new_base->lock); + spin_lock(&old_base->lock); + + BUG_ON(old_base->running_timer); - if (old_base->t_base.running_timer) - BUG(); for (i = 0; i < TVR_SIZE; i++) migrate_timer_list(new_base, old_base->tv1.vec + i); for (i = 0; i < TVN_SIZE; i++) { @@ -1298,20 +1589,21 @@ static void __devinit migrate_timers(int cpu) migrate_timer_list(new_base, old_base->tv5.vec + i); } - spin_unlock(&old_base->t_base.lock); - spin_unlock(&new_base->t_base.lock); + spin_unlock(&old_base->lock); + spin_unlock(&new_base->lock); local_irq_enable(); put_cpu_var(tvec_bases); } #endif /* CONFIG_HOTPLUG_CPU */ -static int __devinit timer_cpu_notify(struct notifier_block *self, +static int __cpuinit timer_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) { long cpu = (long)hcpu; switch(action) { case CPU_UP_PREPARE: - init_timers_cpu(cpu); + if (init_timers_cpu(cpu) < 0) + return NOTIFY_BAD; break; #ifdef CONFIG_HOTPLUG_CPU case CPU_DEAD: @@ -1324,15 +1616,17 @@ static int __devinit timer_cpu_notify(struct notifier_block *self, return NOTIFY_OK; } -static struct notifier_block __devinitdata timers_nb = { +static struct notifier_block __cpuinitdata timers_nb = { .notifier_call = timer_cpu_notify, }; void __init init_timers(void) { - timer_cpu_notify(&timers_nb, (unsigned long)CPU_UP_PREPARE, + int err = timer_cpu_notify(&timers_nb, (unsigned long)CPU_UP_PREPARE, (void *)(long)smp_processor_id()); + + BUG_ON(err == NOTIFY_BAD); register_cpu_notifier(&timers_nb); open_softirq(TIMER_SOFTIRQ, run_timer_softirq, NULL); } @@ -1417,7 +1711,7 @@ unsigned long time_interpolator_get_offset(void) #define INTERPOLATOR_ADJUST 65536 #define INTERPOLATOR_MAX_SKIP 10*INTERPOLATOR_ADJUST -static void time_interpolator_update(long delta_nsec) +void time_interpolator_update(long delta_nsec) { u64 counter; unsigned long offset; @@ -1454,7 +1748,7 @@ static void time_interpolator_update(long delta_nsec) */ if (jiffies % INTERPOLATOR_ADJUST == 0) { - if (time_interpolator->skips == 0 && time_interpolator->offset > TICK_NSEC) + if (time_interpolator->skips == 0 && time_interpolator->offset > tick_nsec) time_interpolator->nsec_per_cyc--; if (time_interpolator->ns_skipped > INTERPOLATOR_MAX_SKIP && time_interpolator->offset == 0) time_interpolator->nsec_per_cyc++; @@ -1478,8 +1772,7 @@ register_time_interpolator(struct time_interpolator *ti) unsigned long flags; /* Sanity check */ - if (ti->frequency == 0 || ti->mask == 0) - BUG(); + BUG_ON(ti->frequency == 0 || ti->mask == 0); ti->nsec_per_cyc = ((u64)NSEC_PER_SEC << ti->shift) / ti->frequency; spin_lock(&time_interpolator_lock);