diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2012-03-20 10:32:09 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2012-03-20 10:32:09 -0700 |
commit | 161f7a7161191ab9c2e97f787829ef8dd2b95771 (patch) | |
tree | 9776d3f963c7f0d247b7fb324eab4811a1302f67 | |
parent | 2ba68940c893c8f0bfc8573c041254251bb6aeab (diff) | |
parent | a078c6d0e6288fad6d83fb6d5edd91ddb7b6ab33 (diff) |
Merge branch 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer changes for v3.4 from Ingo Molnar
* 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (32 commits)
ntp: Fix integer overflow when setting time
math: Introduce div64_long
cs5535-clockevt: Allow the MFGPT IRQ to be shared
cs5535-clockevt: Don't ignore MFGPT on SMP-capable kernels
x86/time: Eliminate unused irq0_irqs counter
clocksource: scx200_hrt: Fix the build
x86/tsc: Reduce the TSC sync check time for core-siblings
timer: Fix bad idle check on irq entry
nohz: Remove ts->Einidle checks before restarting the tick
nohz: Remove update_ts_time_stat from tick_nohz_start_idle
clockevents: Leave the broadcast device in shutdown mode when not needed
clocksource: Load the ACPI PM clocksource asynchronously
clocksource: scx200_hrt: Convert scx200 to use clocksource_register_hz
clocksource: Get rid of clocksource_calc_mult_shift()
clocksource: dbx500: convert to clocksource_register_hz()
clocksource: scx200_hrt: use pr_<level> instead of printk
time: Move common updates to a function
time: Reorder so the hot data is together
time: Remove most of xtime_lock usage in timekeeping.c
ntp: Add ntp_lock to replace xtime_locking
...
-rw-r--r-- | arch/x86/include/asm/hardirq.h | 1 | ||||
-rw-r--r-- | arch/x86/kernel/time.c | 3 | ||||
-rw-r--r-- | arch/x86/kernel/tsc_sync.c | 29 | ||||
-rw-r--r-- | drivers/clocksource/acpi_pm.c | 24 | ||||
-rw-r--r-- | drivers/clocksource/clksrc-dbx500-prcmu.c | 5 | ||||
-rw-r--r-- | drivers/clocksource/cs5535-clockevt.c | 3 | ||||
-rw-r--r-- | drivers/clocksource/cyclone.c | 2 | ||||
-rw-r--r-- | drivers/clocksource/scx200_hrt.c | 24 | ||||
-rw-r--r-- | drivers/rtc/interface.c | 30 | ||||
-rw-r--r-- | include/linux/clocksource.h | 7 | ||||
-rw-r--r-- | include/linux/math64.h | 4 | ||||
-rw-r--r-- | include/linux/timex.h | 17 | ||||
-rw-r--r-- | kernel/softirq.c | 2 | ||||
-rw-r--r-- | kernel/time/ntp.c | 85 | ||||
-rw-r--r-- | kernel/time/tick-broadcast.c | 4 | ||||
-rw-r--r-- | kernel/time/tick-sched.c | 17 | ||||
-rw-r--r-- | kernel/time/timekeeping.c | 338 |
17 files changed, 343 insertions, 252 deletions
diff --git a/arch/x86/include/asm/hardirq.h b/arch/x86/include/asm/hardirq.h index da0b3ca815b7..382f75d735f3 100644 --- a/arch/x86/include/asm/hardirq.h +++ b/arch/x86/include/asm/hardirq.h @@ -7,7 +7,6 @@ typedef struct { unsigned int __softirq_pending; unsigned int __nmi_count; /* arch dependent */ - unsigned int irq0_irqs; #ifdef CONFIG_X86_LOCAL_APIC unsigned int apic_timer_irqs; /* arch dependent */ unsigned int irq_spurious_count; diff --git a/arch/x86/kernel/time.c b/arch/x86/kernel/time.c index dd5fbf4101fc..c6eba2b42673 100644 --- a/arch/x86/kernel/time.c +++ b/arch/x86/kernel/time.c @@ -57,9 +57,6 @@ EXPORT_SYMBOL(profile_pc); */ static irqreturn_t timer_interrupt(int irq, void *dev_id) { - /* Keep nmi watchdog up to date */ - inc_irq_stat(irq0_irqs); - global_clock_event->event_handler(global_clock_event); /* MCA bus quirk: Acknowledge irq0 by setting bit 7 in port 0x61 */ diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index 9eba29b46cb7..fc25e60a5884 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -42,7 +42,7 @@ static __cpuinitdata int nr_warps; /* * TSC-warp measurement loop running on both CPUs: */ -static __cpuinit void check_tsc_warp(void) +static __cpuinit void check_tsc_warp(unsigned int timeout) { cycles_t start, now, prev, end; int i; @@ -51,9 +51,9 @@ static __cpuinit void check_tsc_warp(void) start = get_cycles(); rdtsc_barrier(); /* - * The measurement runs for 20 msecs: + * The measurement runs for 'timeout' msecs: */ - end = start + tsc_khz * 20ULL; + end = start + (cycles_t) tsc_khz * timeout; now = start; for (i = 0; ; i++) { @@ -99,6 +99,25 @@ static __cpuinit void check_tsc_warp(void) } /* + * If the target CPU coming online doesn't have any of its core-siblings + * online, a timeout of 20msec will be used for the TSC-warp measurement + * loop. Otherwise a smaller timeout of 2msec will be used, as we have some + * information about this socket already (and this information grows as we + * have more and more logical-siblings in that socket). + * + * Ideally we should be able to skip the TSC sync check on the other + * core-siblings, if the first logical CPU in a socket passed the sync test. + * But as the TSC is per-logical CPU and can potentially be modified wrongly + * by the bios, TSC sync test for smaller duration should be able + * to catch such errors. Also this will catch the condition where all the + * cores in the socket doesn't get reset at the same time. + */ +static inline unsigned int loop_timeout(int cpu) +{ + return (cpumask_weight(cpu_core_mask(cpu)) > 1) ? 2 : 20; +} + +/* * Source CPU calls into this - it waits for the freshly booted * target CPU to arrive and then starts the measurement: */ @@ -135,7 +154,7 @@ void __cpuinit check_tsc_sync_source(int cpu) */ atomic_inc(&start_count); - check_tsc_warp(); + check_tsc_warp(loop_timeout(cpu)); while (atomic_read(&stop_count) != cpus-1) cpu_relax(); @@ -183,7 +202,7 @@ void __cpuinit check_tsc_sync_target(void) while (atomic_read(&start_count) != cpus) cpu_relax(); - check_tsc_warp(); + check_tsc_warp(loop_timeout(smp_processor_id())); /* * Ok, we are done: diff --git a/drivers/clocksource/acpi_pm.c b/drivers/clocksource/acpi_pm.c index 6b5cf02c35c8..82e882028fcf 100644 --- a/drivers/clocksource/acpi_pm.c +++ b/drivers/clocksource/acpi_pm.c @@ -23,6 +23,7 @@ #include <linux/init.h> #include <linux/pci.h> #include <linux/delay.h> +#include <linux/async.h> #include <asm/io.h> /* @@ -179,17 +180,15 @@ static int verify_pmtmr_rate(void) /* Number of reads we try to get two different values */ #define ACPI_PM_READ_CHECKS 10000 -static int __init init_acpi_pm_clocksource(void) +static void __init acpi_pm_clocksource_async(void *unused, async_cookie_t cookie) { cycle_t value1, value2; unsigned int i, j = 0; - if (!pmtmr_ioport) - return -ENODEV; /* "verify" this timing source: */ for (j = 0; j < ACPI_PM_MONOTONICITY_CHECKS; j++) { - udelay(100 * j); + usleep_range(100 * j, 100 * j + 100); value1 = clocksource_acpi_pm.read(&clocksource_acpi_pm); for (i = 0; i < ACPI_PM_READ_CHECKS; i++) { value2 = clocksource_acpi_pm.read(&clocksource_acpi_pm); @@ -203,25 +202,34 @@ static int __init init_acpi_pm_clocksource(void) " 0x%#llx, 0x%#llx - aborting.\n", value1, value2); pmtmr_ioport = 0; - return -EINVAL; + return; } if (i == ACPI_PM_READ_CHECKS) { printk(KERN_INFO "PM-Timer failed consistency check " " (0x%#llx) - aborting.\n", value1); pmtmr_ioport = 0; - return -ENODEV; + return; } } if (verify_pmtmr_rate() != 0){ pmtmr_ioport = 0; - return -ENODEV; + return; } - return clocksource_register_hz(&clocksource_acpi_pm, + clocksource_register_hz(&clocksource_acpi_pm, PMTMR_TICKS_PER_SEC); } +static int __init init_acpi_pm_clocksource(void) +{ + if (!pmtmr_ioport) + return -ENODEV; + + async_schedule(acpi_pm_clocksource_async, NULL); + return 0; +} + /* We use fs_initcall because we want the PCI fixups to have run * but we still need to load before device_initcall */ diff --git a/drivers/clocksource/clksrc-dbx500-prcmu.c b/drivers/clocksource/clksrc-dbx500-prcmu.c index fb6b6d28b60e..c26c369eb9e6 100644 --- a/drivers/clocksource/clksrc-dbx500-prcmu.c +++ b/drivers/clocksource/clksrc-dbx500-prcmu.c @@ -52,7 +52,6 @@ static struct clocksource clocksource_dbx500_prcmu = { .name = "dbx500-prcmu-timer", .rating = 300, .read = clksrc_dbx500_prcmu_read, - .shift = 10, .mask = CLOCKSOURCE_MASK(32), .flags = CLOCK_SOURCE_IS_CONTINUOUS, }; @@ -90,7 +89,5 @@ void __init clksrc_dbx500_prcmu_init(void __iomem *base) setup_sched_clock(dbx500_prcmu_sched_clock_read, 32, RATE_32K); #endif - clocksource_calc_mult_shift(&clocksource_dbx500_prcmu, - RATE_32K, SCHED_CLOCK_MIN_WRAP); - clocksource_register(&clocksource_dbx500_prcmu); + clocksource_register_hz(&clocksource_dbx500_prcmu, RATE_32K); } diff --git a/drivers/clocksource/cs5535-clockevt.c b/drivers/clocksource/cs5535-clockevt.c index b7dab32ce63c..540795cd0760 100644 --- a/drivers/clocksource/cs5535-clockevt.c +++ b/drivers/clocksource/cs5535-clockevt.c @@ -100,7 +100,6 @@ static struct clock_event_device cs5535_clockevent = { .set_mode = mfgpt_set_mode, .set_next_event = mfgpt_next_event, .rating = 250, - .cpumask = cpu_all_mask, .shift = 32 }; @@ -133,7 +132,7 @@ static irqreturn_t mfgpt_tick(int irq, void *dev_id) static struct irqaction mfgptirq = { .handler = mfgpt_tick, - .flags = IRQF_DISABLED | IRQF_NOBALANCING | IRQF_TIMER, + .flags = IRQF_DISABLED | IRQF_NOBALANCING | IRQF_TIMER | IRQF_SHARED, .name = DRV_NAME, }; diff --git a/drivers/clocksource/cyclone.c b/drivers/clocksource/cyclone.c index 72f811f73e9c..9e0998f22885 100644 --- a/drivers/clocksource/cyclone.c +++ b/drivers/clocksource/cyclone.c @@ -55,11 +55,11 @@ static int __init init_cyclone_clocksource(void) } /* even on 64bit systems, this is only 32bits: */ base = readl(reg); + iounmap(reg); if (!base) { printk(KERN_ERR "Summit chipset: Could not find valid CBAR value.\n"); return -ENODEV; } - iounmap(reg); /* setup PMCC: */ offset = base + CYCLONE_PMCC_OFFSET; diff --git a/drivers/clocksource/scx200_hrt.c b/drivers/clocksource/scx200_hrt.c index 27f4d9637b62..64f9e8294434 100644 --- a/drivers/clocksource/scx200_hrt.c +++ b/drivers/clocksource/scx200_hrt.c @@ -49,9 +49,6 @@ static cycle_t read_hrt(struct clocksource *cs) return (cycle_t) inl(scx200_cb_base + SCx200_TIMER_OFFSET); } -#define HRT_SHIFT_1 22 -#define HRT_SHIFT_27 26 - static struct clocksource cs_hrt = { .name = "scx200_hrt", .rating = 250, @@ -63,6 +60,7 @@ static struct clocksource cs_hrt = { static int __init init_hrt_clocksource(void) { + u32 freq; /* Make sure scx200 has initialized the configuration block */ if (!scx200_cb_present()) return -ENODEV; @@ -71,7 +69,7 @@ static int __init init_hrt_clocksource(void) if (!request_region(scx200_cb_base + SCx200_TIMER_OFFSET, SCx200_TIMER_SIZE, "NatSemi SCx200 High-Resolution Timer")) { - printk(KERN_WARNING NAME ": unable to lock timer region\n"); + pr_warn("unable to lock timer region\n"); return -ENODEV; } @@ -79,19 +77,13 @@ static int __init init_hrt_clocksource(void) outb(HR_TMEN | (mhz27 ? HR_TMCLKSEL : 0), scx200_cb_base + SCx200_TMCNFG_OFFSET); - if (mhz27) { - cs_hrt.shift = HRT_SHIFT_27; - cs_hrt.mult = clocksource_hz2mult((HRT_FREQ + ppm) * 27, - cs_hrt.shift); - } else { - cs_hrt.shift = HRT_SHIFT_1; - cs_hrt.mult = clocksource_hz2mult(HRT_FREQ + ppm, - cs_hrt.shift); - } - printk(KERN_INFO "enabling scx200 high-res timer (%s MHz +%d ppm)\n", - mhz27 ? "27":"1", ppm); + freq = (HRT_FREQ + ppm); + if (mhz27) + freq *= 27; + + pr_info("enabling scx200 high-res timer (%s MHz +%d ppm)\n", mhz27 ? "27":"1", ppm); - return clocksource_register(&cs_hrt); + return clocksource_register_hz(&cs_hrt, freq); } module_init(init_hrt_clocksource); diff --git a/drivers/rtc/interface.c b/drivers/rtc/interface.c index 8a1c031391d6..dc87eda65814 100644 --- a/drivers/rtc/interface.c +++ b/drivers/rtc/interface.c @@ -73,6 +73,8 @@ int rtc_set_time(struct rtc_device *rtc, struct rtc_time *tm) err = -EINVAL; mutex_unlock(&rtc->ops_lock); + /* A timer might have just expired */ + schedule_work(&rtc->irqwork); return err; } EXPORT_SYMBOL_GPL(rtc_set_time); @@ -112,6 +114,8 @@ int rtc_set_mmss(struct rtc_device *rtc, unsigned long secs) err = -EINVAL; mutex_unlock(&rtc->ops_lock); + /* A timer might have just expired */ + schedule_work(&rtc->irqwork); return err; } @@ -380,18 +384,27 @@ EXPORT_SYMBOL_GPL(rtc_set_alarm); int rtc_initialize_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alarm) { int err; + struct rtc_time now; err = rtc_valid_tm(&alarm->time); if (err != 0) return err; + err = rtc_read_time(rtc, &now); + if (err) + return err; + err = mutex_lock_interruptible(&rtc->ops_lock); if (err) return err; rtc->aie_timer.node.expires = rtc_tm_to_ktime(alarm->time); rtc->aie_timer.period = ktime_set(0, 0); - if (alarm->enabled) { + + /* Alarm has to be enabled & in the futrure for us to enqueue it */ + if (alarm->enabled && (rtc_tm_to_ktime(now).tv64 < + rtc->aie_timer.node.expires.tv64)) { + rtc->aie_timer.enabled = 1; timerqueue_add(&rtc->timerqueue, &rtc->aie_timer.node); } @@ -763,6 +776,14 @@ static int rtc_timer_enqueue(struct rtc_device *rtc, struct rtc_timer *timer) return 0; } +static void rtc_alarm_disable(struct rtc_device *rtc) +{ + if (!rtc->ops || !rtc->ops->alarm_irq_enable) + return; + + rtc->ops->alarm_irq_enable(rtc->dev.parent, false); +} + /** * rtc_timer_remove - Removes a rtc_timer from the rtc_device timerqueue * @rtc rtc device @@ -784,8 +805,10 @@ static void rtc_timer_remove(struct rtc_device *rtc, struct rtc_timer *timer) struct rtc_wkalrm alarm; int err; next = timerqueue_getnext(&rtc->timerqueue); - if (!next) + if (!next) { + rtc_alarm_disable(rtc); return; + } alarm.time = rtc_ktime_to_tm(next->expires); alarm.enabled = 1; err = __rtc_set_alarm(rtc, &alarm); @@ -847,7 +870,8 @@ again: err = __rtc_set_alarm(rtc, &alarm); if (err == -ETIME) goto again; - } + } else + rtc_alarm_disable(rtc); mutex_unlock(&rtc->ops_lock); } diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h index 081147da0564..fbe89e17124e 100644 --- a/include/linux/clocksource.h +++ b/include/linux/clocksource.h @@ -319,13 +319,6 @@ static inline void __clocksource_updatefreq_khz(struct clocksource *cs, u32 khz) __clocksource_updatefreq_scale(cs, 1000, khz); } -static inline void -clocksource_calc_mult_shift(struct clocksource *cs, u32 freq, u32 minsec) -{ - return clocks_calc_mult_shift(&cs->mult, &cs->shift, freq, - NSEC_PER_SEC, minsec); -} - #ifdef CONFIG_GENERIC_TIME_VSYSCALL extern void update_vsyscall(struct timespec *ts, struct timespec *wtm, diff --git a/include/linux/math64.h b/include/linux/math64.h index 23fcdfcba81b..b8ba85544721 100644 --- a/include/linux/math64.h +++ b/include/linux/math64.h @@ -6,6 +6,8 @@ #if BITS_PER_LONG == 64 +#define div64_long(x,y) div64_s64((x),(y)) + /** * div_u64_rem - unsigned 64bit divide with 32bit divisor with remainder * @@ -45,6 +47,8 @@ static inline s64 div64_s64(s64 dividend, s64 divisor) #elif BITS_PER_LONG == 32 +#define div64_long(x,y) div_s64((x),(y)) + #ifndef div_u64_rem static inline u64 div_u64_rem(u64 dividend, u32 divisor, u32 *remainder) { diff --git a/include/linux/timex.h b/include/linux/timex.h index aa60fe7b6ed6..b75e1864ed19 100644 --- a/include/linux/timex.h +++ b/include/linux/timex.h @@ -234,23 +234,9 @@ struct timex { extern unsigned long tick_usec; /* USER_HZ period (usec) */ extern unsigned long tick_nsec; /* ACTHZ period (nsec) */ -/* - * phase-lock loop variables - */ -extern int time_status; /* clock synchronization status bits */ - extern void ntp_init(void); extern void ntp_clear(void); -/** - * ntp_synced - Returns 1 if the NTP status is not UNSYNC - * - */ -static inline int ntp_synced(void) -{ - return !(time_status & STA_UNSYNC); -} - /* Required to safely shift negative values */ #define shift_right(x, s) ({ \ __typeof__(x) __x = (x); \ @@ -264,10 +250,9 @@ static inline int ntp_synced(void) #define NTP_INTERVAL_LENGTH (NSEC_PER_SEC/NTP_INTERVAL_FREQ) /* Returns how long ticks are at present, in ns / 2^NTP_SCALE_SHIFT. */ -extern u64 tick_length; +extern u64 ntp_tick_length(void); extern void second_overflow(void); -extern void update_ntp_one_tick(void); extern int do_adjtimex(struct timex *); extern void hardpps(const struct timespec *, const struct timespec *); diff --git a/kernel/softirq.c b/kernel/softirq.c index 15352e0cbd5d..671f9594e368 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -297,7 +297,7 @@ void irq_enter(void) int cpu = smp_processor_id(); rcu_irq_enter(); - if (idle_cpu(cpu) && !in_interrupt()) { + if (is_idle_task(current) && !in_interrupt()) { /* * Prevent raise_softirq from needlessly waking up ksoftirqd * here, as softirq will be serviced on return from interrupt. diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c index f6117a4c7cb8..6e039b144daf 100644 --- a/kernel/time/ntp.c +++ b/kernel/time/ntp.c @@ -22,13 +22,16 @@ * NTP timekeeping variables: */ +DEFINE_SPINLOCK(ntp_lock); + + /* USER_HZ period (usecs): */ unsigned long tick_usec = TICK_USEC; /* ACTHZ period (nsecs): */ unsigned long tick_nsec; -u64 tick_length; +static u64 tick_length; static u64 tick_length_base; static struct hrtimer leap_timer; @@ -49,7 +52,7 @@ static struct hrtimer leap_timer; static int time_state = TIME_OK; /* clock status bits: */ -int time_status = STA_UNSYNC; +static int time_status = STA_UNSYNC; /* TAI offset (secs): */ static long time_tai; @@ -133,7 +136,7 @@ static inline void pps_reset_freq_interval(void) /** * pps_clear - Clears the PPS state variables * - * Must be called while holding a write on the xtime_lock + * Must be called while holding a write on the ntp_lock */ static inline void pps_clear(void) { @@ -149,7 +152,7 @@ static inline void pps_clear(void) * the last PPS signal. When it reaches 0, indicate that PPS signal is * missing. * - * Must be called while holding a write on the xtime_lock + * Must be called while holding a write on the ntp_lock */ static inline void pps_dec_valid(void) { @@ -233,6 +236,17 @@ static inline void pps_fill_timex(struct timex *txc) #endif /* CONFIG_NTP_PPS */ + +/** + * ntp_synced - Returns 1 if the NTP status is not UNSYNC + * + */ +static inline int ntp_synced(void) +{ + return !(time_status & STA_UNSYNC); +} + + /* * NTP methods: */ @@ -275,7 +289,7 @@ static inline s64 ntp_update_offset_fll(s64 offset64, long secs) time_status |= STA_MODE; - return div_s64(offset64 << (NTP_SCALE_SHIFT - SHIFT_FLL), secs); + return div64_long(offset64 << (NTP_SCALE_SHIFT - SHIFT_FLL), secs); } static void ntp_update_offset(long offset) @@ -330,11 +344,13 @@ static void ntp_update_offset(long offset) /** * ntp_clear - Clears the NTP state variables - * - * Must be called while holding a write on the xtime_lock */ void ntp_clear(void) { + unsigned long flags; + + spin_lock_irqsave(&ntp_lock, flags); + time_adjust = 0; /* stop active adjtime() */ time_status |= STA_UNSYNC; time_maxerror = NTP_PHASE_LIMIT; @@ -347,8 +363,23 @@ void ntp_clear(void) /* Clear PPS state variables */ pps_clear(); + spin_unlock_irqrestore(&ntp_lock, flags); + } + +u64 ntp_tick_length(void) +{ + unsigned long flags; + s64 ret; + + spin_lock_irqsave(&ntp_lock, flags); + ret = tick_length; + spin_unlock_irqrestore(&ntp_lock, flags); + return ret; +} + + /* * 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 @@ -357,14 +388,15 @@ void ntp_clear(void) static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) { enum hrtimer_restart res = HRTIMER_NORESTART; + unsigned long flags; + int leap = 0; - write_seqlock(&xtime_lock); - + spin_lock_irqsave(&ntp_lock, flags); switch (time_state) { case TIME_OK: break; case TIME_INS: - timekeeping_leap_insert(-1); + leap = -1; time_state = TIME_OOP; printk(KERN_NOTICE "Clock: inserting leap second 23:59:60 UTC\n"); @@ -372,7 +404,7 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) res = HRTIMER_RESTART; break; case TIME_DEL: - timekeeping_leap_insert(1); + leap = 1; time_tai--; time_state = TIME_WAIT; printk(KERN_NOTICE @@ -387,8 +419,14 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) time_state = TIME_OK; break; } + spin_unlock_irqrestore(&ntp_lock, flags); - write_sequnlock(&xtime_lock); + /* + * We have to call this outside of the ntp_lock to keep + * the proper locking hierarchy + */ + if (leap) + timekeeping_leap_insert(leap); return res; } @@ -404,6 +442,9 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer) void second_overflow(void) { s64 delta; + unsigned long flags; + + spin_lock_irqsave(&ntp_lock, flags); /* Bump the maxerror field */ time_maxerror += MAXFREQ / NSEC_PER_USEC; @@ -423,23 +464,25 @@ void second_overflow(void) pps_dec_valid(); if (!time_adjust) - return; + goto out; if (time_adjust > MAX_TICKADJ) { time_adjust -= MAX_TICKADJ; tick_length += MAX_TICKADJ_SCALED; - return; + goto out; } if (time_adjust < -MAX_TICKADJ) { time_adjust += MAX_TICKADJ; tick_length -= MAX_TICKADJ_SCALED; - return; + goto out; } tick_length += (s64)(time_adjust * NSEC_PER_USEC / NTP_INTERVAL_FREQ) << NTP_SCALE_SHIFT; time_adjust = 0; +out: + spin_unlock_irqrestore(&ntp_lock, flags); } #ifdef CONFIG_GENERIC_CMOS_UPDATE @@ -663,7 +706,7 @@ int do_adjtimex(struct timex *txc) getnstimeofday(&ts); - write_seqlock_irq(&xtime_lock); + spin_lock_irq(&ntp_lock); if (txc->modes & ADJ_ADJTIME) { long save_adjust = time_adjust; @@ -705,7 +748,7 @@ int do_adjtimex(struct timex *txc) /* fill PPS status fields */ pps_fill_timex(txc); - write_sequnlock_irq(&xtime_lock); + spin_unlock_irq(&ntp_lock); txc->time.tv_sec = ts.tv_sec; txc->time.tv_usec = ts.tv_nsec; @@ -903,7 +946,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts) pts_norm = pps_normalize_ts(*phase_ts); - write_seqlock_irqsave(&xtime_lock, flags); + spin_lock_irqsave(&ntp_lock, flags); /* clear the error bits, they will be set again if needed */ time_status &= ~(STA_PPSJITTER | STA_PPSWANDER | STA_PPSERROR); @@ -916,7 +959,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts) * just start the frequency interval */ if (unlikely(pps_fbase.tv_sec == 0)) { pps_fbase = *raw_ts; - write_sequnlock_irqrestore(&xtime_lock, flags); + spin_unlock_irqrestore(&ntp_lock, flags); return; } @@ -931,7 +974,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts) time_status |= STA_PPSJITTER; /* restart the frequency calibration interval */ pps_fbase = *raw_ts; - write_sequnlock_irqrestore(&xtime_lock, flags); + spin_unlock_irqrestore(&ntp_lock, flags); pr_err("hardpps: PPSJITTER: bad pulse\n"); return; } @@ -948,7 +991,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts) hardpps_update_phase(pts_norm.nsec); - write_sequnlock_irqrestore(&xtime_lock, flags); + spin_unlock_irqrestore(&ntp_lock, flags); } EXPORT_SYMBOL(hardpps); diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c index fd4a7b1625a2..e883f57a3cd3 100644 --- a/kernel/time/tick-broadcast.c +++ b/kernel/time/tick-broadcast.c @@ -575,11 +575,15 @@ void tick_broadcast_switch_to_oneshot(void) unsigned long flags; raw_spin_lock_irqsave(&tick_broadcast_lock, flags); + if (cpumask_empty(tick_get_broadcast_mask())) + goto end; tick_broadcast_device.mode = TICKDEV_MODE_ONESHOT; bc = tick_broadcast_device.evtdev; if (bc) tick_broadcast_setup_oneshot(bc); + +end: raw_spin_unlock_irqrestore(&tick_broadcast_lock, flags); } diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index 7656642e4b8e..3526038f2836 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -182,11 +182,7 @@ static void tick_nohz_stop_idle(int cpu, ktime_t now) static ktime_t tick_nohz_start_idle(int cpu, struct tick_sched *ts) { - ktime_t now; - - now = ktime_get(); - - update_ts_time_stats(cpu, ts, now, NULL); + ktime_t now = ktime_get(); ts->idle_entrytime = now; ts->idle_active = 1; @@ -562,20 +558,21 @@ void tick_nohz_idle_exit(void) local_irq_disable(); - if (ts->idle_active || (ts->inidle && ts->tick_stopped)) + WARN_ON_ONCE(!ts->inidle); + + ts->inidle = 0; + + if (ts->idle_active || ts->tick_stopped) now = ktime_get(); if (ts->idle_active) tick_nohz_stop_idle(cpu, now); - if (!ts->inidle || !ts->tick_stopped) { - ts->inidle = 0; + if (!ts->tick_stopped) { local_irq_enable(); return; } - ts->inidle = 0; - /* Update jiffies first */ select_nohz_load_balancer(0); tick_do_update_jiffies64(now); diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 0c6358186401..403c2a092830 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -25,6 +25,8 @@ struct timekeeper { /* Current clocksource used for timekeeping. */ struct clocksource *clock; + /* NTP adjusted clock multiplier */ + u32 mult; /* The shift value of the current clocksource. */ int shift; @@ -45,12 +47,47 @@ struct timekeeper { /* Shift conversion between clock shifted nano seconds and * ntp shifted nano seconds. */ int ntp_error_shift; - /* NTP adjusted clock multiplier */ - u32 mult; + + /* The current time */ + struct timespec xtime; + /* + * wall_to_monotonic is what we need to add to xtime (or xtime corrected + * for sub jiffie times) to get to monotonic time. Monotonic is pegged + * at zero at system boot time, so wall_to_monotonic will be negative, + * however, we will ALWAYS keep the tv_nsec part positive so we can use + * the usual normalization. + * + * wall_to_monotonic is moved after resume from suspend for the + * monotonic time not to jump. We need to add total_sleep_time to + * wall_to_monotonic to get the real boot based time offset. + * + * - wall_to_monotonic is no longer the boot time, getboottime must be + * used instead. + */ + struct timespec wall_to_monotonic; + /* time spent in suspend */ + struct timespec total_sleep_time; + /* The raw monotonic time for the CLOCK_MONOTONIC_RAW posix clock. */ + struct timespec raw_time; + + /* Seqlock for all timekeeper values */ + seqlock_t lock; }; static struct timekeeper timekeeper; +/* + * This read-write spinlock protects us from races in SMP while + * playing with xtime. + */ +__cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); + + +/* flag for if timekeeping is suspended */ +int __read_mostly timekeeping_suspended; + + + /** * timekeeper_setup_internals - Set up internals to use clocksource clock. * @@ -135,47 +172,28 @@ static inline s64 timekeeping_get_ns_raw(void) return clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); } -/* - * This read-write spinlock protects us from races in SMP while - * playing with xtime. - */ -__cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); - - -/* - * The current time - * wall_to_monotonic is what we need to add to xtime (or xtime corrected - * for sub jiffie times) to get to monotonic time. Monotonic is pegged - * at zero at system boot time, so wall_to_monotonic will be negative, - * however, we will ALWAYS keep the tv_nsec part positive so we can use - * the usual normalization. - * - * wall_to_monotonic is moved after resume from suspend for the monotonic - * time not to jump. We need to add total_sleep_time to wall_to_monotonic - * to get the real boot based time offset. - * - * - wall_to_monotonic is no longer the boot time, getboottime must be - * used instead. - */ -static struct timespec xtime __attribute__ ((aligned (16))); -static struct timespec wall_to_monotonic __attribute__ ((aligned (16))); -static struct timespec total_sleep_time; - -/* - * The raw monotonic time for the CLOCK_MONOTONIC_RAW posix clock. - */ -static struct timespec raw_time; +/* must hold write on timekeeper.lock */ +static void timekeeping_update(bool clearntp) +{ + if (clearntp) { + timekeeper.ntp_error = 0; + ntp_clear(); + } + update_vsyscall(&timekeeper.xtime, &timekeeper.wall_to_monotonic, + timekeeper.clock, timekeeper.mult); +} -/* flag for if timekeeping is suspended */ -int __read_mostly timekeeping_suspended; -/* must hold xtime_lock */ void timekeeping_leap_insert(int leapsecond) { - xtime.tv_sec += leapsecond; - wall_to_monotonic.tv_sec -= leapsecond; - update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock, - timekeeper.mult); + unsigned long flags; + + write_seqlock_irqsave(&timekeeper.lock, flags); + timekeeper.xtime.tv_sec += leapsecond; + timekeeper.wall_to_monotonic.tv_sec -= leapsecond; + timekeeping_update(false); + write_sequnlock_irqrestore(&timekeeper.lock, flags); + } /** @@ -202,10 +220,10 @@ static void timekeeping_forward_now(void) /* If arch requires, add in gettimeoffset() */ nsec += arch_gettimeoffset(); - timespec_add_ns(&xtime, nsec); + timespec_add_ns(&timekeeper.xtime, nsec); nsec = clocksource_cyc2ns(cycle_delta, clock->mult, clock->shift); - timespec_add_ns(&raw_time, nsec); + timespec_add_ns(&timekeeper.raw_time, nsec); } /** @@ -222,15 +240,15 @@ void getnstimeofday(struct timespec *ts) WARN_ON(timekeeping_suspended); do { - seq = read_seqbegin(&xtime_lock); + seq = read_seqbegin(&timekeeper.lock); - *ts = xtime; + *ts = timekeeper.xtime; nsecs = timekeeping_get_ns(); /* If arch requires, add in gettimeoffset() */ nsecs += arch_gettimeoffset(); - } while (read_seqretry(&xtime_lock, seq)); + } while (read_seqretry(&timekeeper.lock, seq)); timespec_add_ns(ts, nsecs); } @@ -245,14 +263,16 @@ ktime_t ktime_get(void) WARN_ON(timekeeping_suspended); do { - seq = read_seqbegin(&xtime_lock); - secs = xtime.tv_sec + wall_to_monotonic.tv_sec; - nsecs = xtime.tv_nsec + wall_to_monotonic.tv_nsec; + seq = read_seqbegin(&timekeeper.lock); |