Updates for timers and timekeeping core code:

- Expose CLOCK_TAI to instrumentation to aid with TSN debugging.
 
   - Ensure that the clockevent is stopped when there is no timer armed to
     avoid pointless wakeups.
 
   - Make the sched clock frequency handling and rounding consistent.
 
   - Provide a better debugobject hint for delayed works. The timer callback
     is always the same, which makes it difficult to identify the underlying
     work. Use the work function as a hint instead.
 
   - Move the timer specific sysctl code into the timer subsystem.
 
   - The usual set of improvements and cleanups
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCgAxFiEEQp8+kY+LLUocC4bMphj1TA10mKEFAmKLPHMTHHRnbHhAbGlu
 dXRyb25peC5kZQAKCRCmGPVMDXSYoZBoEACIURtS8w9PFZ6q/2mFq0pTYi/uI/HQ
 vqbB6gCbrjfL6QwInd7jxDc/UoqEOllG9pTaGdWx/0Gi9syDosEbeop7cvvt2xi+
 pReoEN1kVI3JAVrQFIAuGw4EMuzYB8PfuZkm1PdozcCP9qkgDmtippVxe05sFQ+/
 RPdA29vE3g63eXkSFBhEID23pQR8yKLbqVq6KcH87OipZedL+2fry3yB+/9sLuuU
 /PFLbI6B9f43S2sfo6szzpFkpd6tJlBlu02IrB6gh4IxKrslmZb5onpvcf6iT+19
 rFh5A15GFWoZUC8EjH1sBpATq3wA/jfGEOPWgy07N5SmobtJvWSM5yvT+gC3qXqm
 C/bjyjqXzLKftG7KIXo/hWewtsjdovMbdfcMBsGiatytNBZfI1GR/4Pq60/qpTHZ
 qJo35trOUcP6o1njphwONy3lisq78S7xaozpWO1hIMTcAqGgBkm/lOieGMM4hGnE
 Ps0Im3ZsOXNGllulN+3h+UHstM5/y6f/vzBsw7pfIG66i6KqebAiNjbMfHCr22sX
 7UavNCoFggUQgZVgUYX/AscdW4/Dwx6R5YUqj1EBqztknd70Ac4TqjaIz4Xa6ZER
 z+eQSSt5XqqV2eKWA4FsQYmCIc+BvQ4apSA6+whz9vmsvCYtB7zzSfeh+xkgcl1/
 Cc0N6G5+L9v0Gw==
 =De28
 -----END PGP SIGNATURE-----

Merge tag 'timers-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull timer and timekeeping updates from Thomas Gleixner:

 - Expose CLOCK_TAI to instrumentation to aid with TSN debugging.

 - Ensure that the clockevent is stopped when there is no timer armed to
   avoid pointless wakeups.

 - Make the sched clock frequency handling and rounding consistent.

 - Provide a better debugobject hint for delayed works. The timer
   callback is always the same, which makes it difficult to identify the
   underlying work. Use the work function as a hint instead.

 - Move the timer specific sysctl code into the timer subsystem.

 - The usual set of improvements and cleanups

* tag 'timers-core-2022-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  timers: Provide a better debugobjects hint for delayed works
  time/sched_clock: Fix formatting of frequency reporting code
  time/sched_clock: Use Hz as the unit for clock rate reporting below 4kHz
  time/sched_clock: Round the frequency reported to nearest rather than down
  timekeeping: Consolidate fast timekeeper
  timekeeping: Annotate ktime_get_boot_fast_ns() with data_race()
  timers/nohz: Switch to ONESHOT_STOPPED in the low-res handler when the tick is stopped
  timekeeping: Introduce fast accessor to clock tai
  tracing/timer: Add missing argument documentation of trace points
  clocksource: Replace cpumask_weight() with cpumask_empty()
  timers: Move timer sysctl into the timer code
  clockevents: Use dedicated list iterator variable
  timers: Simplify calc_index()
  timers: Initialize base::next_expiry_recalc in timers_prepare_cpu()

Documentation/core-api/timekeeping.rst

@@ -132,6 +132,7 @@ Some additional variants exist for more specialized cases:
 .. c:function:: u64 ktime_get_mono_fast_ns( void )
 		u64 ktime_get_raw_fast_ns( void )
 		u64 ktime_get_boot_fast_ns( void )
+		u64 ktime_get_tai_fast_ns( void )
 		u64 ktime_get_real_fast_ns( void )
 
 	These variants are safe to call from any context, including from

include/linux/timekeeping.h

@@ -177,6 +177,7 @@ static inline u64 ktime_get_raw_ns(void)
 extern u64 ktime_get_mono_fast_ns(void);
 extern u64 ktime_get_raw_fast_ns(void);
 extern u64 ktime_get_boot_fast_ns(void);
+extern u64 ktime_get_tai_fast_ns(void);
 extern u64 ktime_get_real_fast_ns(void);
 
 /*

include/linux/timer.h

@@ -196,14 +196,6 @@ extern void init_timers(void);
 struct hrtimer;
 extern enum hrtimer_restart it_real_fn(struct hrtimer *);
 
-#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON)
-struct ctl_table;
-
-extern unsigned int sysctl_timer_migration;
-int timer_migration_handler(struct ctl_table *table, int write,
-			    void *buffer, size_t *lenp, loff_t *ppos);
-#endif
-
 unsigned long __round_jiffies(unsigned long j, int cpu);
 unsigned long __round_jiffies_relative(unsigned long j, int cpu);
 unsigned long round_jiffies(unsigned long j);

include/trace/events/timer.h

@@ -48,6 +48,7 @@ DEFINE_EVENT(timer_class, timer_init,
  * timer_start - called when the timer is started
  * @timer:	pointer to struct timer_list
  * @expires:	the timers expiry time
+ * @flags:	the timers flags
  */
 TRACE_EVENT(timer_start,
 
@@ -84,6 +85,7 @@ TRACE_EVENT(timer_start,
 /**
  * timer_expire_entry - called immediately before the timer callback
  * @timer:	pointer to struct timer_list
+ * @baseclk:	value of timer_base::clk when timer expires
  *
  * Allows to determine the timer latency.
  */
@@ -190,7 +192,8 @@ TRACE_EVENT(hrtimer_init,
 
 /**
  * hrtimer_start - called when the hrtimer is started
- * @hrtimer: pointer to struct hrtimer
+ * @hrtimer:	pointer to struct hrtimer
+ * @mode:	the hrtimers mode
  */
 TRACE_EVENT(hrtimer_start,
 

kernel/sysctl.c

@@ -2288,17 +2288,6 @@ static struct ctl_table kern_table[] = {
 		.extra1		= SYSCTL_ZERO,
 		.extra2		= SYSCTL_ONE,
 	},
-#if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ_COMMON)
-	{
-		.procname	= "timer_migration",
-		.data		= &sysctl_timer_migration,
-		.maxlen		= sizeof(unsigned int),
-		.mode		= 0644,
-		.proc_handler	= timer_migration_handler,
-		.extra1		= SYSCTL_ZERO,
-		.extra2		= SYSCTL_ONE,
-	},
-#endif
 #ifdef CONFIG_BPF_SYSCALL
 	{
 		.procname	= "unprivileged_bpf_disabled",

kernel/time/clockevents.c

@@ -690,7 +690,7 @@ static ssize_t unbind_device_store(struct device *dev,
 {
 	char name[CS_NAME_LEN];
 	ssize_t ret = sysfs_get_uname(buf, name, count);
-	struct clock_event_device *ce;
+	struct clock_event_device *ce = NULL, *iter;
 
 	if (ret < 0)
 		return ret;
@@ -698,9 +698,10 @@ static ssize_t unbind_device_store(struct device *dev,
 	ret = -ENODEV;
 	mutex_lock(&clockevents_mutex);
 	raw_spin_lock_irq(&clockevents_lock);
-	list_for_each_entry(ce, &clockevent_devices, list) {
-		if (!strcmp(ce->name, name)) {
-			ret = __clockevents_try_unbind(ce, dev->id);
+	list_for_each_entry(iter, &clockevent_devices, list) {
+		if (!strcmp(iter->name, name)) {
+			ret = __clockevents_try_unbind(iter, dev->id);
+			ce = iter;
 			break;
 		}
 	}

kernel/time/clocksource.c

@@ -343,7 +343,7 @@ void clocksource_verify_percpu(struct clocksource *cs)
 	cpus_read_lock();
 	preempt_disable();
 	clocksource_verify_choose_cpus();
-	if (cpumask_weight(&cpus_chosen) == 0) {
+	if (cpumask_empty(&cpus_chosen)) {
 		preempt_enable();
 		cpus_read_unlock();
 		pr_warn("Not enough CPUs to check clocksource '%s'.\n", cs->name);

kernel/time/sched_clock.c

@@ -8,6 +8,7 @@
 #include <linux/jiffies.h>
 #include <linux/ktime.h>
 #include <linux/kernel.h>
+#include <linux/math.h>
 #include <linux/moduleparam.h>
 #include <linux/sched.h>
 #include <linux/sched/clock.h>
@@ -199,15 +200,13 @@ sched_clock_register(u64 (*read)(void), int bits, unsigned long rate)
 
 	r = rate;
 	if (r >= 4000000) {
-		r /= 1000000;
+		r = DIV_ROUND_CLOSEST(r, 1000000);
 		r_unit = 'M';
+	} else if (r >= 4000) {
+		r = DIV_ROUND_CLOSEST(r, 1000);
+		r_unit = 'k';
 	} else {
-		if (r >= 1000) {
-			r /= 1000;
-			r_unit = 'k';
-		} else {
-			r_unit = ' ';
-		}
+		r_unit = ' ';
 	}
 
 	/* Calculate the ns resolution of this counter */

kernel/time/tick-sched.c

@@ -928,6 +928,8 @@ static void tick_nohz_stop_tick(struct tick_sched *ts, int cpu)
 	if (unlikely(expires == KTIME_MAX)) {
 		if (ts->nohz_mode == NOHZ_MODE_HIGHRES)
 			hrtimer_cancel(&ts->sched_timer);
+		else
+			tick_program_event(KTIME_MAX, 1);
 		return;
 	}
 
@@ -1364,9 +1366,15 @@ static void tick_nohz_handler(struct clock_event_device *dev)
 	tick_sched_do_timer(ts, now);
 	tick_sched_handle(ts, regs);
 
-	/* No need to reprogram if we are running tickless  */
-	if (unlikely(ts->tick_stopped))
+	if (unlikely(ts->tick_stopped)) {
+		/*
+		 * The clockevent device is not reprogrammed, so change the
+		 * clock event device to ONESHOT_STOPPED to avoid spurious
+		 * interrupts on devices which might not be truly one shot.
+		 */
+		tick_program_event(KTIME_MAX, 1);
 		return;
+	}
 
 	hrtimer_forward(&ts->sched_timer, now, TICK_NSEC);
 	tick_program_event(hrtimer_get_expires(&ts->sched_timer), 1);

kernel/time/timekeeping.c

@@ -429,6 +429,14 @@ static void update_fast_timekeeper(const struct tk_read_base *tkr,
 	memcpy(base + 1, base, sizeof(*base));
 }
 
+static __always_inline u64 fast_tk_get_delta_ns(struct tk_read_base *tkr)
+{
+	u64 delta, cycles = tk_clock_read(tkr);
+
+	delta = clocksource_delta(cycles, tkr->cycle_last, tkr->mask);
+	return timekeeping_delta_to_ns(tkr, delta);
+}
+
 static __always_inline u64 __ktime_get_fast_ns(struct tk_fast *tkf)
 {
 	struct tk_read_base *tkr;
@@ -439,12 +447,7 @@ static __always_inline u64 __ktime_get_fast_ns(struct tk_fast *tkf)
 		seq = raw_read_seqcount_latch(&tkf->seq);
 		tkr = tkf->base + (seq & 0x01);
 		now = ktime_to_ns(tkr->base);
-
-		now += timekeeping_delta_to_ns(tkr,
-				clocksource_delta(
-					tk_clock_read(tkr),
-					tkr->cycle_last,
-					tkr->mask));
+		now += fast_tk_get_delta_ns(tkr);
 	} while (read_seqcount_latch_retry(&tkf->seq, seq));
 
 	return now;
@@ -528,10 +531,27 @@ u64 notrace ktime_get_boot_fast_ns(void)
 {
 	struct timekeeper *tk = &tk_core.timekeeper;
 
-	return (ktime_get_mono_fast_ns() + ktime_to_ns(tk->offs_boot));
+	return (ktime_get_mono_fast_ns() + ktime_to_ns(data_race(tk->offs_boot)));
 }
 EXPORT_SYMBOL_GPL(ktime_get_boot_fast_ns);
 
+/**
+ * ktime_get_tai_fast_ns - NMI safe and fast access to tai clock.
+ *
+ * The same limitations as described for ktime_get_boot_fast_ns() apply. The
+ * mono time and the TAI offset are not read atomically which may yield wrong
+ * readouts. However, an update of the TAI offset is an rare event e.g., caused
+ * by settime or adjtimex with an offset. The user of this function has to deal
+ * with the possibility of wrong timestamps in post processing.
+ */
+u64 notrace ktime_get_tai_fast_ns(void)
+{
+	struct timekeeper *tk = &tk_core.timekeeper;
+
+	return (ktime_get_mono_fast_ns() + ktime_to_ns(data_race(tk->offs_tai)));
+}
+EXPORT_SYMBOL_GPL(ktime_get_tai_fast_ns);
+
 static __always_inline u64 __ktime_get_real_fast(struct tk_fast *tkf, u64 *mono)
 {
 	struct tk_read_base *tkr;
@@ -543,10 +563,7 @@ static __always_inline u64 __ktime_get_real_fast(struct tk_fast *tkf, u64 *mono)
 		tkr = tkf->base + (seq & 0x01);
 		basem = ktime_to_ns(tkr->base);
 		baser = ktime_to_ns(tkr->base_real);
-
-		delta = timekeeping_delta_to_ns(tkr,
-				clocksource_delta(tk_clock_read(tkr),
-				tkr->cycle_last, tkr->mask));
+		delta = fast_tk_get_delta_ns(tkr);
 	} while (read_seqcount_latch_retry(&tkf->seq, seq));
 
 	if (mono)

kernel/time/timer.c

@@ -44,6 +44,7 @@
 #include <linux/slab.h>
 #include <linux/compat.h>
 #include <linux/random.h>
+#include <linux/sysctl.h>
 
 #include <linux/uaccess.h>
 #include <asm/unistd.h>
@@ -223,7 +224,7 @@ static void timer_update_keys(struct work_struct *work);
 static DECLARE_WORK(timer_update_work, timer_update_keys);
 
 #ifdef CONFIG_SMP
-unsigned int sysctl_timer_migration = 1;
+static unsigned int sysctl_timer_migration = 1;
 
 DEFINE_STATIC_KEY_FALSE(timers_migration_enabled);
 
@@ -234,7 +235,42 @@ static void timers_update_migration(void)
 	else
 		static_branch_disable(&timers_migration_enabled);
 }
-#else
+
+#ifdef CONFIG_SYSCTL
+static int timer_migration_handler(struct ctl_table *table, int write,
+			    void *buffer, size_t *lenp, loff_t *ppos)
+{
+	int ret;
+
+	mutex_lock(&timer_keys_mutex);
+	ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
+	if (!ret && write)
+		timers_update_migration();
+	mutex_unlock(&timer_keys_mutex);
+	return ret;
+}
+
+static struct ctl_table timer_sysctl[] = {
+	{
+		.procname	= "timer_migration",
+		.data		= &sysctl_timer_migration,
+		.maxlen		= sizeof(unsigned int),
+		.mode		= 0644,
+		.proc_handler	= timer_migration_handler,
+		.extra1		= SYSCTL_ZERO,
+		.extra2		= SYSCTL_ONE,
+	},
+	{}
+};
+
+static int __init timer_sysctl_init(void)
+{
+	register_sysctl("kernel", timer_sysctl);
+	return 0;
+}
+device_initcall(timer_sysctl_init);
+#endif /* CONFIG_SYSCTL */
+#else /* CONFIG_SMP */
 static inline void timers_update_migration(void) { }
 #endif /* !CONFIG_SMP */
 
@@ -251,19 +287,6 @@ void timers_update_nohz(void)
 	schedule_work(&timer_update_work);
 }
 
-int timer_migration_handler(struct ctl_table *table, int write,
-			    void *buffer, size_t *lenp, loff_t *ppos)
-{
-	int ret;
-
-	mutex_lock(&timer_keys_mutex);
-	ret = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
-	if (!ret && write)
-		timers_update_migration();
-	mutex_unlock(&timer_keys_mutex);
-	return ret;
-}
-
 static inline bool is_timers_nohz_active(void)
 {
 	return static_branch_unlikely(&timers_nohz_active);
@@ -502,7 +525,7 @@ static inline unsigned calc_index(unsigned long expires, unsigned lvl,
 	 *
 	 * Round up with level granularity to prevent this.
 	 */
-	expires = (expires + LVL_GRAN(lvl)) >> LVL_SHIFT(lvl);
+	expires = (expires >> LVL_SHIFT(lvl)) + 1;
 	*bucket_expiry = expires << LVL_SHIFT(lvl);
 	return LVL_OFFS(lvl) + (expires & LVL_MASK);
 }
@@ -615,9 +638,39 @@ static void internal_add_timer(struct timer_base *base, struct timer_list *timer
 
 static const struct debug_obj_descr timer_debug_descr;
 
+struct timer_hint {
+	void	(*function)(struct timer_list *t);
+	long	offset;
+};
+
+#define TIMER_HINT(fn, container, timr, hintfn)			\
+	{							\
+		.function = fn,					\
+		.offset	  = offsetof(container, hintfn) -	\
+			    offsetof(container, timr)		\
+	}
+
+static const struct timer_hint timer_hints[] = {
+	TIMER_HINT(delayed_work_timer_fn,
+		   struct delayed_work, timer, work.func),
+	TIMER_HINT(kthread_delayed_work_timer_fn,
+		   struct kthread_delayed_work, timer, work.func),
+};
+
 static void *timer_debug_hint(void *addr)
 {
-	return ((struct timer_list *) addr)->function;
+	struct timer_list *timer = addr;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(timer_hints); i++) {
+		if (timer_hints[i].function == timer->function) {
+			void (**fn)(void) = addr + timer_hints[i].offset;
+
+			return *fn;
+		}
+	}
+
+	return timer->function;
 }
 
 static bool timer_is_static_object(void *addr)
@@ -1953,6 +2006,7 @@ 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->next_expiry_recalc = false;
 		base->timers_pending = false;
 		base->is_idle = false;
 	}