Revert: Unify CLOCK_MONOTONIC and CLOCK_BOOTTIME

Revert commits

92af4dcb4e ("tracing: Unify the "boot" and "mono" tracing clocks")
127bfa5f43 ("hrtimer: Unify MONOTONIC and BOOTTIME clock behavior")
7250a4047a ("posix-timers: Unify MONOTONIC and BOOTTIME clock behavior")
d6c7270e91 ("timekeeping: Remove boot time specific code")
f2d6fdbfd2 ("Input: Evdev - unify MONOTONIC and BOOTTIME clock behavior")
d6ed449afd ("timekeeping: Make the MONOTONIC clock behave like the BOOTTIME clock")
72199320d4 ("timekeeping: Add the new CLOCK_MONOTONIC_ACTIVE clock")

As stated in the pull request for the unification of CLOCK_MONOTONIC and
CLOCK_BOOTTIME, it was clear that we might have to revert the change.

As reported by several folks systemd and other applications rely on the
documented behaviour of CLOCK_MONOTONIC on Linux and break with the above
changes. After resume daemons time out and other timeout related issues are
observed. Rafael compiled this list:

* systemd kills daemons on resume, after >WatchdogSec seconds
  of suspending (Genki Sky).  [Verified that that's because systemd uses
  CLOCK_MONOTONIC and expects it to not include the suspend time.]

* systemd-journald misbehaves after resume:
  systemd-journald[7266]: File /var/log/journal/016627c3c4784cd4812d4b7e96a34226/system.journal
corrupted or uncleanly shut down, renaming and replacing.
  (Mike Galbraith).

* NetworkManager reports "networking disabled" and networking is broken
  after resume 50% of the time (Pavel).  [May be because of systemd.]

* MATE desktop dims the display and starts the screensaver right after
  system resume (Pavel).

* Full system hang during resume (me).  [May be due to systemd or NM or both.]

That happens on debian and open suse systems.

It's sad, that these problems were neither catched in -next nor by those
folks who expressed interest in this change.

Reported-by: Rafael J. Wysocki <rjw@rjwysocki.net>
Reported-by: Genki Sky <sky@genki.is>,
Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kevin Easton <kevin@guarana.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mark Salyzyn <salyzyn@android.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>

kernel/time/hrtimer.c

@@ -90,6 +90,11 @@ DEFINE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases) =
 			.clockid = CLOCK_REALTIME,
 			.get_time = &ktime_get_real,
 		},
+		{
+			.index = HRTIMER_BASE_BOOTTIME,
+			.clockid = CLOCK_BOOTTIME,
+			.get_time = &ktime_get_boottime,
+		},
 		{
 			.index = HRTIMER_BASE_TAI,
 			.clockid = CLOCK_TAI,
@@ -105,6 +110,11 @@ DEFINE_PER_CPU(struct hrtimer_cpu_base, hrtimer_bases) =
 			.clockid = CLOCK_REALTIME,
 			.get_time = &ktime_get_real,
 		},
+		{
+			.index = HRTIMER_BASE_BOOTTIME_SOFT,
+			.clockid = CLOCK_BOOTTIME,
+			.get_time = &ktime_get_boottime,
+		},
 		{
 			.index = HRTIMER_BASE_TAI_SOFT,
 			.clockid = CLOCK_TAI,
@@ -119,7 +129,7 @@ static const int hrtimer_clock_to_base_table[MAX_CLOCKS] = {
 
 	[CLOCK_REALTIME]	= HRTIMER_BASE_REALTIME,
 	[CLOCK_MONOTONIC]	= HRTIMER_BASE_MONOTONIC,
-	[CLOCK_BOOTTIME]	= HRTIMER_BASE_MONOTONIC,
+	[CLOCK_BOOTTIME]	= HRTIMER_BASE_BOOTTIME,
 	[CLOCK_TAI]		= HRTIMER_BASE_TAI,
 };
 
@@ -571,12 +581,14 @@ __hrtimer_get_next_event(struct hrtimer_cpu_base *cpu_base, unsigned int active_
 static inline ktime_t hrtimer_update_base(struct hrtimer_cpu_base *base)
 {
 	ktime_t *offs_real = &base->clock_base[HRTIMER_BASE_REALTIME].offset;
+	ktime_t *offs_boot = &base->clock_base[HRTIMER_BASE_BOOTTIME].offset;
 	ktime_t *offs_tai = &base->clock_base[HRTIMER_BASE_TAI].offset;
 
 	ktime_t now = ktime_get_update_offsets_now(&base->clock_was_set_seq,
-						   offs_real, offs_tai);
+					    offs_real, offs_boot, offs_tai);
 
 	base->clock_base[HRTIMER_BASE_REALTIME_SOFT].offset = *offs_real;
+	base->clock_base[HRTIMER_BASE_BOOTTIME_SOFT].offset = *offs_boot;
 	base->clock_base[HRTIMER_BASE_TAI_SOFT].offset = *offs_tai;
 
 	return now;

kernel/time/posix-stubs.c

@@ -83,8 +83,6 @@ int do_clock_gettime(clockid_t which_clock, struct timespec64 *tp)
 	case CLOCK_BOOTTIME:
 		get_monotonic_boottime64(tp);
 		break;
-	case CLOCK_MONOTONIC_ACTIVE:
-		ktime_get_active_ts64(tp);
 	default:
 		return -EINVAL;
 	}

kernel/time/posix-timers.c

@@ -252,16 +252,15 @@ static int posix_get_coarse_res(const clockid_t which_clock, struct timespec64 *
 	return 0;
 }
 
-static int posix_get_tai(clockid_t which_clock, struct timespec64 *tp)
+static int posix_get_boottime(const clockid_t which_clock, struct timespec64 *tp)
 {
-	timekeeping_clocktai64(tp);
+	get_monotonic_boottime64(tp);
 	return 0;
 }
 
-static int posix_get_monotonic_active(clockid_t which_clock,
-				      struct timespec64 *tp)
+static int posix_get_tai(clockid_t which_clock, struct timespec64 *tp)
 {
-	ktime_get_active_ts64(tp);
+	timekeeping_clocktai64(tp);
 	return 0;
 }
 
@@ -1317,9 +1316,19 @@ static const struct k_clock clock_tai = {
 	.timer_arm		= common_hrtimer_arm,
 };
 
-static const struct k_clock clock_monotonic_active = {
+static const struct k_clock clock_boottime = {
 	.clock_getres		= posix_get_hrtimer_res,
-	.clock_get		= posix_get_monotonic_active,
+	.clock_get		= posix_get_boottime,
+	.nsleep			= common_nsleep,
+	.timer_create		= common_timer_create,
+	.timer_set		= common_timer_set,
+	.timer_get		= common_timer_get,
+	.timer_del		= common_timer_del,
+	.timer_rearm		= common_hrtimer_rearm,
+	.timer_forward		= common_hrtimer_forward,
+	.timer_remaining	= common_hrtimer_remaining,
+	.timer_try_to_cancel	= common_hrtimer_try_to_cancel,
+	.timer_arm		= common_hrtimer_arm,
 };
 
 static const struct k_clock * const posix_clocks[] = {
@@ -1330,11 +1339,10 @@ static const struct k_clock * const posix_clocks[] = {
 	[CLOCK_MONOTONIC_RAW]		= &clock_monotonic_raw,
 	[CLOCK_REALTIME_COARSE]		= &clock_realtime_coarse,
 	[CLOCK_MONOTONIC_COARSE]	= &clock_monotonic_coarse,
-	[CLOCK_BOOTTIME]		= &clock_monotonic,
+	[CLOCK_BOOTTIME]		= &clock_boottime,
 	[CLOCK_REALTIME_ALARM]		= &alarm_clock,
 	[CLOCK_BOOTTIME_ALARM]		= &alarm_clock,
 	[CLOCK_TAI]			= &clock_tai,
-	[CLOCK_MONOTONIC_ACTIVE]	= &clock_monotonic_active,
 };
 
 static const struct k_clock *clockid_to_kclock(const clockid_t id)

kernel/time/tick-common.c

@@ -419,19 +419,6 @@ void tick_suspend_local(void)
 	clockevents_shutdown(td->evtdev);
 }
 
-static void tick_forward_next_period(void)
-{
-	ktime_t delta, now = ktime_get();
-	u64 n;
-
-	delta = ktime_sub(now, tick_next_period);
-	n = ktime_divns(delta, tick_period);
-	tick_next_period += n * tick_period;
-	if (tick_next_period < now)
-		tick_next_period += tick_period;
-	tick_sched_forward_next_period();
-}
-
 /**
  * tick_resume_local - Resume the local tick device
  *
@@ -444,8 +431,6 @@ void tick_resume_local(void)
 	struct tick_device *td = this_cpu_ptr(&tick_cpu_device);
 	bool broadcast = tick_resume_check_broadcast();
 
-	tick_forward_next_period();
-
 	clockevents_tick_resume(td->evtdev);
 	if (!broadcast) {
 		if (td->mode == TICKDEV_MODE_PERIODIC)

kernel/time/tick-internal.h

@@ -141,12 +141,6 @@ static inline void tick_check_oneshot_broadcast_this_cpu(void) { }
 static inline bool tick_broadcast_oneshot_available(void) { return tick_oneshot_possible(); }
 #endif /* !(BROADCAST && ONESHOT) */
 
-#if defined(CONFIG_NO_HZ_COMMON) || defined(CONFIG_HIGH_RES_TIMERS)
-extern void tick_sched_forward_next_period(void);
-#else
-static inline void tick_sched_forward_next_period(void) { }
-#endif
-
 /* NO_HZ_FULL internal */
 #ifdef CONFIG_NO_HZ_FULL
 extern void tick_nohz_init(void);

kernel/time/tick-sched.c

@@ -51,15 +51,6 @@ struct tick_sched *tick_get_tick_sched(int cpu)
  */
 static ktime_t last_jiffies_update;
 
-/*
- * Called after resume. Make sure that jiffies are not fast forwarded due to
- * clock monotonic being forwarded by the suspended time.
- */
-void tick_sched_forward_next_period(void)
-{
-	last_jiffies_update = tick_next_period;
-}
-
 /*
  * Must be called with interrupts disabled !
  */

kernel/time/timekeeping.c

@@ -138,12 +138,7 @@ static void tk_set_wall_to_mono(struct timekeeper *tk, struct timespec64 wtm)
 
 static inline void tk_update_sleep_time(struct timekeeper *tk, ktime_t delta)
 {
-	/* Update both bases so mono and raw stay coupled. */
-	tk->tkr_mono.base += delta;
-	tk->tkr_raw.base += delta;
-
-	/* Accumulate time spent in suspend */
-	tk->time_suspended += delta;
+	tk->offs_boot = ktime_add(tk->offs_boot, delta);
 }
 
 /*
@@ -473,6 +468,36 @@ u64 ktime_get_raw_fast_ns(void)
 }
 EXPORT_SYMBOL_GPL(ktime_get_raw_fast_ns);
 
+/**
+ * ktime_get_boot_fast_ns - NMI safe and fast access to boot clock.
+ *
+ * To keep it NMI safe since we're accessing from tracing, we're not using a
+ * separate timekeeper with updates to monotonic clock and boot offset
+ * protected with seqlocks. This has the following minor side effects:
+ *
+ * (1) Its possible that a timestamp be taken after the boot offset is updated
+ * but before the timekeeper is updated. If this happens, the new boot offset
+ * is added to the old timekeeping making the clock appear to update slightly
+ * earlier:
+ *    CPU 0                                        CPU 1
+ *    timekeeping_inject_sleeptime64()
+ *    __timekeeping_inject_sleeptime(tk, delta);
+ *                                                 timestamp();
+ *    timekeeping_update(tk, TK_CLEAR_NTP...);
+ *
+ * (2) On 32-bit systems, the 64-bit boot offset (tk->offs_boot) may be
+ * partially updated.  Since the tk->offs_boot update is a rare event, this
+ * should be a rare occurrence which postprocessing should be able to handle.
+ */
+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));
+}
+EXPORT_SYMBOL_GPL(ktime_get_boot_fast_ns);
+
+
 /*
  * See comment for __ktime_get_fast_ns() vs. timestamp ordering
  */
@@ -764,6 +789,7 @@ EXPORT_SYMBOL_GPL(ktime_get_resolution_ns);
 
 static ktime_t *offsets[TK_OFFS_MAX] = {
 	[TK_OFFS_REAL]	= &tk_core.timekeeper.offs_real,
+	[TK_OFFS_BOOT]	= &tk_core.timekeeper.offs_boot,
 	[TK_OFFS_TAI]	= &tk_core.timekeeper.offs_tai,
 };
 
@@ -860,39 +886,6 @@ void ktime_get_ts64(struct timespec64 *ts)
 }
 EXPORT_SYMBOL_GPL(ktime_get_ts64);
 
-/**
- * ktime_get_active_ts64 - Get the active non-suspended monotonic clock
- * @ts:		pointer to timespec variable
- *
- * The function calculates the monotonic clock from the realtime clock and
- * the wall_to_monotonic offset, subtracts the accumulated suspend time and
- * stores the result in normalized timespec64 format in the variable
- * pointed to by @ts.
- */
-void ktime_get_active_ts64(struct timespec64 *ts)
-{
-	struct timekeeper *tk = &tk_core.timekeeper;
-	struct timespec64 tomono, tsusp;
-	u64 nsec, nssusp;
-	unsigned int seq;
-
-	WARN_ON(timekeeping_suspended);
-
-	do {
-		seq = read_seqcount_begin(&tk_core.seq);
-		ts->tv_sec = tk->xtime_sec;
-		nsec = timekeeping_get_ns(&tk->tkr_mono);
-		tomono = tk->wall_to_monotonic;
-		nssusp = tk->time_suspended;
-	} while (read_seqcount_retry(&tk_core.seq, seq));
-
-	ts->tv_sec += tomono.tv_sec;
-	ts->tv_nsec = 0;
-	timespec64_add_ns(ts, nsec + tomono.tv_nsec);
-	tsusp = ns_to_timespec64(nssusp);
-	*ts = timespec64_sub(*ts, tsusp);
-}
-
 /**
  * ktime_get_seconds - Get the seconds portion of CLOCK_MONOTONIC
  *
@@ -1593,6 +1586,7 @@ static void __timekeeping_inject_sleeptime(struct timekeeper *tk,
 		return;
 	}
 	tk_xtime_add(tk, delta);
+	tk_set_wall_to_mono(tk, timespec64_sub(tk->wall_to_monotonic, *delta));
 	tk_update_sleep_time(tk, timespec64_to_ktime(*delta));
 	tk_debug_account_sleep_time(delta);
 }
@@ -2125,7 +2119,7 @@ out:
 void getboottime64(struct timespec64 *ts)
 {
 	struct timekeeper *tk = &tk_core.timekeeper;
-	ktime_t t = ktime_sub(tk->offs_real, tk->time_suspended);
+	ktime_t t = ktime_sub(tk->offs_real, tk->offs_boot);
 
 	*ts = ktime_to_timespec64(t);
 }
@@ -2188,6 +2182,7 @@ void do_timer(unsigned long ticks)
  * ktime_get_update_offsets_now - hrtimer helper
  * @cwsseq:	pointer to check and store the clock was set sequence number
  * @offs_real:	pointer to storage for monotonic -> realtime offset
+ * @offs_boot:	pointer to storage for monotonic -> boottime offset
  * @offs_tai:	pointer to storage for monotonic -> clock tai offset
  *
  * Returns current monotonic time and updates the offsets if the
@@ -2197,7 +2192,7 @@ void do_timer(unsigned long ticks)
  * Called from hrtimer_interrupt() or retrigger_next_event()
  */
 ktime_t ktime_get_update_offsets_now(unsigned int *cwsseq, ktime_t *offs_real,
-				     ktime_t *offs_tai)
+				     ktime_t *offs_boot, ktime_t *offs_tai)
 {
 	struct timekeeper *tk = &tk_core.timekeeper;
 	unsigned int seq;
@@ -2214,6 +2209,7 @@ ktime_t ktime_get_update_offsets_now(unsigned int *cwsseq, ktime_t *offs_real,
 		if (*cwsseq != tk->clock_was_set_seq) {
 			*cwsseq = tk->clock_was_set_seq;
 			*offs_real = tk->offs_real;
+			*offs_boot = tk->offs_boot;
 			*offs_tai = tk->offs_tai;
 		}
 

kernel/time/timekeeping.h

@@ -6,6 +6,7 @@
  */
 extern ktime_t ktime_get_update_offsets_now(unsigned int *cwsseq,
 					    ktime_t *offs_real,
+					    ktime_t *offs_boot,
 					    ktime_t *offs_tai);
 
 extern int timekeeping_valid_for_hres(void);

kernel/trace/trace.c

@@ -1165,7 +1165,7 @@ static struct {
 	{ trace_clock,			"perf",		1 },
 	{ ktime_get_mono_fast_ns,	"mono",		1 },
 	{ ktime_get_raw_fast_ns,	"mono_raw",	1 },
-	{ ktime_get_mono_fast_ns,	"boot",		1 },
+	{ ktime_get_boot_fast_ns,	"boot",		1 },
 	ARCH_TRACE_CLOCKS
 };