1 files changed, 499 insertions, 286 deletions

diff --git a/kernel/time/timer_migration.c b/kernel/time/timer_migration.c
index 84413114db5c..18dda1aa782d 100644
--- a/kernel/time/timer_migration.c
+++ b/kernel/time/timer_migration.c
@@ -10,6 +10,7 @@
 #include <linux/spinlock.h>
 #include <linux/timerqueue.h>
 #include <trace/events/ipi.h>
+#include <linux/sched/isolation.h>
 
 #include "timer_migration.h"
 #include "tick-internal.h"
@@ -420,14 +421,54 @@ static struct list_head *tmigr_level_list __read_mostly;
 static unsigned int tmigr_hierarchy_levels __read_mostly;
 static unsigned int tmigr_crossnode_level __read_mostly;
 
+static struct tmigr_group *tmigr_root;
+
 static DEFINE_PER_CPU(struct tmigr_cpu, tmigr_cpu);
 
+/*
+ * CPUs available for timer migration.
+ * Protected by cpuset_mutex (with cpus_read_lock held) or cpus_write_lock.
+ * Additionally tmigr_available_mutex serializes set/clear operations with each other.
+ */
+static cpumask_var_t tmigr_available_cpumask;
+static DEFINE_MUTEX(tmigr_available_mutex);
+
+/* Enabled during late initcall */
+static DEFINE_STATIC_KEY_FALSE(tmigr_exclude_isolated);
+
 #define TMIGR_NONE	0xFF
 #define BIT_CNT		8
 
 static inline bool tmigr_is_not_available(struct tmigr_cpu *tmc)
 {
-	return !(tmc->tmgroup && tmc->online);
+	return !(tmc->tmgroup && tmc->available);
+}
+
+/*
+ * Returns true if @cpu should be excluded from the hierarchy as isolated.
+ * Domain isolated CPUs don't participate in timer migration, nohz_full CPUs
+ * are still part of the hierarchy but become idle (from a tick and timer
+ * migration perspective) when they stop their tick. This lets the timekeeping
+ * CPU handle their global timers. Marking also isolated CPUs as idle would be
+ * too costly, hence they are completely excluded from the hierarchy.
+ * This check is necessary, for instance, to prevent offline isolated CPUs from
+ * being incorrectly marked as available once getting back online.
+ *
+ * This function returns false during early boot and the isolation logic is
+ * enabled only after isolated CPUs are marked as unavailable at late boot.
+ * The tick CPU can be isolated at boot, however we cannot mark it as
+ * unavailable to avoid having no global migrator for the nohz_full CPUs. This
+ * should be ensured by the callers of this function: implicitly from hotplug
+ * callbacks and explicitly in tmigr_init_isolation() and
+ * tmigr_isolated_exclude_cpumask().
+ */
+static inline bool tmigr_is_isolated(int cpu)
+{
+	if (!static_branch_unlikely(&tmigr_exclude_isolated))
+		return false;
+	return (!housekeeping_cpu(cpu, HK_TYPE_DOMAIN) ||
+		cpuset_cpu_is_isolated(cpu)) &&
+	       housekeeping_cpu(cpu, HK_TYPE_KERNEL_NOISE);
 }
 
 /*
@@ -475,44 +516,33 @@ static bool tmigr_check_lonely(struct tmigr_group *group)
 	return bitmap_weight(&active, BIT_CNT) <= 1;
 }
 
-typedef bool (*up_f)(struct tmigr_group *, struct tmigr_group *, void *);
-
-static void __walk_groups(up_f up, void *data,
-			  struct tmigr_cpu *tmc)
-{
-	struct tmigr_group *child = NULL, *group = tmc->tmgroup;
-
-	do {
-		WARN_ON_ONCE(group->level >= tmigr_hierarchy_levels);
-
-		if (up(group, child, data))
-			break;
-
-		child = group;
-		group = group->parent;
-	} while (group);
-}
-
-static void walk_groups(up_f up, void *data, struct tmigr_cpu *tmc)
-{
-	lockdep_assert_held(&tmc->lock);
-
-	__walk_groups(up, data, tmc);
-}
-
 /**
  * struct tmigr_walk - data required for walking the hierarchy
  * @nextexp:		Next CPU event expiry information which is handed into
  *			the timer migration code by the timer code
  *			(get_next_timer_interrupt())
- * @firstexp:		Contains the first event expiry information when last
- *			active CPU of hierarchy is on the way to idle to make
- *			sure CPU will be back in time.
+ * @firstexp:		Contains the first event expiry information when
+ *			hierarchy is completely idle.  When CPU itself was the
+ *			last going idle, information makes sure, that CPU will
+ *			be back in time. When using this value in the remote
+ *			expiry case, firstexp is stored in the per CPU tmigr_cpu
+ *			struct of CPU which expires remote timers. It is updated
+ *			in top level group only. Be aware, there could occur a
+ *			new top level of the hierarchy between the 'top level
+ *			call' in tmigr_update_events() and the check for the
+ *			parent group in walk_groups(). Then @firstexp might
+ *			contain a value != KTIME_MAX even if it was not the
+ *			final top level. This is not a problem, as the worst
+ *			outcome is a CPU which might wake up a little early.
  * @evt:		Pointer to tmigr_event which needs to be queued (of idle
  *			child group)
- * @childmask:		childmask of child group
+ * @childmask:		groupmask of child group
  * @remote:		Is set, when the new timer path is executed in
  *			tmigr_handle_remote_cpu()
+ * @basej:		timer base in jiffies
+ * @now:		timer base monotonic
+ * @check:		is set if there is the need to handle remote timers;
+ *			required in tmigr_requires_handle_remote() only
  */
 struct tmigr_walk {
 	u64			nextexp;
@@ -520,33 +550,45 @@ struct tmigr_walk {
 	struct tmigr_event	*evt;
 	u8			childmask;
 	bool			remote;
+	unsigned long		basej;
+	u64			now;
+	bool			check;
 };
 
-/**
- * struct tmigr_remote_data - data required for remote expiry hierarchy walk
- * @basej:		timer base in jiffies
- * @now:		timer base monotonic
- * @firstexp:		returns expiry of the first timer in the idle timer
- *			migration hierarchy to make sure the timer is handled in
- *			time; it is stored in the per CPU tmigr_cpu struct of
- *			CPU which expires remote timers
- * @childmask:		childmask of child group
- * @check:		is set if there is the need to handle remote timers;
- *			required in tmigr_requires_handle_remote() only
- * @tmc_active:		this flag indicates, whether the CPU which triggers
- *			the hierarchy walk is !idle in the timer migration
- *			hierarchy. When the CPU is idle and the whole hierarchy is
- *			idle, only the first event of the top level has to be
- *			considered.
- */
-struct tmigr_remote_data {
-	unsigned long	basej;
-	u64		now;
-	u64		firstexp;
-	u8		childmask;
-	bool		check;
-	bool		tmc_active;
-};
+typedef bool (*up_f)(struct tmigr_group *, struct tmigr_group *, struct tmigr_walk *);
+
+static void __walk_groups_from(up_f up, struct tmigr_walk *data,
+			       struct tmigr_group *child, struct tmigr_group *group)
+{
+	do {
+		WARN_ON_ONCE(group->level >= tmigr_hierarchy_levels);
+
+		if (up(group, child, data))
+			break;
+
+		child = group;
+		/*
+		 * Pairs with the store release on group connection
+		 * to make sure group initialization is visible.
+		 */
+		group = READ_ONCE(group->parent);
+		data->childmask = child->groupmask;
+		WARN_ON_ONCE(!data->childmask);
+	} while (group);
+}
+
+static void __walk_groups(up_f up, struct tmigr_walk *data,
+			  struct tmigr_cpu *tmc)
+{
+	__walk_groups_from(up, data, NULL, tmc->tmgroup);
+}
+
+static void walk_groups(up_f up, struct tmigr_walk *data, struct tmigr_cpu *tmc)
+{
+	lockdep_assert_held(&tmc->lock);
+
+	__walk_groups(up, data, tmc);
+}
 
 /*
  * Returns the next event of the timerqueue @group->events
@@ -566,7 +608,7 @@ static struct tmigr_event *tmigr_next_groupevt(struct tmigr_group *group)
 	while ((node = timerqueue_getnext(&group->events))) {
 		evt = container_of(node, struct tmigr_event, nextevt);
 
-		if (!evt->ignore) {
+		if (!READ_ONCE(evt->ignore)) {
 			WRITE_ONCE(group->next_expiry, evt->nextevt.expires);
 			return evt;
 		}
@@ -618,10 +660,9 @@ static u64 tmigr_next_groupevt_expires(struct tmigr_group *group)
 
 static bool tmigr_active_up(struct tmigr_group *group,
 			    struct tmigr_group *child,
-			    void *ptr)
+			    struct tmigr_walk *data)
 {
 	union tmigr_state curstate, newstate;
-	struct tmigr_walk *data = ptr;
 	bool walk_done;
 	u8 childmask;
 
@@ -649,8 +690,7 @@ static bool tmigr_active_up(struct tmigr_group *group,
 
 	} while (!atomic_try_cmpxchg(&group->migr_state, &curstate.state, newstate.state));
 
-	if ((walk_done == false) && group->parent)
-		data->childmask = group->childmask;
+	trace_tmigr_group_set_cpu_active(group, newstate, childmask);
 
 	/*
 	 * The group is active (again). The group event might be still queued
@@ -664,9 +704,7 @@ static bool tmigr_active_up(struct tmigr_group *group,
 	 * lock is held while updating the ignore flag in idle path. So this
 	 * state change will not be lost.
 	 */
-	group->groupevt.ignore = true;
-
-	trace_tmigr_group_set_cpu_active(group, newstate, childmask);
+	WRITE_ONCE(group->groupevt.ignore, true);
 
 	return walk_done;
 }
@@ -675,7 +713,7 @@ static void __tmigr_cpu_activate(struct tmigr_cpu *tmc)
 {
 	struct tmigr_walk data;
 
-	data.childmask = tmc->childmask;
+	data.childmask = tmc->groupmask;
 
 	trace_tmigr_cpu_active(tmc);
 
@@ -709,7 +747,7 @@ void tmigr_cpu_activate(void)
 /*
  * Returns true, if there is nothing to be propagated to the next level
  *
- * @data->firstexp is set to expiry of first gobal event of the (top level of
+ * @data->firstexp is set to expiry of first global event of the (top level of
  * the) hierarchy, but only when hierarchy is completely idle.
  *
  * The child and group states need to be read under the lock, to prevent a race
@@ -727,6 +765,7 @@ bool tmigr_update_events(struct tmigr_group *group, struct tmigr_group *child,
 	union tmigr_state childstate, groupstate;
 	bool remote = data->remote;
 	bool walk_done = false;
+	bool ignore;
 	u64 nextexp;
 
 	if (child) {
@@ -745,11 +784,19 @@ bool tmigr_update_events(struct tmigr_group *group, struct tmigr_group *child,
 		nextexp = child->next_expiry;
 		evt = &child->groupevt;
 
-		evt->ignore = (nextexp == KTIME_MAX) ? true : false;
+		/*
+		 * This can race with concurrent idle exit (activate).
+		 * If the current writer wins, a useless remote expiration may
+		 * be scheduled. If the activate wins, the event is properly
+		 * ignored.
+		 */
+		ignore = (nextexp == KTIME_MAX) ? true : false;
+		WRITE_ONCE(evt->ignore, ignore);
 	} else {
 		nextexp = data->nextexp;
 
 		first_childevt = evt = data->evt;
+		ignore = evt->ignore;
 
 		/*
 		 * Walking the hierarchy is required in any case when a
@@ -775,7 +822,7 @@ bool tmigr_update_events(struct tmigr_group *group, struct tmigr_group *child,
 		 * first event information of the group is updated properly and
 		 * also handled properly, so skip this fast return path.
 		 */
-		if (evt->ignore && !remote && group->parent)
+		if (ignore && !remote && group->parent)
 			return true;
 
 		raw_spin_lock(&group->lock);
@@ -789,7 +836,7 @@ bool tmigr_update_events(struct tmigr_group *group, struct tmigr_group *child,
 	 * queue when the expiry time changed only or when it could be ignored.
 	 */
 	if (timerqueue_node_queued(&evt->nextevt)) {
-		if ((evt->nextevt.expires == nextexp) && !evt->ignore) {
+		if ((evt->nextevt.expires == nextexp) && !ignore) {
 			/* Make sure not to miss a new CPU event with the same expiry */
 			evt->cpu = first_childevt->cpu;
 			goto check_toplvl;
@@ -799,7 +846,7 @@ bool tmigr_update_events(struct tmigr_group *group, struct tmigr_group *child,
 			WRITE_ONCE(group->next_expiry, KTIME_MAX);
 	}
 
-	if (evt->ignore) {
+	if (ignore) {
 		/*
 		 * When the next child event could be ignored (nextexp is
 		 * KTIME_MAX) and there was no remote timer handling before or
@@ -860,10 +907,8 @@ unlock:
 
 static bool tmigr_new_timer_up(struct tmigr_group *group,
 			       struct tmigr_group *child,
-			       void *ptr)
+			       struct tmigr_walk *data)
 {
-	struct tmigr_walk *data = ptr;
-
 	return tmigr_update_events(group, child, data);
 }
 
@@ -920,7 +965,7 @@ static void tmigr_handle_remote_cpu(unsigned int cpu, u64 now,
 	 * updated the event takes care when hierarchy is completely
 	 * idle. Otherwise the migrator does it as the event is enqueued.
 	 */
-	if (!tmc->online || tmc->remote || tmc->cpuevt.ignore ||
+	if (!tmc->available || tmc->remote || tmc->cpuevt.ignore ||
 	    now < tmc->cpuevt.nextevt.expires) {
 		raw_spin_unlock_irq(&tmc->lock);
 		return;
@@ -967,7 +1012,7 @@ static void tmigr_handle_remote_cpu(unsigned int cpu, u64 now,
 	 * (See also section "Required event and timerqueue update after a
 	 * remote expiry" in the documentation at the top)
 	 */
-	if (!tmc->online || !tmc->idle) {
+	if (!tmc->available || !tmc->idle) {
 		timer_unlock_remote_bases(cpu);
 		goto unlock;
 	}
@@ -995,9 +1040,8 @@ unlock:
 
 static bool tmigr_handle_remote_up(struct tmigr_group *group,
 				   struct tmigr_group *child,
-				   void *ptr)
+				   struct tmigr_walk *data)
 {
-	struct tmigr_remote_data *data = ptr;
 	struct tmigr_event *evt;
 	unsigned long jif;
 	u8 childmask;
@@ -1034,12 +1078,10 @@ again:
 	}
 
 	/*
-	 * Update of childmask for the next level and keep track of the expiry
-	 * of the first event that needs to be handled (group->next_expiry was
-	 * updated by tmigr_next_expired_groupevt(), next was set by
-	 * tmigr_handle_remote_cpu()).
+	 * Keep track of the expiry of the first event that needs to be handled
+	 * (group->next_expiry was updated by tmigr_next_expired_groupevt(),
+	 * next was set by tmigr_handle_remote_cpu()).
 	 */
-	data->childmask = group->childmask;
 	data->firstexp = group->next_expiry;
 
 	raw_spin_unlock_irq(&group->lock);
@@ -1055,12 +1097,12 @@ again:
 void tmigr_handle_remote(void)
 {
 	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-	struct tmigr_remote_data data;
+	struct tmigr_walk data;
 
 	if (tmigr_is_not_available(tmc))
 		return;
 
-	data.childmask = tmc->childmask;
+	data.childmask = tmc->groupmask;
 	data.firstexp = KTIME_MAX;
 
 	/*
@@ -1068,7 +1110,7 @@ void tmigr_handle_remote(void)
 	 * in tmigr_handle_remote_up() anyway. Keep this check to speed up the
 	 * return when nothing has to be done.
 	 */
-	if (!tmigr_check_migrator(tmc->tmgroup, tmc->childmask)) {
+	if (!tmigr_check_migrator(tmc->tmgroup, tmc->groupmask)) {
 		/*
 		 * If this CPU was an idle migrator, make sure to clear its wakeup
 		 * value so it won't chase timers that have already expired elsewhere.
@@ -1097,9 +1139,8 @@ void tmigr_handle_remote(void)
 
 static bool tmigr_requires_handle_remote_up(struct tmigr_group *group,
 					    struct tmigr_group *child,
-					    void *ptr)
+					    struct tmigr_walk *data)
 {
-	struct tmigr_remote_data *data = ptr;
 	u8 childmask;
 
 	childmask = data->childmask;
@@ -1111,15 +1152,6 @@ static bool tmigr_requires_handle_remote_up(struct tmigr_group *group,
 	 */
 	if (!tmigr_check_migrator(group, childmask))
 		return true;
-
-	/*
-	 * When there is a parent group and the CPU which triggered the
-	 * hierarchy walk is not active, proceed the walk to reach the top level
-	 * group before reading the next_expiry value.
-	 */
-	if (group->parent && !data->tmc_active)
-		goto out;
-
 	/*
 	 * The lock is required on 32bit architectures to read the variable
 	 * consistently with a concurrent writer. On 64bit the lock is not
@@ -1143,9 +1175,6 @@ static bool tmigr_requires_handle_remote_up(struct tmigr_group *group,
 		raw_spin_unlock(&group->lock);
 	}
 
-out:
-	/* Update of childmask for the next level */
-	data->childmask = group->childmask;
 	return false;
 }
 
@@ -1157,7 +1186,7 @@ out:
 bool tmigr_requires_handle_remote(void)
 {
 	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-	struct tmigr_remote_data data;
+	struct tmigr_walk data;
 	unsigned long jif;
 	bool ret = false;
 
@@ -1165,9 +1194,8 @@ bool tmigr_requires_handle_remote(void)
 		return ret;
 
 	data.now = get_jiffies_update(&jif);
-	data.childmask = tmc->childmask;
+	data.childmask = tmc->groupmask;
 	data.firstexp = KTIME_MAX;
-	data.tmc_active = !tmc->idle;
 	data.check = false;
 
 	/*
@@ -1230,14 +1258,13 @@ u64 tmigr_cpu_new_timer(u64 nextexp)
 		if (nextexp != tmc->cpuevt.nextevt.expires ||
 		    tmc->cpuevt.ignore) {
 			ret = tmigr_new_timer(tmc, nextexp);
+			/*
+			 * Make sure the reevaluation of timers in idle path
+			 * will not miss an event.
+			 */
+			WRITE_ONCE(tmc->wakeup, ret);
 		}
 	}
-	/*
-	 * Make sure the reevaluation of timers in idle path will not miss an
-	 * event.
-	 */
-	WRITE_ONCE(tmc->wakeup, ret);
-
 	trace_tmigr_cpu_new_timer_idle(tmc, nextexp);
 	raw_spin_unlock(&tmc->lock);
 	return ret;
@@ -1245,10 +1272,9 @@ u64 tmigr_cpu_new_timer(u64 nextexp)
 
 static bool tmigr_inactive_up(struct tmigr_group *group,
 			      struct tmigr_group *child,
-			      void *ptr)
+			      struct tmigr_walk *data)
 {
 	union tmigr_state curstate, newstate, childstate;
-	struct tmigr_walk *data = ptr;
 	bool walk_done;
 	u8 childmask;
 
@@ -1299,9 +1325,10 @@ static bool tmigr_inactive_up(struct tmigr_group *group,
 
 		WARN_ON_ONCE((newstate.migrator != TMIGR_NONE) && !(newstate.active));
 
-		if (atomic_try_cmpxchg(&group->migr_state, &curstate.state,
-				       newstate.state))
+		if (atomic_try_cmpxchg(&group->migr_state, &curstate.state, newstate.state)) {
+			trace_tmigr_group_set_cpu_inactive(group, newstate, childmask);
 			break;
+		}
 
 		/*
 		 * The memory barrier is paired with the cmpxchg() in
@@ -1317,22 +1344,6 @@ static bool tmigr_inactive_up(struct tmigr_group *group,
 	/* Event Handling */
 	tmigr_update_events(group, child, data);
 
-	if (group->parent && (walk_done == false))
-		data->childmask = group->childmask;
-
-	/*
-	 * data->firstexp was set by tmigr_update_events() and contains the
-	 * expiry of the first global event which needs to be handled. It
-	 * differs from KTIME_MAX if:
-	 * - group is the top level group and
-	 * - group is idle (which means CPU was the last active CPU in the
-	 *   hierarchy) and
-	 * - there is a pending event in the hierarchy
-	 */
-	WARN_ON_ONCE(data->firstexp != KTIME_MAX && group->parent);
-
-	trace_tmigr_group_set_cpu_inactive(group, newstate, childmask);
-
 	return walk_done;
 }
 
@@ -1341,7 +1352,7 @@ static u64 __tmigr_cpu_deactivate(struct tmigr_cpu *tmc, u64 nextexp)
 	struct tmigr_walk data = { .nextexp = nextexp,
 				   .firstexp = KTIME_MAX,
 				   .evt = &tmc->cpuevt,
-				   .childmask = tmc->childmask };
+				   .childmask = tmc->groupmask };
 
 	/*
 	 * If nextexp is KTIME_MAX, the CPU event will be ignored because the
@@ -1400,7 +1411,7 @@ u64 tmigr_cpu_deactivate(u64 nextexp)
  *			  the only one in the level 0 group; and if it is the
  *			  only one in level 0 group, but there are more than a
  *			  single group active on the way to top level)
- * * nextevt		- when CPU is offline and has to handle timer on his own
+ * * nextevt		- when CPU is offline and has to handle timer on its own
  *			  or when on the way to top in every group only a single
  *			  child is active but @nextevt is before the lowest
  *			  next_expiry encountered while walking up to top level.
@@ -1419,29 +1430,198 @@ u64 tmigr_quick_check(u64 nextevt)
 	if (WARN_ON_ONCE(tmc->idle))
 		return nextevt;
 
-	if (!tmigr_check_migrator_and_lonely(tmc->tmgroup, tmc->childmask))
+	if (!tmigr_check_migrator_and_lonely(tmc->tmgroup, tmc->groupmask))
 		return KTIME_MAX;
 
 	do {
-		if (!tmigr_check_lonely(group)) {
+		if (!tmigr_check_lonely(group))
 			return KTIME_MAX;
-		} else {
-			/*
-			 * Since current CPU is active, events may not be sorted
-			 * from bottom to the top because the CPU's event is ignored
-			 * up to the top and its sibling's events not propagated upwards.
-			 * Thus keep track of the lowest observed expiry.
-			 */
-			nextevt = min_t(u64, nextevt, READ_ONCE(group->next_expiry));
-			if (!group->parent)
-				return nextevt;
-		}
+
+		/*
+		 * Since current CPU is active, events may not be sorted
+		 * from bottom to the top because the CPU's event is ignored
+		 * up to the top and its sibling's events not propagated upwards.
+		 * Thus keep track of the lowest observed expiry.
+		 */
+		nextevt = min_t(u64, nextevt, READ_ONCE(group->next_expiry));
 		group = group->parent;
 	} while (group);
 
-	return KTIME_MAX;
+	return nextevt;
+}
+
+/*
+ * tmigr_trigger_active() - trigger a CPU to become active again
+ *
+ * This function is executed on a CPU which is part of cpu_online_mask, when the
+ * last active CPU in the hierarchy is offlining. With this, it is ensured that
+ * the other CPU is active and takes over the migrator duty.
+ */
+static long tmigr_trigger_active(void *unused)
+{
+	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
+
+	WARN_ON_ONCE(!tmc->available || tmc->idle);
+
+	return 0;
+}
+
+static int tmigr_clear_cpu_available(unsigned int cpu)
+{
+	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
+	int migrator;
+	u64 firstexp;
+
+	guard(mutex)(&tmigr_available_mutex);
+
+	cpumask_clear_cpu(cpu, tmigr_available_cpumask);
+	scoped_guard(raw_spinlock_irq, &tmc->lock) {
+		if (!tmc->available)
+			return 0;
+		tmc->available = false;
+		WRITE_ONCE(tmc->wakeup, KTIME_MAX);
+
+		/*
+		 * CPU has to handle the local events on his own, when on the way to
+		 * offline; Therefore nextevt value is set to KTIME_MAX
+		 */
+		firstexp = __tmigr_cpu_deactivate(tmc, KTIME_MAX);
+		trace_tmigr_cpu_unavailable(tmc);
+	}
+
+	if (firstexp != KTIME_MAX) {
+		migrator = cpumask_any(tmigr_available_cpumask);
+		work_on_cpu(migrator, tmigr_trigger_active, NULL);
+	}
+
+	return 0;
+}
+
+static int tmigr_set_cpu_available(unsigned int cpu)
+{
+	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
+
+	/* Check whether CPU data was successfully initialized */
+	if (WARN_ON_ONCE(!tmc->tmgroup))
+		return -EINVAL;
+
+	if (tmigr_is_isolated(cpu))
+		return 0;
+
+	guard(mutex)(&tmigr_available_mutex);
+
+	cpumask_set_cpu(cpu, tmigr_available_cpumask);
+	scoped_guard(raw_spinlock_irq, &tmc->lock) {
+		if (tmc->available)
+			return 0;
+		trace_tmigr_cpu_available(tmc);
+		tmc->idle = timer_base_is_idle();
+		if (!tmc->idle)
+			__tmigr_cpu_activate(tmc);
+		tmc->available = true;
+	}
+	return 0;
+}
+
+static void tmigr_cpu_isolate(struct work_struct *ignored)
+{
+	tmigr_clear_cpu_available(smp_processor_id());
+}
+
+static void tmigr_cpu_unisolate(struct work_struct *ignored)
+{
+	tmigr_set_cpu_available(smp_processor_id());
 }
 
+/**
+ * tmigr_isolated_exclude_cpumask - Exclude given CPUs from hierarchy
+ * @exclude_cpumask: the cpumask to be excluded from timer migration hierarchy
+ *
+ * This function can be called from cpuset code to provide the new set of
+ * isolated CPUs that should be excluded from the hierarchy.
+ * Online CPUs not present in exclude_cpumask but already excluded are brought
+ * back to the hierarchy.
+ * Functions to isolate/unisolate need to be called locally and can sleep.
+ */
+int tmigr_isolated_exclude_cpumask(struct cpumask *exclude_cpumask)
+{
+	struct work_struct __percpu *works __free(free_percpu) =
+		alloc_percpu(struct work_struct);
+	cpumask_var_t cpumask __free(free_cpumask_var) = CPUMASK_VAR_NULL;
+	int cpu;
+
+	lockdep_assert_cpus_held();
+
+	if (!works)
+		return -ENOMEM;
+	if (!alloc_cpumask_var(&cpumask, GFP_KERNEL))
+		return -ENOMEM;
+
+	/*
+	 * First set previously isolated CPUs as available (unisolate).
+	 * This cpumask contains only CPUs that switched to available now.
+	 */
+	cpumask_andnot(cpumask, cpu_online_mask, exclude_cpumask);
+	cpumask_andnot(cpumask, cpumask, tmigr_available_cpumask);
+
+	for_each_cpu(cpu, cpumask) {
+		struct work_struct *work = per_cpu_ptr(works, cpu);
+
+		INIT_WORK(work, tmigr_cpu_unisolate);
+		schedule_work_on(cpu, work);
+	}
+	for_each_cpu(cpu, cpumask)
+		flush_work(per_cpu_ptr(works, cpu));
+
+	/*
+	 * Then clear previously available CPUs (isolate).
+	 * This cpumask contains only CPUs that switched to not available now.
+	 * There cannot be overlap with the newly available ones.
+	 */
+	cpumask_and(cpumask, exclude_cpumask, tmigr_available_cpumask);
+	cpumask_and(cpumask, cpumask, housekeeping_cpumask(HK_TYPE_KERNEL_NOISE));
+	/*
+	 * Handle this here and not in the cpuset code because exclude_cpumask
+	 * might include also the tick CPU if included in isolcpus.
+	 */
+	for_each_cpu(cpu, cpumask) {
+		if (!tick_nohz_cpu_hotpluggable(cpu)) {
+			cpumask_clear_cpu(cpu, cpumask);
+			break;
+		}
+	}
+
+	for_each_cpu(cpu, cpumask) {
+		struct work_struct *work = per_cpu_ptr(works, cpu);
+
+		INIT_WORK(work, tmigr_cpu_isolate);
+		schedule_work_on(cpu, work);
+	}
+	for_each_cpu(cpu, cpumask)
+		flush_work(per_cpu_ptr(works, cpu));
+
+	return 0;
+}
+
+static int __init tmigr_init_isolation(void)
+{
+	cpumask_var_t cpumask __free(free_cpumask_var) = CPUMASK_VAR_NULL;
+
+	static_branch_enable(&tmigr_exclude_isolated);
+
+	if (!housekeeping_enabled(HK_TYPE_DOMAIN))
+		return 0;
+	if (!alloc_cpumask_var(&cpumask, GFP_KERNEL))
+		return -ENOMEM;
+
+	cpumask_andnot(cpumask, cpu_possible_mask, housekeeping_cpumask(HK_TYPE_DOMAIN));
+
+	/* Protect against RCU torture hotplug testing */
+	guard(cpus_read_lock)();
+	return tmigr_isolated_exclude_cpumask(cpumask);
+}
+late_initcall(tmigr_init_isolation);
+
 static void tmigr_init_group(struct tmigr_group *group, unsigned int lvl,
 			     int node)
 {
@@ -1466,8 +1646,7 @@ static void tmigr_init_group(struct tmigr_group *group, unsigned int lvl,
 	group->groupevt.ignore = true;
 }
 
-static struct tmigr_group *tmigr_get_group(unsigned int cpu, int node,
-					   unsigned int lvl)
+static struct tmigr_group *tmigr_get_group(int node, unsigned int lvl)
 {
 	struct tmigr_group *tmp, *group = NULL;
 
@@ -1513,91 +1692,116 @@ static struct tmigr_group *tmigr_get_group(unsigned int cpu, int node,
 	return group;
 }
 
-static void tmigr_connect_child_parent(struct tmigr_group *child,
-				       struct tmigr_group *parent)
+static bool tmigr_init_root(struct tmigr_group *group, bool activate)
 {
-	union tmigr_state childstate;
-
-	raw_spin_lock_irq(&child->lock);
-	raw_spin_lock_nested(&parent->lock, SINGLE_DEPTH_NESTING);
-
-	child->parent = parent;
-	child->childmask = BIT(parent->num_children++);
+	if (!group->parent && group != tmigr_root) {
+		/*
+		 * This is the new top-level, prepare its groupmask in advance
+		 * to avoid accidents where yet another new top-level is
+		 * created in the future and made visible before this groupmask.
+		 */
+		group->groupmask = BIT(0);
+		WARN_ON_ONCE(activate);
 
-	raw_spin_unlock(&parent->lock);
-	raw_spin_unlock_irq(&child->lock);
+		return true;
+	}
 
-	trace_tmigr_connect_child_parent(child);
+	return false;
 
-	/*
-	 * To prevent inconsistent states, active children need to be active in
-	 * the new parent as well. Inactive children are already marked inactive
-	 * in the parent group:
-	 *
-	 * * When new groups were created by tmigr_setup_groups() starting from
-	 *   the lowest level (and not higher then one level below the current
-	 *   top level), then they are not active. They will be set active when
-	 *   the new online CPU comes active.
-	 *
-	 * * But if a new group above the current top level is required, it is
-	 *   mandatory to propagate the active state of the already existing
-	 *   child to the new parent. So tmigr_connect_child_parent() is
-	 *   executed with the formerly top level group (child) and the newly
-	 *   created group (parent).
-	 */
-	childstate.state = atomic_read(&child->migr_state);
-	if (childstate.migrator != TMIGR_NONE) {
-		struct tmigr_walk data;
+}
 
-		data.childmask = child->childmask;
+static void tmigr_connect_child_parent(struct tmigr_group *child,
+				       struct tmigr_group *parent,
+				       bool activate)
+{
+	if (tmigr_init_root(parent, activate)) {
+		/*
+		 * The previous top level had prepared its groupmask already,
+		 * simply account it in advance as the first child. If some groups
+		 * have been created between the old and new root due to node
+		 * mismatch, the new root's child will be intialized accordingly.
+		 */
+		parent->num_children = 1;
+	}
 
+	/* Connecting old root to new root ? */
+	if (!parent->parent && activate) {
 		/*
-		 * There is only one new level per time. When connecting the
-		 * child and the parent and set the child active when the parent
-		 * is inactive, the parent needs to be the uppermost
-		 * level. Otherwise there went something wrong!
+		 * @child is the old top, or in case of node mismatch, some
+		 * intermediate group between the old top and the new one in
+		 * @parent. In this case the @child must be pre-accounted above
+		 * as the first child. Its new inactive sibling corresponding
+		 * to the CPU going up has been accounted as the second child.
 		 */
-		WARN_ON(!tmigr_active_up(parent, child, &data) && parent->parent);
+		WARN_ON_ONCE(parent->num_children != 2);
+		child->groupmask = BIT(0);
+	} else {
+		/* Common case adding @child for the CPU going up to @parent. */
+		child->groupmask = BIT(parent->num_children++);
 	}
+
+	/*
+	 * Make sure parent initialization is visible before publishing it to a
+	 * racing CPU entering/exiting idle. This RELEASE barrier enforces an
+	 * address dependency that pairs with the READ_ONCE() in __walk_groups().
+	 */
+	smp_store_release(&child->parent, parent);
+
+	trace_tmigr_connect_child_parent(child);
 }
 
-static int tmigr_setup_groups(unsigned int cpu, unsigned int node)
+static int tmigr_setup_groups(unsigned int cpu, unsigned int node,
+			      struct tmigr_group *start, bool activate)
 {
 	struct tmigr_group *group, *child, **stack;
-	int top = 0, err = 0, i = 0;
-	struct list_head *lvllist;
+	int i, top = 0, err = 0, start_lvl = 0;
+	bool root_mismatch = false;
 
 	stack = kcalloc(tmigr_hierarchy_levels, sizeof(*stack), GFP_KERNEL);
 	if (!stack)
 		return -ENOMEM;
 
-	do {
-		group = tmigr_get_group(cpu, node, i);
+	if (start) {
+		stack[start->level] = start;
+		start_lvl = start->level + 1;
+	}
+
+	if (tmigr_root)
+		root_mismatch = tmigr_root->numa_node != node;
+
+	for (i = start_lvl; i < tmigr_hierarchy_levels; i++) {
+		group = tmigr_get_group(node, i);
 		if (IS_ERR(group)) {
 			err = PTR_ERR(group);
+			i--;
 			break;
 		}
 
 		top = i;
-		stack[i++] = group;
+		stack[i] = group;
 
 		/*
 		 * When booting only less CPUs of a system than CPUs are
-		 * available, not all calculated hierarchy levels are required.
+		 * available, not all calculated hierarchy levels are required,
+		 * unless a node mismatch is detected.
 		 *
 		 * The loop is aborted as soon as the highest level, which might
 		 * be different from tmigr_hierarchy_levels, contains only a
-		 * single group.
+		 * single group, unless the nodes mismatch below tmigr_crossnode_level
 		 */
-		if (group->parent || i == tmigr_hierarchy_levels ||
-		    (list_empty(&tmigr_level_list[i]) &&
-		     list_is_singular(&tmigr_level_list[i - 1])))
+		if (group->parent)
 			break;
+		if ((!root_mismatch || i >= tmigr_crossnode_level) &&
+		    list_is_singular(&tmigr_level_list[i]))
+			break;
+	}
 
-	} while (i < tmigr_hierarchy_levels);
+	/* Assert single root without parent */
+	if (WARN_ON_ONCE(i >= tmigr_hierarchy_levels))
+		return -EINVAL;
 
-	while (i > 0) {
-		group = stack[--i];
+	for (; i >= start_lvl; i--) {
+		group = stack[i];
 
 		if (err < 0) {
 			list_del(&group->list);
@@ -1611,14 +1815,12 @@ static int tmigr_setup_groups(unsigned int cpu, unsigned int node)
 		 * Update tmc -> group / child -> group connection
 		 */
 		if (i == 0) {
-			struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-
-			raw_spin_lock_irq(&group->lock);
+			struct tmigr_cpu *tmc = per_cpu_ptr(&tmigr_cpu, cpu);
 
 			tmc->tmgroup = group;
-			tmc->childmask = BIT(group->num_children++);
+			tmc->groupmask = BIT(group->num_children++);
 
-			raw_spin_unlock_irq(&group->lock);
+			tmigr_init_root(group, activate);
 
 			trace_tmigr_connect_cpu_parent(tmc);
 
@@ -1626,27 +1828,58 @@ static int tmigr_setup_groups(unsigned int cpu, unsigned int node)
 			continue;
 		} else {
 			child = stack[i - 1];
-			tmigr_connect_child_parent(child, group);
+			tmigr_connect_child_parent(child, group, activate);
 		}
+	}
 
-		/* check if uppermost level was newly created */
-		if (top != i)
-			continue;
+	if (err < 0)
+		goto out;
 
-		WARN_ON_ONCE(top == 0);
+	if (activate) {
+		struct tmigr_walk data;
+		union tmigr_state state;
 
-		lvllist = &tmigr_level_list[top];
-		if (group->num_children == 1 && list_is_singular(lvllist)) {
-			lvllist = &tmigr_level_list[top - 1];
-			list_for_each_entry(child, lvllist, list) {
-				if (child->parent)
-					continue;
+		/*
+		 * To prevent inconsistent states, active children need to be active in
+		 * the new parent as well. Inactive children are already marked inactive
+		 * in the parent group:
+		 *
+		 * * When new groups were created by tmigr_setup_groups() starting from
+		 *   the lowest level, then they are not active. They will be set active
+		 *   when the new online CPU comes active.
+		 *
+		 * * But if new groups above the current top level are required, it is
+		 *   mandatory to propagate the active state of the already existing
+		 *   child to the new parents. So tmigr_active_up() activates the
+		 *   new parents while walking up from the old root to the new.
+		 *
+		 * * It is ensured that @start is active, as this setup path is
+		 *   executed in hotplug prepare callback. This is executed by an
+		 *   already connected and !idle CPU. Even if all other CPUs go idle,
+		 *   the CPU executing the setup will be responsible up to current top
+		 *   level group. And the next time it goes inactive, it will release
+		 *   the new childmask and parent to subsequent walkers through this
+		 *   @child. Therefore propagate active state unconditionally.
+		 */
+		state.state = atomic_read(&start->migr_state);
+		WARN_ON_ONCE(!state.active);
+		WARN_ON_ONCE(!start->parent);
+		data.childmask = start->groupmask;
+		__walk_groups_from(tmigr_active_up, &data, start, start->parent);
+	}
 
-				tmigr_connect_child_parent(child, group);
-			}
+	/* Root update */
+	if (list_is_singular(&tmigr_level_list[top])) {
+		group = list_first_entry(&tmigr_level_list[top],
+					 typeof(*group), list);
+		WARN_ON_ONCE(group->parent);
+		if (tmigr_root) {
+			/* Old root should be the same or below */
+			WARN_ON_ONCE(tmigr_root->level > top);
 		}
+		tmigr_root = group;
 	}
-
+out:
 	kfree(stack);
 
 	return err;
@@ -1654,90 +1887,60 @@ static int tmigr_setup_groups(unsigned int cpu, unsigned int node)
 
 static int tmigr_add_cpu(unsigned int cpu)
 {
+	struct tmigr_group *old_root = tmigr_root;
 	int node = cpu_to_node(cpu);
 	int ret;
 
-	mutex_lock(&tmigr_mutex);
-	ret = tmigr_setup_groups(cpu, node);
-	mutex_unlock(&tmigr_mutex);
-
-	return ret;
-}
-
-static int tmigr_cpu_online(unsigned int cpu)
-{
-	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-	int ret;
-
-	/* First online attempt? Initialize CPU data */
-	if (!tmc->tmgroup) {
-		raw_spin_lock_init(&tmc->lock);
-
-		ret = tmigr_add_cpu(cpu);
-		if (ret < 0)
-			return ret;
+	guard(mutex)(&tmigr_mutex);
 
-		if (tmc->childmask == 0)
-			return -EINVAL;
+	ret = tmigr_setup_groups(cpu, node, NULL, false);
 
-		timerqueue_init(&tmc->cpuevt.nextevt);
-		tmc->cpuevt.nextevt.expires = KTIME_MAX;
-		tmc->cpuevt.ignore = true;
-		tmc->cpuevt.cpu = cpu;
-
-		tmc->remote = false;
-		WRITE_ONCE(tmc->wakeup, KTIME_MAX);
+	/* Root has changed? Connect the old one to the new */
+	if (ret >= 0 && old_root && old_root != tmigr_root) {
+		/*
+		 * The target CPU must never do the prepare work, except
+		 * on early boot when the boot CPU is the target. Otherwise
+		 * it may spuriously activate the old top level group inside
+		 * the new one (nevertheless whether old top level group is
+		 * active or not) and/or release an uninitialized childmask.
+		 */
+		WARN_ON_ONCE(cpu == raw_smp_processor_id());
+		/*
+		 * The (likely) current CPU is expected to be online in the hierarchy,
+		 * otherwise the old root may not be active as expected.
+		 */
+		WARN_ON_ONCE(!per_cpu_ptr(&tmigr_cpu, raw_smp_processor_id())->available);
+		ret = tmigr_setup_groups(-1, old_root->numa_node, old_root, true);
 	}
-	raw_spin_lock_irq(&tmc->lock);
-	trace_tmigr_cpu_online(tmc);
-	tmc->idle = timer_base_is_idle();
-	if (!tmc->idle)
-		__tmigr_cpu_activate(tmc);
-	tmc->online = true;
-	raw_spin_unlock_irq(&tmc->lock);
-	return 0;
-}
-
-/*
- * tmigr_trigger_active() - trigger a CPU to become active again
- *
- * This function is executed on a CPU which is part of cpu_online_mask, when the
- * last active CPU in the hierarchy is offlining. With this, it is ensured that
- * the other CPU is active and takes over the migrator duty.
- */
-static long tmigr_trigger_active(void *unused)
-{
-	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-
-	WARN_ON_ONCE(!tmc->online || tmc->idle);
 
-	return 0;
+	return ret;
 }
 
-static int tmigr_cpu_offline(unsigned int cpu)
+static int tmigr_cpu_prepare(unsigned int cpu)
 {
-	struct tmigr_cpu *tmc = this_cpu_ptr(&tmigr_cpu);
-	int migrator;
-	u64 firstexp;
+	struct tmigr_cpu *tmc = per_cpu_ptr(&tmigr_cpu, cpu);
+	int ret = 0;
 
-	raw_spin_lock_irq(&tmc->lock);
-	tmc->online = false;
+	/* Not first online attempt? */
+	if (tmc->tmgroup)
+		return ret;
+
+	raw_spin_lock_init(&tmc->lock);
+	timerqueue_init(&tmc->cpuevt.nextevt);
+	tmc->cpuevt.nextevt.expires = KTIME_MAX;
+	tmc->cpuevt.ignore = true;
+	tmc->cpuevt.cpu = cpu;
+	tmc->remote = false;
 	WRITE_ONCE(tmc->wakeup, KTIME_MAX);
 
-	/*
-	 * CPU has to handle the local events on his own, when on the way to
-	 * offline; Therefore nextevt value is set to KTIME_MAX
-	 */
-	firstexp = __tmigr_cpu_deactivate(tmc, KTIME_MAX);
-	trace_tmigr_cpu_offline(tmc);
-	raw_spin_unlock_irq(&tmc->lock);
+	ret = tmigr_add_cpu(cpu);
+	if (ret < 0)
+		return ret;
 
-	if (firstexp != KTIME_MAX) {
-		migrator = cpumask_any_but(cpu_online_mask, cpu);
-		work_on_cpu(migrator, tmigr_trigger_active, NULL);
-	}
+	if (tmc->groupmask == 0)
+		return -EINVAL;
 
-	return 0;
+	return ret;
 }
 
 static int __init tmigr_init(void)
@@ -1753,6 +1956,11 @@ static int __init tmigr_init(void)
 	if (ncpus == 1)
 		return 0;
 
+	if (!zalloc_cpumask_var(&tmigr_available_cpumask, GFP_KERNEL)) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
 	/*
 	 * Calculate the required hierarchy levels. Unfortunately there is no
 	 * reliable information available, unless all possible CPUs have been
@@ -1796,8 +2004,13 @@ static int __init tmigr_init(void)
 		tmigr_hierarchy_levels, TMIGR_CHILDREN_PER_GROUP,
 		tmigr_crossnode_level);
 
+	ret = cpuhp_setup_state(CPUHP_TMIGR_PREPARE, "tmigr:prepare",
+				tmigr_cpu_prepare, NULL);
+	if (ret)
+		goto err;
+
 	ret = cpuhp_setup_state(CPUHP_AP_TMIGR_ONLINE, "tmigr:online",
-				tmigr_cpu_online, tmigr_cpu_offline);
+				tmigr_set_cpu_available, tmigr_clear_cpu_available);
 	if (ret)
 		goto err;
 
@@ -1807,4 +2020,4 @@ err:
 	pr_err("Timer migration setup failed\n");
 	return ret;
 }
-late_initcall(tmigr_init);
+early_initcall(tmigr_init);