Merge remote-tracking branch 'tip/core/rcu' into next.2012.09.25b

Resolved conflict in kernel/sched/core.c using Peter Zijlstra's
approach from https://lkml.org/lkml/2012/9/5/585.

10 files changed, 198 insertions, 141 deletions

diff --git a/kernel/events/core.c b/kernel/events/core.c
index b7935fcec7d9..7fee567153f0 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -1253,7 +1253,7 @@ retry:
 /*
  * Cross CPU call to disable a performance event
  */
-static int __perf_event_disable(void *info)
+int __perf_event_disable(void *info)
 {
 	struct perf_event *event = info;
 	struct perf_event_context *ctx = event->ctx;
@@ -2935,12 +2935,12 @@ EXPORT_SYMBOL_GPL(perf_event_release_kernel);
 /*
  * Called when the last reference to the file is gone.
  */
-static int perf_release(struct inode *inode, struct file *file)
+static void put_event(struct perf_event *event)
 {
-	struct perf_event *event = file->private_data;
 	struct task_struct *owner;
 
-	file->private_data = NULL;
+	if (!atomic_long_dec_and_test(&event->refcount))
+		return;
 
 	rcu_read_lock();
 	owner = ACCESS_ONCE(event->owner);
@@ -2975,7 +2975,13 @@ static int perf_release(struct inode *inode, struct file *file)
 		put_task_struct(owner);
 	}
 
-	return perf_event_release_kernel(event);
+	perf_event_release_kernel(event);
+}
+
+static int perf_release(struct inode *inode, struct file *file)
+{
+	put_event(file->private_data);
+	return 0;
 }
 
 u64 perf_event_read_value(struct perf_event *event, u64 *enabled, u64 *running)
@@ -3227,7 +3233,7 @@ unlock:
 
 static const struct file_operations perf_fops;
 
-static struct perf_event *perf_fget_light(int fd, int *fput_needed)
+static struct file *perf_fget_light(int fd, int *fput_needed)
 {
 	struct file *file;
 
@@ -3241,7 +3247,7 @@ static struct perf_event *perf_fget_light(int fd, int *fput_needed)
 		return ERR_PTR(-EBADF);
 	}
 
-	return file->private_data;
+	return file;
 }
 
 static int perf_event_set_output(struct perf_event *event,
@@ -3273,19 +3279,21 @@ static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 
 	case PERF_EVENT_IOC_SET_OUTPUT:
 	{
+		struct file *output_file = NULL;
 		struct perf_event *output_event = NULL;
 		int fput_needed = 0;
 		int ret;
 
 		if (arg != -1) {
-			output_event = perf_fget_light(arg, &fput_needed);
-			if (IS_ERR(output_event))
-				return PTR_ERR(output_event);
+			output_file = perf_fget_light(arg, &fput_needed);
+			if (IS_ERR(output_file))
+				return PTR_ERR(output_file);
+			output_event = output_file->private_data;
 		}
 
 		ret = perf_event_set_output(event, output_event);
 		if (output_event)
-			fput_light(output_event->filp, fput_needed);
+			fput_light(output_file, fput_needed);
 
 		return ret;
 	}
@@ -5950,6 +5958,7 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu,
 
 	mutex_init(&event->mmap_mutex);
 
+	atomic_long_set(&event->refcount, 1);
 	event->cpu		= cpu;
 	event->attr		= *attr;
 	event->group_leader	= group_leader;
@@ -6260,12 +6269,12 @@ SYSCALL_DEFINE5(perf_event_open,
 		return event_fd;
 
 	if (group_fd != -1) {
-		group_leader = perf_fget_light(group_fd, &fput_needed);
-		if (IS_ERR(group_leader)) {
-			err = PTR_ERR(group_leader);
+		group_file = perf_fget_light(group_fd, &fput_needed);
+		if (IS_ERR(group_file)) {
+			err = PTR_ERR(group_file);
 			goto err_fd;
 		}
-		group_file = group_leader->filp;
+		group_leader = group_file->private_data;
 		if (flags & PERF_FLAG_FD_OUTPUT)
 			output_event = group_leader;
 		if (flags & PERF_FLAG_FD_NO_GROUP)
@@ -6402,7 +6411,6 @@ SYSCALL_DEFINE5(perf_event_open,
 		put_ctx(gctx);
 	}
 
-	event->filp = event_file;
 	WARN_ON_ONCE(ctx->parent_ctx);
 	mutex_lock(&ctx->mutex);
 
@@ -6496,7 +6504,6 @@ perf_event_create_kernel_counter(struct perf_event_attr *attr, int cpu,
 		goto err_free;
 	}
 
-	event->filp = NULL;
 	WARN_ON_ONCE(ctx->parent_ctx);
 	mutex_lock(&ctx->mutex);
 	perf_install_in_context(ctx, event, cpu);
@@ -6578,7 +6585,7 @@ static void sync_child_event(struct perf_event *child_event,
 	 * Release the parent event, if this was the last
 	 * reference to it.
 	 */
-	fput(parent_event->filp);
+	put_event(parent_event);
 }
 
 static void
@@ -6654,9 +6661,8 @@ static void perf_event_exit_task_context(struct task_struct *child, int ctxn)
 	 *
 	 *   __perf_event_exit_task()
 	 *     sync_child_event()
-	 *       fput(parent_event->filp)
-	 *         perf_release()
-	 *           mutex_lock(&ctx->mutex)
+	 *       put_event()
+	 *         mutex_lock(&ctx->mutex)
 	 *
 	 * But since its the parent context it won't be the same instance.
 	 */
@@ -6724,7 +6730,7 @@ static void perf_free_event(struct perf_event *event,
 	list_del_init(&event->child_list);
 	mutex_unlock(&parent->child_mutex);
 
-	fput(parent->filp);
+	put_event(parent);
 
 	perf_group_detach(event);
 	list_del_event(event, ctx);
@@ -6804,6 +6810,12 @@ inherit_event(struct perf_event *parent_event,
 				           NULL, NULL);
 	if (IS_ERR(child_event))
 		return child_event;
+
+	if (!atomic_long_inc_not_zero(&parent_event->refcount)) {
+		free_event(child_event);
+		return NULL;
+	}
+
 	get_ctx(child_ctx);
 
 	/*
@@ -6845,14 +6857,6 @@ inherit_event(struct perf_event *parent_event,
 	raw_spin_unlock_irqrestore(&child_ctx->lock, flags);
 
 	/*
-	 * Get a reference to the parent filp - we will fput it
-	 * when the child event exits. This is safe to do because
-	 * we are in the parent and we know that the filp still
-	 * exists and has a nonzero count:
-	 */
-	atomic_long_inc(&parent_event->filp->f_count);
-
-	/*
 	 * Link this into the parent event's child list
 	 */
 	WARN_ON_ONCE(parent_event->ctx->parent_ctx);
diff --git a/kernel/events/hw_breakpoint.c b/kernel/events/hw_breakpoint.c
index bb38c4d3ee12..9a7b487c6fe2 100644
--- a/kernel/events/hw_breakpoint.c
+++ b/kernel/events/hw_breakpoint.c
@@ -453,7 +453,16 @@ int modify_user_hw_breakpoint(struct perf_event *bp, struct perf_event_attr *att
 	int old_type = bp->attr.bp_type;
 	int err = 0;
 
-	perf_event_disable(bp);
+	/*
+	 * modify_user_hw_breakpoint can be invoked with IRQs disabled and hence it
+	 * will not be possible to raise IPIs that invoke __perf_event_disable.
+	 * So call the function directly after making sure we are targeting the
+	 * current task.
+	 */
+	if (irqs_disabled() && bp->ctx && bp->ctx->task == current)
+		__perf_event_disable(bp);
+	else
+		perf_event_disable(bp);
 
 	bp->attr.bp_addr = attr->bp_addr;
 	bp->attr.bp_type = attr->bp_type;
diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c
index b3c7fd554250..6144bab8fd8e 100644
--- a/kernel/pid_namespace.c
+++ b/kernel/pid_namespace.c
@@ -232,15 +232,19 @@ static int pid_ns_ctl_handler(struct ctl_table *table, int write,
 	 */
 
 	tmp.data = &current->nsproxy->pid_ns->last_pid;
-	return proc_dointvec(&tmp, write, buffer, lenp, ppos);
+	return proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos);
 }
 
+extern int pid_max;
+static int zero = 0;
 static struct ctl_table pid_ns_ctl_table[] = {
 	{
 		.procname = "ns_last_pid",
 		.maxlen = sizeof(int),
 		.mode = 0666, /* permissions are checked in the handler */
 		.proc_handler = pid_ns_ctl_handler,
+		.extra1 = &zero,
+		.extra2 = &pid_max,
 	},
 	{ }
 };
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 8c38b5e7ce47..1a48cdbc8631 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -5342,9 +5342,6 @@ static void migrate_tasks(unsigned int dead_cpu)
 	 */
 	rq->stop = NULL;
 
-	/* Ensure any throttled groups are reachable by pick_next_task */
-	unthrottle_offline_cfs_rqs(rq);
-
 	for ( ; ; ) {
 		/*
 		 * There's this thread running, bail when that's the only
@@ -5610,15 +5607,7 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu)
 		break;
 
 	case CPU_DEAD:
-		{
-			struct rq *dest_rq;
-
-			local_irq_save(flags);
-			dest_rq = cpu_rq(smp_processor_id());
-			raw_spin_lock(&dest_rq->lock);
-			calc_load_migrate(rq);
-			raw_spin_unlock_irqrestore(&dest_rq->lock, flags);
-		}
+		calc_load_migrate(rq);
 		break;
 #endif
 	}
@@ -6027,11 +6016,6 @@ static void destroy_sched_domains(struct sched_domain *sd, int cpu)
  * SD_SHARE_PKG_RESOURCE set (Last Level Cache Domain) for this
  * allows us to avoid some pointer chasing select_idle_sibling().
  *
- * Iterate domains and sched_groups downward, assigning CPUs to be
- * select_idle_sibling() hw buddy.  Cross-wiring hw makes bouncing
- * due to random perturbation self canceling, ie sw buddies pull
- * their counterpart to their CPU's hw counterpart.
- *
  * Also keep a unique ID per domain (we use the first cpu number in
  * the cpumask of the domain), this allows us to quickly tell if
  * two cpus are in the same cache domain, see cpus_share_cache().
@@ -6045,40 +6029,8 @@ static void update_top_cache_domain(int cpu)
 	int id = cpu;
 
 	sd = highest_flag_domain(cpu, SD_SHARE_PKG_RESOURCES);
-	if (sd) {
-		struct sched_domain *tmp = sd;
-		struct sched_group *sg, *prev;
-		bool right;
-
-		/*
-		 * Traverse to first CPU in group, and count hops
-		 * to cpu from there, switching direction on each
-		 * hop, never ever pointing the last CPU rightward.
-		 */
-		do {
-			id = cpumask_first(sched_domain_span(tmp));
-			prev = sg = tmp->groups;
-			right = 1;
-
-			while (cpumask_first(sched_group_cpus(sg)) != id)
-				sg = sg->next;
-
-			while (!cpumask_test_cpu(cpu, sched_group_cpus(sg))) {
-				prev = sg;
-				sg = sg->next;
-				right = !right;
-			}
-
-			/* A CPU went down, never point back to domain start. */
-			if (right && cpumask_first(sched_group_cpus(sg->next)) == id)
-				right = false;
-
-			sg = right ? sg->next : prev;
-			tmp->idle_buddy = cpumask_first(sched_group_cpus(sg));
-		} while ((tmp = tmp->child));
-
+	if (sd)
 		id = cpumask_first(sched_domain_span(sd));
-	}
 
 	rcu_assign_pointer(per_cpu(sd_llc, cpu), sd);
 	per_cpu(sd_llc_id, cpu) = id;
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index c219bf8d704c..96e2b18b6283 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -2052,7 +2052,7 @@ static void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
 	hrtimer_cancel(&cfs_b->slack_timer);
 }
 
-void unthrottle_offline_cfs_rqs(struct rq *rq)
+static void unthrottle_offline_cfs_rqs(struct rq *rq)
 {
 	struct cfs_rq *cfs_rq;
 
@@ -2106,7 +2106,7 @@ static inline struct cfs_bandwidth *tg_cfs_bandwidth(struct task_group *tg)
 	return NULL;
 }
 static inline void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b) {}
-void unthrottle_offline_cfs_rqs(struct rq *rq) {}
+static inline void unthrottle_offline_cfs_rqs(struct rq *rq) {}
 
 #endif /* CONFIG_CFS_BANDWIDTH */
 
@@ -2637,6 +2637,8 @@ static int select_idle_sibling(struct task_struct *p, int target)
 	int cpu = smp_processor_id();
 	int prev_cpu = task_cpu(p);
 	struct sched_domain *sd;
+	struct sched_group *sg;
+	int i;
 
 	/*
 	 * If the task is going to be woken-up on this cpu and if it is
@@ -2653,17 +2655,29 @@ static int select_idle_sibling(struct task_struct *p, int target)
 		return prev_cpu;
 
 	/*
-	 * Otherwise, check assigned siblings to find an elegible idle cpu.
+	 * Otherwise, iterate the domains and find an elegible idle cpu.
 	 */
 	sd = rcu_dereference(per_cpu(sd_llc, target));
-
 	for_each_lower_domain(sd) {
-		if (!cpumask_test_cpu(sd->idle_buddy, tsk_cpus_allowed(p)))
-			continue;
-		if (idle_cpu(sd->idle_buddy))
-			return sd->idle_buddy;
-	}
+		sg = sd->groups;
+		do {
+			if (!cpumask_intersects(sched_group_cpus(sg),
+						tsk_cpus_allowed(p)))
+				goto next;
 
+			for_each_cpu(i, sched_group_cpus(sg)) {
+				if (!idle_cpu(i))
+					goto next;
+			}
+
+			target = cpumask_first_and(sched_group_cpus(sg),
+					tsk_cpus_allowed(p));
+			goto done;
+next:
+			sg = sg->next;
+		} while (sg != sd->groups);
+	}
+done:
 	return target;
 }
 
@@ -3658,7 +3672,6 @@ fix_small_capacity(struct sched_domain *sd, struct sched_group *group)
  * @group: sched_group whose statistics are to be updated.
  * @load_idx: Load index of sched_domain of this_cpu for load calc.
  * @local_group: Does group contain this_cpu.
- * @cpus: Set of cpus considered for load balancing.
  * @balance: Should we balance.
  * @sgs: variable to hold the statistics for this group.
  */
@@ -3805,7 +3818,6 @@ static bool update_sd_pick_busiest(struct lb_env *env,
 /**
  * update_sd_lb_stats - Update sched_domain's statistics for load balancing.
  * @env: The load balancing environment.
- * @cpus: Set of cpus considered for load balancing.
  * @balance: Should we balance.
  * @sds: variable to hold the statistics for this sched_domain.
  */
@@ -4956,6 +4968,9 @@ static void rq_online_fair(struct rq *rq)
 static void rq_offline_fair(struct rq *rq)
 {
 	update_sysctl();
+
+	/* Ensure any throttled groups are reachable by pick_next_task */
+	unthrottle_offline_cfs_rqs(rq);
 }
 
 #endif /* CONFIG_SMP */
diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index 944cb68420e9..e0b7ba9c040f 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
@@ -691,6 +691,7 @@ balanced:
 		 * runtime - in which case borrowing doesn't make sense.
 		 */
 		rt_rq->rt_runtime = RUNTIME_INF;
+		rt_rq->rt_throttled = 0;
 		raw_spin_unlock(&rt_rq->rt_runtime_lock);
 		raw_spin_unlock(&rt_b->rt_runtime_lock);
 	}
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index f6714d009e77..0848fa36c383 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1144,7 +1144,6 @@ extern void print_rt_stats(struct seq_file *m, int cpu);
 
 extern void init_cfs_rq(struct cfs_rq *cfs_rq);
 extern void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq);
-extern void unthrottle_offline_cfs_rqs(struct rq *rq);
 
 extern void account_cfs_bandwidth_used(int enabled, int was_enabled);
 
diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
index 4b1785a7bb83..cf5f6b262673 100644
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@ -574,6 +574,7 @@ static void tick_nohz_restart_sched_tick(struct tick_sched *ts, ktime_t now)
 	tick_do_update_jiffies64(now);
 	update_cpu_load_nohz();
 
+	calc_load_exit_idle();
 	touch_softlockup_watchdog();
 	/*
 	 * Cancel the scheduled timer and restore the tick
diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
index 34e5eac81424..d3b91e75cecd 100644
--- a/kernel/time/timekeeping.c
+++ b/kernel/time/timekeeping.c
@@ -303,10 +303,11 @@ void getnstimeofday(struct timespec *ts)
 		seq = read_seqbegin(&tk->lock);
 
 		ts->tv_sec = tk->xtime_sec;
-		ts->tv_nsec = timekeeping_get_ns(tk);
+		nsecs = timekeeping_get_ns(tk);
 
 	} while (read_seqretry(&tk->lock, seq));
 
+	ts->tv_nsec = 0;
 	timespec_add_ns(ts, nsecs);
 }
 EXPORT_SYMBOL(getnstimeofday);
@@ -345,6 +346,7 @@ void ktime_get_ts(struct timespec *ts)
 {
 	struct timekeeper *tk = &timekeeper;
 	struct timespec tomono;
+	s64 nsec;
 	unsigned int seq;
 
 	WARN_ON(timekeeping_suspended);
@@ -352,13 +354,14 @@ void ktime_get_ts(struct timespec *ts)
 	do {
 		seq = read_seqbegin(&tk->lock);
 		ts->tv_sec = tk->xtime_sec;
-		ts->tv_nsec = timekeeping_get_ns(tk);
+		nsec = timekeeping_get_ns(tk);
 		tomono = tk->wall_to_monotonic;
 
 	} while (read_seqretry(&tk->lock, seq));
 
-	set_normalized_timespec(ts, ts->tv_sec + tomono.tv_sec,
-				ts->tv_nsec + tomono.tv_nsec);
+	ts->tv_sec += tomono.tv_sec;
+	ts->tv_nsec = 0;
+	timespec_add_ns(ts, nsec + tomono.tv_nsec);
 }
 EXPORT_SYMBOL_GPL(ktime_get_ts);
 
@@ -1244,6 +1247,7 @@ void get_monotonic_boottime(struct timespec *ts)
 {
 	struct timekeeper *tk = &timekeeper;
 	struct timespec tomono, sleep;
+	s64 nsec;
 	unsigned int seq;
 
 	WARN_ON(timekeeping_suspended);
@@ -1251,14 +1255,15 @@ void get_monotonic_boottime(struct timespec *ts)
 	do {
 		seq = read_seqbegin(&tk->lock);
 		ts->tv_sec = tk->xtime_sec;
-		ts->tv_nsec = timekeeping_get_ns(tk);
+		nsec = timekeeping_get_ns(tk);
 		tomono = tk->wall_to_monotonic;
 		sleep = tk->total_sleep_time;
 
 	} while (read_seqretry(&tk->lock, seq));
 
-	set_normalized_timespec(ts, ts->tv_sec + tomono.tv_sec + sleep.tv_sec,
-			ts->tv_nsec + tomono.tv_nsec + sleep.tv_nsec);
+	ts->tv_sec += tomono.tv_sec + sleep.tv_sec;
+	ts->tv_nsec = 0;
+	timespec_add_ns(ts, nsec + tomono.tv_nsec + sleep.tv_nsec);
 }
 EXPORT_SYMBOL_GPL(get_monotonic_boottime);
 
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 692d97628a10..3c5a79e2134c 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -66,6 +66,7 @@ enum {
 
 	/* pool flags */
 	POOL_MANAGE_WORKERS	= 1 << 0,	/* need to manage workers */
+	POOL_MANAGING_WORKERS   = 1 << 1,       /* managing workers */
 
 	/* worker flags */
 	WORKER_STARTED		= 1 << 0,	/* started */
@@ -652,7 +653,7 @@ static bool need_to_manage_workers(struct worker_pool *pool)
 /* Do we have too many workers and should some go away? */
 static bool too_many_workers(struct worker_pool *pool)
 {
-	bool managing = mutex_is_locked(&pool->manager_mutex);
+	bool managing = pool->flags & POOL_MANAGING_WORKERS;
 	int nr_idle = pool->nr_idle + managing; /* manager is considered idle */
 	int nr_busy = pool->nr_workers - nr_idle;
 
@@ -1326,6 +1327,15 @@ static void idle_worker_rebind(struct worker *worker)
 
 	/* we did our part, wait for rebind_workers() to finish up */
 	wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND));
+
+	/*
+	 * rebind_workers() shouldn't finish until all workers passed the
+	 * above WORKER_REBIND wait.  Tell it when done.
+	 */
+	spin_lock_irq(&worker->pool->gcwq->lock);
+	if (!--worker->idle_rebind->cnt)
+		complete(&worker->idle_rebind->done);
+	spin_unlock_irq(&worker->pool->gcwq->lock);
 }
 
 /*
@@ -1339,8 +1349,16 @@ static void busy_worker_rebind_fn(struct work_struct *work)
 	struct worker *worker = container_of(work, struct worker, rebind_work);
 	struct global_cwq *gcwq = worker->pool->gcwq;
 
-	if (worker_maybe_bind_and_lock(worker))
-		worker_clr_flags(worker, WORKER_REBIND);
+	worker_maybe_bind_and_lock(worker);
+
+	/*
+	 * %WORKER_REBIND must be cleared even if the above binding failed;
+	 * otherwise, we may confuse the next CPU_UP cycle or oops / get
+	 * stuck by calling idle_worker_rebind() prematurely.  If CPU went
+	 * down again inbetween, %WORKER_UNBOUND would be set, so clearing
+	 * %WORKER_REBIND is always safe.
+	 */
+	worker_clr_flags(worker, WORKER_REBIND);
 
 	spin_unlock_irq(&gcwq->lock);
 }
@@ -1396,12 +1414,15 @@ retry:
 	/* set REBIND and kick idle ones, we'll wait for these later */
 	for_each_worker_pool(pool, gcwq) {
 		list_for_each_entry(worker, &pool->idle_list, entry) {
+			unsigned long worker_flags = worker->flags;
+
 			if (worker->flags & WORKER_REBIND)
 				continue;
 
-			/* morph UNBOUND to REBIND */
-			worker->flags &= ~WORKER_UNBOUND;
-			worker->flags |= WORKER_REBIND;
+			/* morph UNBOUND to REBIND atomically */
+			worker_flags &= ~WORKER_UNBOUND;
+			worker_flags |= WORKER_REBIND;
+			ACCESS_ONCE(worker->flags) = worker_flags;
 
 			idle_rebind.cnt++;
 			worker->idle_rebind = &idle_rebind;
@@ -1419,25 +1440,15 @@ retry:
 		goto retry;
 	}
 
-	/*
-	 * All idle workers are rebound and waiting for %WORKER_REBIND to
-	 * be cleared inside idle_worker_rebind().  Clear and release.
-	 * Clearing %WORKER_REBIND from this foreign context is safe
-	 * because these workers are still guaranteed to be idle.
-	 */
-	for_each_worker_pool(pool, gcwq)
-		list_for_each_entry(worker, &pool->idle_list, entry)
-			worker->flags &= ~WORKER_REBIND;
-
-	wake_up_all(&gcwq->rebind_hold);
-
-	/* rebind busy workers */
+	/* all idle workers are rebound, rebind busy workers */
 	for_each_busy_worker(worker, i, pos, gcwq) {
 		struct work_struct *rebind_work = &worker->rebind_work;
+		unsigned long worker_flags = worker->flags;
 
-		/* morph UNBOUND to REBIND */
-		worker->flags &= ~WORKER_UNBOUND;
-		worker->flags |= WORKER_REBIND;
+		/* morph UNBOUND to REBIND atomically */
+		worker_flags &= ~WORKER_UNBOUND;
+		worker_flags |= WORKER_REBIND;
+		ACCESS_ONCE(worker->flags) = worker_flags;
 
 		if (test_and_set_bit(WORK_STRUCT_PENDING_BIT,
 				     work_data_bits(rebind_work)))
@@ -1449,6 +1460,34 @@ retry:
 			    worker->scheduled.next,
 			    work_color_to_flags(WORK_NO_COLOR));
 	}
+
+	/*
+	 * All idle workers are rebound and waiting for %WORKER_REBIND to
+	 * be cleared inside idle_worker_rebind().  Clear and release.
+	 * Clearing %WORKER_REBIND from this foreign context is safe
+	 * because these workers are still guaranteed to be idle.
+	 *
+	 * We need to make sure all idle workers passed WORKER_REBIND wait
+	 * in idle_worker_rebind() before returning; otherwise, workers can
+	 * get stuck at the wait if hotplug cycle repeats.
+	 */
+	idle_rebind.cnt = 1;
+	INIT_COMPLETION(idle_rebind.done);
+
+	for_each_worker_pool(pool, gcwq) {
+		list_for_each_entry(worker, &pool->idle_list, entry) {
+			worker->flags &= ~WORKER_REBIND;
+			idle_rebind.cnt++;
+		}
+	}
+
+	wake_up_all(&gcwq->rebind_hold);
+
+	if (--idle_rebind.cnt) {
+		spin_unlock_irq(&gcwq->lock);
+		wait_for_completion(&idle_rebind.done);
+		spin_lock_irq(&gcwq->lock);
+	}
 }
 
 static struct worker *alloc_worker(void)
@@ -1794,9 +1833,45 @@ static bool manage_workers(struct worker *worker)
 	struct worker_pool *pool = worker->pool;
 	bool ret = false;
 
-	if (!mutex_trylock(&pool->manager_mutex))
+	if (pool->flags & POOL_MANAGING_WORKERS)
 		return ret;
 
+	pool->flags |= POOL_MANAGING_WORKERS;
+
+	/*
+	 * To simplify both worker management and CPU hotplug, hold off
+	 * management while hotplug is in progress.  CPU hotplug path can't
+	 * grab %POOL_MANAGING_WORKERS to achieve this because that can
+	 * lead to idle worker depletion (all become busy thinking someone
+	 * else is managing) which in turn can result in deadlock under
+	 * extreme circumstances.  Use @pool->manager_mutex to synchronize
+	 * manager against CPU hotplug.
+	 *
+	 * manager_mutex would always be free unless CPU hotplug is in
+	 * progress.  trylock first without dropping @gcwq->lock.
+	 */
+	if (unlikely(!mutex_trylock(&pool->manager_mutex))) {
+		spin_unlock_irq(&pool->gcwq->lock);
+		mutex_lock(&pool->manager_mutex);
+		/*
+		 * CPU hotplug could have happened while we were waiting
+		 * for manager_mutex.  Hotplug itself can't handle us
+		 * because manager isn't either on idle or busy list, and
+		 * @gcwq's state and ours could have deviated.
+		 *
+		 * As hotplug is now excluded via manager_mutex, we can
+		 * simply try to bind.  It will succeed or fail depending
+		 * on @gcwq's current state.  Try it and adjust
+		 * %WORKER_UNBOUND accordingly.
+		 */
+		if (worker_maybe_bind_and_lock(worker))
+			worker->flags &= ~WORKER_UNBOUND;
+		else
+			worker->flags |= WORKER_UNBOUND;
+
+		ret = true;
+	}
+
 	pool->flags &= ~POOL_MANAGE_WORKERS;
 
 	/*
@@ -1806,6 +1881,7 @@ static bool manage_workers(struct worker *worker)
 	ret |= maybe_destroy_workers(pool);
 	ret |= maybe_create_worker(pool);
 
+	pool->flags &= ~POOL_MANAGING_WORKERS;
 	mutex_unlock(&pool->manager_mutex);
 	return ret;
 }
@@ -3500,18 +3576,17 @@ static int __devinit workqueue_cpu_down_callback(struct notifier_block *nfb,
 #ifdef CONFIG_SMP
 
 struct work_for_cpu {
-	struct completion completion;
+	struct work_struct work;
 	long (*fn)(void *);
 	void *arg;
 	long ret;
 };
 
-static int do_work_for_cpu(void *_wfc)
+static void work_for_cpu_fn(struct work_struct *work)
 {
-	struct work_for_cpu *wfc = _wfc;
+	struct work_for_cpu *wfc = container_of(work, struct work_for_cpu, work);
+
 	wfc->ret = wfc->fn(wfc->arg);
-	complete(&wfc->completion);
-	return 0;
 }
 
 /**
@@ -3526,19 +3601,11 @@ static int do_work_for_cpu(void *_wfc)
  */
 long work_on_cpu(unsigned int cpu, long (*fn)(void *), void *arg)
 {
-	struct task_struct *sub_thread;
-	struct work_for_cpu wfc = {
-		.completion = COMPLETION_INITIALIZER_ONSTACK(wfc.completion),
-		.fn = fn,
-		.arg = arg,
-	};
+	struct work_for_cpu wfc = { .fn = fn, .arg = arg };
 
-	sub_thread = kthread_create(do_work_for_cpu, &wfc, "work_for_cpu");
-	if (IS_ERR(sub_thread))
-		return PTR_ERR(sub_thread);
-	kthread_bind(sub_thread, cpu);
-	wake_up_process(sub_thread);
-	wait_for_completion(&wfc.completion);
+	INIT_WORK_ONSTACK(&wfc.work, work_for_cpu_fn);
+	schedule_work_on(cpu, &wfc.work);
+	flush_work(&wfc.work);
 	return wfc.ret;
 }
 EXPORT_SYMBOL_GPL(work_on_cpu);