[PATCH tip/core/rcu 0/10] Dyntick changes for v4.20/v5.0

[PATCH tip/core/rcu 0/10] Dyntick changes for v4.20/v5.0

[Paul E. McKenney]

Hello!

Now that there is only one rcu_data structure per CPU, there isn't a whole
lot of point in maintaining a separate per-CPU rcu_dynticks structure.
This series therefore merges the rcu_dynticks structure into the rcu_data
structure.

1.	Remove unused rcu_dynticks_snap() from Tiny RCU.

2.	Merge rcu_dynticks structure into rcu_data structure, but
	continue using the rcu_dynticks structure's fields.

3-8.	Change the code to use the fields in the rcu_data structure
	instead of those in the rcu_dynticks structure.

9.	Remove the now-obsolete ->dynticks_fqs and ->cond_resched_completed
	fields from the rcu_data structure.

10.	Convert rcu_state.ofl_lock to raw_spinlock_t to make -rt feel
	better, courtesy of Mike Galbraith.

							Thanx, Paul

------------------------------------------------------------------------

 include/linux/rcutiny.h  |    6 -
 kernel/rcu/tree.c        |  196 ++++++++++++++++++++++-------------------------
 kernel/rcu/tree.h        |   82 +++++++------------
 kernel/rcu/tree_exp.h    |    8 -
 kernel/rcu/tree_plugin.h |   69 +++++++---------
 5 files changed, 163 insertions(+), 198 deletions(-)

[PATCH tip/core/rcu 01/10] rcu: Remove unused rcu_dynticks_snap() from Tiny RCU

[Paul E. McKenney]

The rcu_dynticks_snap() function is defined in include/linux/rcutiny.h,
but is no longer used by Tiny RCU.  This commit therefore removes it.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 include/linux/rcutiny.h | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h
index 7fa4fb9e899e..f183683bdf79 100644
--- a/include/linux/rcutiny.h
+++ b/include/linux/rcutiny.h
@@ -27,12 +27,6 @@
 
 #include <linux/ktime.h>
 
-struct rcu_dynticks;
-static inline int rcu_dynticks_snap(struct rcu_dynticks *rdtp)
-{
-	return 0;
-}
-
 /* Never flag non-existent other CPUs! */
 static inline bool rcu_eqs_special_set(int cpu) { return false; }
 
-- 
2.17.1

[PATCH tip/core/rcu 02/10] rcu: Merge rcu_dynticks structure into rcu_data structure

[Paul E. McKenney]

Now that there is only ever one rcu_data structure per CPU, there is no
need for a separate rcu_dynticks structure.  This commit therefore adds
the rcu_dynticks fields into the rcu_data structure in preparation for
removing the rcu_dynticks structure entirely.  Note that the ->dynticks
field will be handled specially because there is a field by that name
in both structures.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 5e561f1339d4..d35cd9677b08 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -212,6 +212,23 @@ struct rcu_data {
 	/* 3) dynticks interface. */
 	struct rcu_dynticks *dynticks;	/* Shared per-CPU dynticks state. */
 	int dynticks_snap;		/* Per-GP tracking for dynticks. */
+	long dynticks_nesting;      /* Track process nesting level. */
+	long dynticks_nmi_nesting;  /* Track irq/NMI nesting level. */
+	// atomic_t dynticks;	    /* Even value for idle, else odd. */
+	bool rcu_need_heavy_qs;     /* GP old, need heavy quiescent state. */
+	bool rcu_urgent_qs;	    /* GP old need light quiescent state. */
+#ifdef CONFIG_RCU_FAST_NO_HZ
+	bool all_lazy;		    /* Are all CPU's CBs lazy? */
+	unsigned long nonlazy_posted;
+				    /* # times non-lazy CBs posted to CPU. */
+	unsigned long nonlazy_posted_snap;
+				    /* idle-period nonlazy_posted snapshot. */
+	unsigned long last_accelerate;
+				    /* Last jiffy CBs were accelerated. */
+	unsigned long last_advance_all;
+				    /* Last jiffy CBs were all advanced. */
+	int tick_nohz_enabled_snap; /* Previously seen value from sysfs. */
+#endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 
 	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
 	unsigned long dynticks_fqs;	/* Kicked due to dynticks idle. */
-- 
2.17.1

[PATCH tip/core/rcu 03/10] rcu: Switch ->tick_nohz_enabled_snap to rcu_data structure

[Paul E. McKenney]

This commit removes ->tick_nohz_enabled_snap from the rcu_dynticks
structure and updates the code to access it from the rcu_data
structure.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.h        |  1 -
 kernel/rcu/tree_plugin.h | 10 +++++-----
 2 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index d35cd9677b08..5d447ceba769 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -53,7 +53,6 @@ struct rcu_dynticks {
 				    /* Last jiffy CBs were accelerated. */
 	unsigned long last_advance_all;
 				    /* Last jiffy CBs were all advanced. */
-	int tick_nohz_enabled_snap; /* Previously seen value from sysfs. */
 #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 };
 
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 978ce3539809..6511032371c1 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1681,7 +1681,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt)
 static void rcu_prepare_for_idle(void)
 {
 	bool needwake;
-	struct rcu_data *rdp;
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	struct rcu_node *rnp;
 	int tne;
@@ -1692,10 +1692,10 @@ static void rcu_prepare_for_idle(void)
 
 	/* Handle nohz enablement switches conservatively. */
 	tne = READ_ONCE(tick_nohz_active);
-	if (tne != rdtp->tick_nohz_enabled_snap) {
+	if (tne != rdp->tick_nohz_enabled_snap) {
 		if (rcu_cpu_has_callbacks(NULL))
 			invoke_rcu_core(); /* force nohz to see update. */
-		rdtp->tick_nohz_enabled_snap = tne;
+		rdp->tick_nohz_enabled_snap = tne;
 		return;
 	}
 	if (!tne)
@@ -1721,7 +1721,6 @@ static void rcu_prepare_for_idle(void)
 	if (rdtp->last_accelerate == jiffies)
 		return;
 	rdtp->last_accelerate = jiffies;
-	rdp = this_cpu_ptr(&rcu_data);
 	if (rcu_segcblist_pend_cbs(&rdp->cblist)) {
 		rnp = rdp->mynode;
 		raw_spin_lock_rcu_node(rnp); /* irqs already disabled. */
@@ -1765,6 +1764,7 @@ static void rcu_idle_count_callbacks_posted(void)
 
 static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
 {
+	struct rcu_data *rdp = &per_cpu(rcu_data, cpu);
 	struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
 	unsigned long nlpd = rdtp->nonlazy_posted - rdtp->nonlazy_posted_snap;
 
@@ -1772,7 +1772,7 @@ static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
 		rdtp->last_accelerate & 0xffff, jiffies & 0xffff,
 		ulong2long(nlpd),
 		rdtp->all_lazy ? 'L' : '.',
-		rdtp->tick_nohz_enabled_snap ? '.' : 'D');
+		rdp->tick_nohz_enabled_snap ? '.' : 'D');
 }
 
 #else /* #ifdef CONFIG_RCU_FAST_NO_HZ */
-- 
2.17.1

[PATCH tip/core/rcu 04/10] rcu: Switch last accelerate/advance to rcu_data structure

[Paul E. McKenney]

This commit removes ->last_accelerate and ->last_advance_all from the
rcu_dynticks structure and updates the code to access them from the
rcu_data structure.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.h        |  4 ----
 kernel/rcu/tree_plugin.h | 17 ++++++++---------
 2 files changed, 8 insertions(+), 13 deletions(-)

diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 5d447ceba769..69bd6bec05bb 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -49,10 +49,6 @@ struct rcu_dynticks {
 				    /* # times non-lazy CBs posted to CPU. */
 	unsigned long nonlazy_posted_snap;
 				    /* idle-period nonlazy_posted snapshot. */
-	unsigned long last_accelerate;
-				    /* Last jiffy CBs were accelerated. */
-	unsigned long last_advance_all;
-				    /* Last jiffy CBs were all advanced. */
 #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 };
 
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 6511032371c1..45708164ddf9 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1597,16 +1597,14 @@ module_param(rcu_idle_lazy_gp_delay, int, 0644);
 static bool __maybe_unused rcu_try_advance_all_cbs(void)
 {
 	bool cbs_ready = false;
-	struct rcu_data *rdp;
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	struct rcu_node *rnp;
 
 	/* Exit early if we advanced recently. */
-	if (jiffies == rdtp->last_advance_all)
+	if (jiffies == rdp->last_advance_all)
 		return false;
-	rdtp->last_advance_all = jiffies;
+	rdp->last_advance_all = jiffies;
 
-	rdp = this_cpu_ptr(&rcu_data);
 	rnp = rdp->mynode;
 
 	/*
@@ -1635,6 +1633,7 @@ static bool __maybe_unused rcu_try_advance_all_cbs(void)
  */
 int rcu_needs_cpu(u64 basemono, u64 *nextevt)
 {
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	unsigned long dj;
 
@@ -1655,7 +1654,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt)
 		invoke_rcu_core();
 		return 1;
 	}
-	rdtp->last_accelerate = jiffies;
+	rdp->last_accelerate = jiffies;
 
 	/* Request timer delay depending on laziness, and round. */
 	if (!rdtp->all_lazy) {
@@ -1718,9 +1717,9 @@ static void rcu_prepare_for_idle(void)
 	 * If we have not yet accelerated this jiffy, accelerate all
 	 * callbacks on this CPU.
 	 */
-	if (rdtp->last_accelerate == jiffies)
+	if (rdp->last_accelerate == jiffies)
 		return;
-	rdtp->last_accelerate = jiffies;
+	rdp->last_accelerate = jiffies;
 	if (rcu_segcblist_pend_cbs(&rdp->cblist)) {
 		rnp = rdp->mynode;
 		raw_spin_lock_rcu_node(rnp); /* irqs already disabled. */
@@ -1769,7 +1768,7 @@ static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
 	unsigned long nlpd = rdtp->nonlazy_posted - rdtp->nonlazy_posted_snap;
 
 	sprintf(cp, "last_accelerate: %04lx/%04lx, nonlazy_posted: %ld, %c%c",
-		rdtp->last_accelerate & 0xffff, jiffies & 0xffff,
+		rdp->last_accelerate & 0xffff, jiffies & 0xffff,
 		ulong2long(nlpd),
 		rdtp->all_lazy ? 'L' : '.',
 		rdp->tick_nohz_enabled_snap ? '.' : 'D');
-- 
2.17.1

[PATCH tip/core/rcu 05/10] rcu: Switch lazy counts to rcu_data structure

[Paul E. McKenney]

This commit removes ->all_lazy, ->nonlazy_posted and ->nonlazy_posted_snap
from the rcu_dynticks structure and updates the code to access them from
the rcu_data structure.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.h        |  7 -------
 kernel/rcu/tree_plugin.h | 23 ++++++++++-------------
 2 files changed, 10 insertions(+), 20 deletions(-)

diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 69bd6bec05bb..36a47c7bd882 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -43,13 +43,6 @@ struct rcu_dynticks {
 	atomic_t dynticks;	    /* Even value for idle, else odd. */
 	bool rcu_need_heavy_qs;     /* GP old, need heavy quiescent state. */
 	bool rcu_urgent_qs;	    /* GP old need light quiescent state. */
-#ifdef CONFIG_RCU_FAST_NO_HZ
-	bool all_lazy;		    /* Are all CPU's CBs lazy? */
-	unsigned long nonlazy_posted;
-				    /* # times non-lazy CBs posted to CPU. */
-	unsigned long nonlazy_posted_snap;
-				    /* idle-period nonlazy_posted snapshot. */
-#endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 };
 
 /* Communicate arguments to a workqueue handler. */
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 45708164ddf9..b5aeb2fe4cfe 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1634,16 +1634,15 @@ static bool __maybe_unused rcu_try_advance_all_cbs(void)
 int rcu_needs_cpu(u64 basemono, u64 *nextevt)
 {
 	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	unsigned long dj;
 
 	lockdep_assert_irqs_disabled();
 
 	/* Snapshot to detect later posting of non-lazy callback. */
-	rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted;
+	rdp->nonlazy_posted_snap = rdp->nonlazy_posted;
 
 	/* If no callbacks, RCU doesn't need the CPU. */
-	if (!rcu_cpu_has_callbacks(&rdtp->all_lazy)) {
+	if (!rcu_cpu_has_callbacks(&rdp->all_lazy)) {
 		*nextevt = KTIME_MAX;
 		return 0;
 	}
@@ -1657,7 +1656,7 @@ int rcu_needs_cpu(u64 basemono, u64 *nextevt)
 	rdp->last_accelerate = jiffies;
 
 	/* Request timer delay depending on laziness, and round. */
-	if (!rdtp->all_lazy) {
+	if (!rdp->all_lazy) {
 		dj = round_up(rcu_idle_gp_delay + jiffies,
 			       rcu_idle_gp_delay) - jiffies;
 	} else {
@@ -1681,7 +1680,6 @@ static void rcu_prepare_for_idle(void)
 {
 	bool needwake;
 	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	struct rcu_node *rnp;
 	int tne;
 
@@ -1705,10 +1703,10 @@ static void rcu_prepare_for_idle(void)
 	 * callbacks, invoke RCU core for the side-effect of recalculating
 	 * idle duration on re-entry to idle.
 	 */
-	if (rdtp->all_lazy &&
-	    rdtp->nonlazy_posted != rdtp->nonlazy_posted_snap) {
-		rdtp->all_lazy = false;
-		rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted;
+	if (rdp->all_lazy &&
+	    rdp->nonlazy_posted != rdp->nonlazy_posted_snap) {
+		rdp->all_lazy = false;
+		rdp->nonlazy_posted_snap = rdp->nonlazy_posted;
 		invoke_rcu_core();
 		return;
 	}
@@ -1754,7 +1752,7 @@ static void rcu_cleanup_after_idle(void)
  */
 static void rcu_idle_count_callbacks_posted(void)
 {
-	__this_cpu_add(rcu_dynticks.nonlazy_posted, 1);
+	__this_cpu_add(rcu_data.nonlazy_posted, 1);
 }
 
 #endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */
@@ -1764,13 +1762,12 @@ static void rcu_idle_count_callbacks_posted(void)
 static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
 {
 	struct rcu_data *rdp = &per_cpu(rcu_data, cpu);
-	struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
-	unsigned long nlpd = rdtp->nonlazy_posted - rdtp->nonlazy_posted_snap;
+	unsigned long nlpd = rdp->nonlazy_posted - rdp->nonlazy_posted_snap;
 
 	sprintf(cp, "last_accelerate: %04lx/%04lx, nonlazy_posted: %ld, %c%c",
 		rdp->last_accelerate & 0xffff, jiffies & 0xffff,
 		ulong2long(nlpd),
-		rdtp->all_lazy ? 'L' : '.',
+		rdp->all_lazy ? 'L' : '.',
 		rdp->tick_nohz_enabled_snap ? '.' : 'D');
 }
 
-- 
2.17.1

[PATCH tip/core/rcu 06/10] rcu: Switch urgent quiescent-state requests to rcu_data structure

[Paul E. McKenney]

This commit removes ->rcu_need_heavy_qs and ->rcu_urgent_qs from the
rcu_dynticks structure and updates the code to access them from the
rcu_data structure.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.c        | 12 ++++++------
 kernel/rcu/tree.h        |  2 --
 kernel/rcu/tree_exp.h    |  2 +-
 kernel/rcu/tree_plugin.h | 14 +++++++-------
 4 files changed, 14 insertions(+), 16 deletions(-)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index 70f760f28432..44f5e0a99e5a 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -362,7 +362,7 @@ static void __maybe_unused rcu_momentary_dyntick_idle(void)
 	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	int special;
 
-	raw_cpu_write(rcu_dynticks.rcu_need_heavy_qs, false);
+	raw_cpu_write(rcu_data.rcu_need_heavy_qs, false);
 	special = atomic_add_return(2 * RCU_DYNTICK_CTRL_CTR, &rdtp->dynticks);
 	/* It is illegal to call this from idle state. */
 	WARN_ON_ONCE(!(special & RCU_DYNTICK_CTRL_CTR));
@@ -927,7 +927,7 @@ void rcu_request_urgent_qs_task(struct task_struct *t)
 	cpu = task_cpu(t);
 	if (!task_curr(t))
 		return; /* This task is not running on that CPU. */
-	smp_store_release(per_cpu_ptr(&rcu_dynticks.rcu_urgent_qs, cpu), true);
+	smp_store_release(per_cpu_ptr(&rcu_data.rcu_urgent_qs, cpu), true);
 }
 
 #if defined(CONFIG_PROVE_RCU) && defined(CONFIG_HOTPLUG_CPU)
@@ -1080,8 +1080,8 @@ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp)
 	 * is set way high.
 	 */
 	jtsq = READ_ONCE(jiffies_to_sched_qs);
-	ruqp = per_cpu_ptr(&rcu_dynticks.rcu_urgent_qs, rdp->cpu);
-	rnhqp = &per_cpu(rcu_dynticks.rcu_need_heavy_qs, rdp->cpu);
+	ruqp = per_cpu_ptr(&rcu_data.rcu_urgent_qs, rdp->cpu);
+	rnhqp = &per_cpu(rcu_data.rcu_need_heavy_qs, rdp->cpu);
 	if (!READ_ONCE(*rnhqp) &&
 	    (time_after(jiffies, rcu_state.gp_start + jtsq * 2) ||
 	     time_after(jiffies, rcu_state.jiffies_resched))) {
@@ -2498,13 +2498,13 @@ void rcu_check_callbacks(int user)
 	trace_rcu_utilization(TPS("Start scheduler-tick"));
 	raw_cpu_inc(rcu_data.ticks_this_gp);
 	/* The load-acquire pairs with the store-release setting to true. */
-	if (smp_load_acquire(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs))) {
+	if (smp_load_acquire(this_cpu_ptr(&rcu_data.rcu_urgent_qs))) {
 		/* Idle and userspace execution already are quiescent states. */
 		if (!rcu_is_cpu_rrupt_from_idle() && !user) {
 			set_tsk_need_resched(current);
 			set_preempt_need_resched();
 		}
-		__this_cpu_write(rcu_dynticks.rcu_urgent_qs, false);
+		__this_cpu_write(rcu_data.rcu_urgent_qs, false);
 	}
 	rcu_flavor_check_callbacks(user);
 	if (rcu_pending())
diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 36a47c7bd882..4c31066ddb94 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -41,8 +41,6 @@ struct rcu_dynticks {
 	long dynticks_nesting;      /* Track process nesting level. */
 	long dynticks_nmi_nesting;  /* Track irq/NMI nesting level. */
 	atomic_t dynticks;	    /* Even value for idle, else odd. */
-	bool rcu_need_heavy_qs;     /* GP old, need heavy quiescent state. */
-	bool rcu_urgent_qs;	    /* GP old need light quiescent state. */
 };
 
 /* Communicate arguments to a workqueue handler. */
diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h
index 030df96e0d3c..11387fcd4d85 100644
--- a/kernel/rcu/tree_exp.h
+++ b/kernel/rcu/tree_exp.h
@@ -780,7 +780,7 @@ static void sync_sched_exp_handler(void *unused)
 	}
 	__this_cpu_write(rcu_data.cpu_no_qs.b.exp, true);
 	/* Store .exp before .rcu_urgent_qs. */
-	smp_store_release(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs), true);
+	smp_store_release(this_cpu_ptr(&rcu_data.rcu_urgent_qs), true);
 	set_tsk_need_resched(current);
 	set_preempt_need_resched();
 }
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index b5aeb2fe4cfe..161760957a07 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -967,17 +967,17 @@ void rcu_all_qs(void)
 {
 	unsigned long flags;
 
-	if (!raw_cpu_read(rcu_dynticks.rcu_urgent_qs))
+	if (!raw_cpu_read(rcu_data.rcu_urgent_qs))
 		return;
 	preempt_disable();
 	/* Load rcu_urgent_qs before other flags. */
-	if (!smp_load_acquire(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs))) {
+	if (!smp_load_acquire(this_cpu_ptr(&rcu_data.rcu_urgent_qs))) {
 		preempt_enable();
 		return;
 	}
-	this_cpu_write(rcu_dynticks.rcu_urgent_qs, false);
+	this_cpu_write(rcu_data.rcu_urgent_qs, false);
 	barrier(); /* Avoid RCU read-side critical sections leaking down. */
-	if (unlikely(raw_cpu_read(rcu_dynticks.rcu_need_heavy_qs))) {
+	if (unlikely(raw_cpu_read(rcu_data.rcu_need_heavy_qs))) {
 		local_irq_save(flags);
 		rcu_momentary_dyntick_idle();
 		local_irq_restore(flags);
@@ -997,10 +997,10 @@ void rcu_note_context_switch(bool preempt)
 	trace_rcu_utilization(TPS("Start context switch"));
 	rcu_qs();
 	/* Load rcu_urgent_qs before other flags. */
-	if (!smp_load_acquire(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs)))
+	if (!smp_load_acquire(this_cpu_ptr(&rcu_data.rcu_urgent_qs)))
 		goto out;
-	this_cpu_write(rcu_dynticks.rcu_urgent_qs, false);
-	if (unlikely(raw_cpu_read(rcu_dynticks.rcu_need_heavy_qs)))
+	this_cpu_write(rcu_data.rcu_urgent_qs, false);
+	if (unlikely(raw_cpu_read(rcu_data.rcu_need_heavy_qs)))
 		rcu_momentary_dyntick_idle();
 	if (!preempt)
 		rcu_tasks_qs(current);
-- 
2.17.1

[PATCH tip/core/rcu 07/10] rcu: Switch dyntick nesting counters to rcu_data structure

[Paul E. McKenney]

This commit removes ->dynticks_nesting and ->dynticks_nmi_nesting from
the rcu_dynticks structure and updates the code to access them from the
rcu_data structure.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.c        | 79 +++++++++++++++++++++-------------------
 kernel/rcu/tree.h        |  2 -
 kernel/rcu/tree_plugin.h |  2 +-
 3 files changed, 43 insertions(+), 40 deletions(-)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index 44f5e0a99e5a..35326d1005eb 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -73,7 +73,10 @@
 
 /* Data structures. */
 
-static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data);
+static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data) = {
+	.dynticks_nesting = 1,
+	.dynticks_nmi_nesting = DYNTICK_IRQ_NONIDLE,
+};
 struct rcu_state rcu_state = {
 	.level = { &rcu_state.node[0] },
 	.gp_state = RCU_GP_IDLE,
@@ -210,8 +213,6 @@ void rcu_softirq_qs(void)
 #endif
 
 static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = {
-	.dynticks_nesting = 1,
-	.dynticks_nmi_nesting = DYNTICK_IRQ_NONIDLE,
 	.dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR),
 };
 
@@ -378,8 +379,8 @@ static void __maybe_unused rcu_momentary_dyntick_idle(void)
  */
 static int rcu_is_cpu_rrupt_from_idle(void)
 {
-	return __this_cpu_read(rcu_dynticks.dynticks_nesting) <= 0 &&
-	       __this_cpu_read(rcu_dynticks.dynticks_nmi_nesting) <= 1;
+	return __this_cpu_read(rcu_data.dynticks_nesting) <= 0 &&
+	       __this_cpu_read(rcu_data.dynticks_nmi_nesting) <= 1;
 }
 
 #define DEFAULT_RCU_BLIMIT 10     /* Maximum callbacks per rcu_do_batch. */
@@ -571,27 +572,27 @@ static struct rcu_node *rcu_get_root(void)
  */
 static void rcu_eqs_enter(bool user)
 {
-	struct rcu_data *rdp;
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	struct rcu_dynticks *rdtp;
 
 	rdtp = this_cpu_ptr(&rcu_dynticks);
-	WARN_ON_ONCE(rdtp->dynticks_nmi_nesting != DYNTICK_IRQ_NONIDLE);
-	WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0);
+	WARN_ON_ONCE(rdp->dynticks_nmi_nesting != DYNTICK_IRQ_NONIDLE);
+	WRITE_ONCE(rdp->dynticks_nmi_nesting, 0);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) &&
-		     rdtp->dynticks_nesting == 0);
-	if (rdtp->dynticks_nesting != 1) {
-		rdtp->dynticks_nesting--;
+		     rdp->dynticks_nesting == 0);
+	if (rdp->dynticks_nesting != 1) {
+		rdp->dynticks_nesting--;
 		return;
 	}
 
 	lockdep_assert_irqs_disabled();
-	trace_rcu_dyntick(TPS("Start"), rdtp->dynticks_nesting, 0, rdtp->dynticks);
+	trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, rdtp->dynticks);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
 	rdp = this_cpu_ptr(&rcu_data);
 	do_nocb_deferred_wakeup(rdp);
 	rcu_prepare_for_idle();
 	rcu_preempt_deferred_qs(current);
-	WRITE_ONCE(rdtp->dynticks_nesting, 0); /* Avoid irq-access tearing. */
+	WRITE_ONCE(rdp->dynticks_nesting, 0); /* Avoid irq-access tearing. */
 	rcu_dynticks_eqs_enter();
 	rcu_dynticks_task_enter();
 }
@@ -637,7 +638,7 @@ void rcu_user_enter(void)
  * @irq: Is this call from rcu_irq_exit?
  *
  * If we are returning from the outermost NMI handler that interrupted an
- * RCU-idle period, update rdtp->dynticks and rdtp->dynticks_nmi_nesting
+ * RCU-idle period, update rdtp->dynticks and rdp->dynticks_nmi_nesting
  * to let the RCU grace-period handling know that the CPU is back to
  * being RCU-idle.
  *
@@ -646,30 +647,31 @@ void rcu_user_enter(void)
  */
 static __always_inline void rcu_nmi_exit_common(bool irq)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
+	struct rcu_dynticks __maybe_unused *rdtp = this_cpu_ptr(&rcu_dynticks);
 
 	/*
 	 * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks.
 	 * (We are exiting an NMI handler, so RCU better be paying attention
 	 * to us!)
 	 */
-	WARN_ON_ONCE(rdtp->dynticks_nmi_nesting <= 0);
+	WARN_ON_ONCE(rdp->dynticks_nmi_nesting <= 0);
 	WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs());
 
 	/*
 	 * If the nesting level is not 1, the CPU wasn't RCU-idle, so
 	 * leave it in non-RCU-idle state.
 	 */
-	if (rdtp->dynticks_nmi_nesting != 1) {
-		trace_rcu_dyntick(TPS("--="), rdtp->dynticks_nmi_nesting, rdtp->dynticks_nmi_nesting - 2, rdtp->dynticks);
-		WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* No store tearing. */
-			   rdtp->dynticks_nmi_nesting - 2);
+	if (rdp->dynticks_nmi_nesting != 1) {
+		trace_rcu_dyntick(TPS("--="), rdp->dynticks_nmi_nesting, rdp->dynticks_nmi_nesting - 2, rdtp->dynticks);
+		WRITE_ONCE(rdp->dynticks_nmi_nesting, /* No store tearing. */
+			   rdp->dynticks_nmi_nesting - 2);
 		return;
 	}
 
 	/* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */
-	trace_rcu_dyntick(TPS("Startirq"), rdtp->dynticks_nmi_nesting, 0, rdtp->dynticks);
-	WRITE_ONCE(rdtp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */
+	trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, rdtp->dynticks);
+	WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */
 
 	if (irq)
 		rcu_prepare_for_idle();
@@ -738,25 +740,27 @@ void rcu_irq_exit_irqson(void)
  */
 static void rcu_eqs_exit(bool user)
 {
+	struct rcu_data *rdp;
 	struct rcu_dynticks *rdtp;
 	long oldval;
 
 	lockdep_assert_irqs_disabled();
 	rdtp = this_cpu_ptr(&rcu_dynticks);
-	oldval = rdtp->dynticks_nesting;
+	rdp = this_cpu_ptr(&rcu_data);
+	oldval = rdp->dynticks_nesting;
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0);
 	if (oldval) {
-		rdtp->dynticks_nesting++;
+		rdp->dynticks_nesting++;
 		return;
 	}
 	rcu_dynticks_task_exit();
 	rcu_dynticks_eqs_exit();
 	rcu_cleanup_after_idle();
-	trace_rcu_dyntick(TPS("End"), rdtp->dynticks_nesting, 1, rdtp->dynticks);
+	trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, rdtp->dynticks);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
-	WRITE_ONCE(rdtp->dynticks_nesting, 1);
-	WARN_ON_ONCE(rdtp->dynticks_nmi_nesting);
-	WRITE_ONCE(rdtp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE);
+	WRITE_ONCE(rdp->dynticks_nesting, 1);
+	WARN_ON_ONCE(rdp->dynticks_nmi_nesting);
+	WRITE_ONCE(rdp->dynticks_nmi_nesting, DYNTICK_IRQ_NONIDLE);
 }
 
 /**
@@ -798,7 +802,7 @@ void rcu_user_exit(void)
  * @irq: Is this call from rcu_irq_enter?
  *
  * If the CPU was idle from RCU's viewpoint, update rdtp->dynticks and
- * rdtp->dynticks_nmi_nesting to let the RCU grace-period handling know
+ * rdp->dynticks_nmi_nesting to let the RCU grace-period handling know
  * that the CPU is active.  This implementation permits nested NMIs, as
  * long as the nesting level does not overflow an int.  (You will probably
  * run out of stack space first.)
@@ -808,11 +812,12 @@ void rcu_user_exit(void)
  */
 static __always_inline void rcu_nmi_enter_common(bool irq)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
+	struct rcu_dynticks __maybe_unused *rdtp = this_cpu_ptr(&rcu_dynticks);
 	long incby = 2;
 
 	/* Complain about underflow. */
-	WARN_ON_ONCE(rdtp->dynticks_nmi_nesting < 0);
+	WARN_ON_ONCE(rdp->dynticks_nmi_nesting < 0);
 
 	/*
 	 * If idle from RCU viewpoint, atomically increment ->dynticks
@@ -835,10 +840,10 @@ static __always_inline void rcu_nmi_enter_common(bool irq)
 		incby = 1;
 	}
 	trace_rcu_dyntick(incby == 1 ? TPS("Endirq") : TPS("++="),
-			  rdtp->dynticks_nmi_nesting,
-			  rdtp->dynticks_nmi_nesting + incby, rdtp->dynticks);
-	WRITE_ONCE(rdtp->dynticks_nmi_nesting, /* Prevent store tearing. */
-		   rdtp->dynticks_nmi_nesting + incby);
+			  rdp->dynticks_nmi_nesting,
+			  rdp->dynticks_nmi_nesting + incby, rdtp->dynticks);
+	WRITE_ONCE(rdp->dynticks_nmi_nesting, /* Prevent store tearing. */
+		   rdp->dynticks_nmi_nesting + incby);
 	barrier();
 }
 
@@ -3193,7 +3198,7 @@ rcu_boot_init_percpu_data(int cpu)
 	/* Set up local state, ensuring consistent view of global state. */
 	rdp->grpmask = leaf_node_cpu_bit(rdp->mynode, cpu);
 	rdp->dynticks = &per_cpu(rcu_dynticks, cpu);
-	WARN_ON_ONCE(rdp->dynticks->dynticks_nesting != 1);
+	WARN_ON_ONCE(rdp->dynticks_nesting != 1);
 	WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp->dynticks)));
 	rdp->rcu_ofl_gp_seq = rcu_state.gp_seq;
 	rdp->rcu_ofl_gp_flags = RCU_GP_CLEANED;
@@ -3226,7 +3231,7 @@ int rcutree_prepare_cpu(unsigned int cpu)
 	if (rcu_segcblist_empty(&rdp->cblist) && /* No early-boot CBs? */
 	    !init_nocb_callback_list(rdp))
 		rcu_segcblist_init(&rdp->cblist);  /* Re-enable callbacks. */
-	rdp->dynticks->dynticks_nesting = 1;	/* CPU not up, no tearing. */
+	rdp->dynticks_nesting = 1;	/* CPU not up, no tearing. */
 	rcu_dynticks_eqs_online();
 	raw_spin_unlock_rcu_node(rnp);		/* irqs remain disabled. */
 
diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 4c31066ddb94..2e5eec48a94a 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -38,8 +38,6 @@
  * Dynticks per-CPU state.
  */
 struct rcu_dynticks {
-	long dynticks_nesting;      /* Track process nesting level. */
-	long dynticks_nmi_nesting;  /* Track irq/NMI nesting level. */
 	atomic_t dynticks;	    /* Even value for idle, else odd. */
 };
 
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 161760957a07..7087ee3e1ea5 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1831,7 +1831,7 @@ static void print_cpu_stall_info(int cpu)
 				"!."[!delta],
 	       ticks_value, ticks_title,
 	       rcu_dynticks_snap(rdtp) & 0xfff,
-	       rdtp->dynticks_nesting, rdtp->dynticks_nmi_nesting,
+	       rdp->dynticks_nesting, rdp->dynticks_nmi_nesting,
 	       rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu),
 	       READ_ONCE(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart,
 	       fast_no_hz);
-- 
2.17.1

[PATCH tip/core/rcu 08/10] rcu: Switch ->dynticks to rcu_data structure, remove rcu_dynticks

[Paul E. McKenney]

This commit move ->dynticks from the rcu_dynticks structure to the
rcu_data structure, replacing the field of the same name.  It also updates
the code to access ->dynticks from the rcu_data structure and to use the
rcu_data structure rather than following to now-gone ->dynticks field
to the now-gone rcu_dynticks structure.  While in the area, this commit
also fixes up comments.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.c        | 92 ++++++++++++++++++----------------------
 kernel/rcu/tree.h        | 35 ++++++---------
 kernel/rcu/tree_exp.h    |  6 +--
 kernel/rcu/tree_plugin.h |  3 +-
 4 files changed, 56 insertions(+), 80 deletions(-)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index 35326d1005eb..ad09707d99ee 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -73,9 +73,20 @@
 
 /* Data structures. */
 
+/*
+ * Steal a bit from the bottom of ->dynticks for idle entry/exit
+ * control.  Initially this is for TLB flushing.
+ */
+#define RCU_DYNTICK_CTRL_MASK 0x1
+#define RCU_DYNTICK_CTRL_CTR  (RCU_DYNTICK_CTRL_MASK + 1)
+#ifndef rcu_eqs_special_exit
+#define rcu_eqs_special_exit() do { } while (0)
+#endif
+
 static DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, rcu_data) = {
 	.dynticks_nesting = 1,
 	.dynticks_nmi_nesting = DYNTICK_IRQ_NONIDLE,
+	.dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR),
 };
 struct rcu_state rcu_state = {
 	.level = { &rcu_state.node[0] },
@@ -202,27 +213,13 @@ void rcu_softirq_qs(void)
 	rcu_preempt_deferred_qs(current);
 }
 
-/*
- * Steal a bit from the bottom of ->dynticks for idle entry/exit
- * control.  Initially this is for TLB flushing.
- */
-#define RCU_DYNTICK_CTRL_MASK 0x1
-#define RCU_DYNTICK_CTRL_CTR  (RCU_DYNTICK_CTRL_MASK + 1)
-#ifndef rcu_eqs_special_exit
-#define rcu_eqs_special_exit() do { } while (0)
-#endif
-
-static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = {
-	.dynticks = ATOMIC_INIT(RCU_DYNTICK_CTRL_CTR),
-};
-
 /*
  * Record entry into an extended quiescent state.  This is only to be
  * called when not already in an extended quiescent state.
  */
 static void rcu_dynticks_eqs_enter(void)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	int seq;
 
 	/*
@@ -230,7 +227,7 @@ static void rcu_dynticks_eqs_enter(void)
 	 * critical sections, and we also must force ordering with the
 	 * next idle sojourn.
 	 */
-	seq = atomic_add_return(RCU_DYNTICK_CTRL_CTR, &rdtp->dynticks);
+	seq = atomic_add_return(RCU_DYNTICK_CTRL_CTR, &rdp->dynticks);
 	/* Better be in an extended quiescent state! */
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) &&
 		     (seq & RCU_DYNTICK_CTRL_CTR));
@@ -245,7 +242,7 @@ static void rcu_dynticks_eqs_enter(void)
  */
 static void rcu_dynticks_eqs_exit(void)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 	int seq;
 
 	/*
@@ -253,11 +250,11 @@ static void rcu_dynticks_eqs_exit(void)
 	 * and we also must force ordering with the next RCU read-side
 	 * critical section.
 	 */
-	seq = atomic_add_return(RCU_DYNTICK_CTRL_CTR, &rdtp->dynticks);
+	seq = atomic_add_return(RCU_DYNTICK_CTRL_CTR, &rdp->dynticks);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) &&
 		     !(seq & RCU_DYNTICK_CTRL_CTR));
 	if (seq & RCU_DYNTICK_CTRL_MASK) {
-		atomic_andnot(RCU_DYNTICK_CTRL_MASK, &rdtp->dynticks);
+		atomic_andnot(RCU_DYNTICK_CTRL_MASK, &rdp->dynticks);
 		smp_mb__after_atomic(); /* _exit after clearing mask. */
 		/* Prefer duplicate flushes to losing a flush. */
 		rcu_eqs_special_exit();
@@ -276,11 +273,11 @@ static void rcu_dynticks_eqs_exit(void)
  */
 static void rcu_dynticks_eqs_online(void)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 
-	if (atomic_read(&rdtp->dynticks) & RCU_DYNTICK_CTRL_CTR)
+	if (atomic_read(&rdp->dynticks) & RCU_DYNTICK_CTRL_CTR)
 		return;
-	atomic_add(RCU_DYNTICK_CTRL_CTR, &rdtp->dynticks);
+	atomic_add(RCU_DYNTICK_CTRL_CTR, &rdp->dynticks);
 }
 
 /*
@@ -290,18 +287,18 @@ static void rcu_dynticks_eqs_online(void)
  */
 bool rcu_dynticks_curr_cpu_in_eqs(void)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
+	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
 
-	return !(atomic_read(&rdtp->dynticks) & RCU_DYNTICK_CTRL_CTR);
+	return !(atomic_read(&rdp->dynticks) & RCU_DYNTICK_CTRL_CTR);
 }
 
 /*
  * Snapshot the ->dynticks counter with full ordering so as to allow
  * stable comparison of this counter with past and future snapshots.
  */
-int rcu_dynticks_snap(struct rcu_dynticks *rdtp)
+int rcu_dynticks_snap(struct rcu_data *rdp)
 {
-	int snap = atomic_add_return(0, &rdtp->dynticks);
+	int snap = atomic_add_return(0, &rdp->dynticks);
 
 	return snap & ~RCU_DYNTICK_CTRL_MASK;
 }
@@ -316,13 +313,13 @@ static bool rcu_dynticks_in_eqs(int snap)
 }
 
 /*
- * Return true if the CPU corresponding to the specified rcu_dynticks
+ * Return true if the CPU corresponding to the specified rcu_data
  * structure has spent some time in an extended quiescent state since
  * rcu_dynticks_snap() returned the specified snapshot.
  */
-static bool rcu_dynticks_in_eqs_since(struct rcu_dynticks *rdtp, int snap)
+static bool rcu_dynticks_in_eqs_since(struct rcu_data *rdp, int snap)
 {
-	return snap != rcu_dynticks_snap(rdtp);
+	return snap != rcu_dynticks_snap(rdp);
 }
 
 /*
@@ -336,14 +333,14 @@ bool rcu_eqs_special_set(int cpu)
 {
 	int old;
 	int new;
-	struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
+	struct rcu_data *rdp = &per_cpu(rcu_data, cpu);
 
 	do {
-		old = atomic_read(&rdtp->dynticks);
+		old = atomic_read(&rdp->dynticks);
 		if (old & RCU_DYNTICK_CTRL_CTR)
 			return false;
 		new = old | RCU_DYNTICK_CTRL_MASK;
-	} while (atomic_cmpxchg(&rdtp->dynticks, old, new) != old);
+	} while (atomic_cmpxchg(&rdp->dynticks, old, new) != old);
 	return true;
 }
 
@@ -360,11 +357,11 @@ bool rcu_eqs_special_set(int cpu)
  */
 static void __maybe_unused rcu_momentary_dyntick_idle(void)
 {
-	struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
 	int special;
 
 	raw_cpu_write(rcu_data.rcu_need_heavy_qs, false);
-	special = atomic_add_return(2 * RCU_DYNTICK_CTRL_CTR, &rdtp->dynticks);
+	special = atomic_add_return(2 * RCU_DYNTICK_CTRL_CTR,
+				    &this_cpu_ptr(&rcu_data)->dynticks);
 	/* It is illegal to call this from idle state. */
 	WARN_ON_ONCE(!(special & RCU_DYNTICK_CTRL_CTR));
 	rcu_preempt_deferred_qs(current);
@@ -573,9 +570,7 @@ static struct rcu_node *rcu_get_root(void)
 static void rcu_eqs_enter(bool user)
 {
 	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-	struct rcu_dynticks *rdtp;
 
-	rdtp = this_cpu_ptr(&rcu_dynticks);
 	WARN_ON_ONCE(rdp->dynticks_nmi_nesting != DYNTICK_IRQ_NONIDLE);
 	WRITE_ONCE(rdp->dynticks_nmi_nesting, 0);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) &&
@@ -586,7 +581,7 @@ static void rcu_eqs_enter(bool user)
 	}
 
 	lockdep_assert_irqs_disabled();
-	trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, rdtp->dynticks);
+	trace_rcu_dyntick(TPS("Start"), rdp->dynticks_nesting, 0, rdp->dynticks);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
 	rdp = this_cpu_ptr(&rcu_data);
 	do_nocb_deferred_wakeup(rdp);
@@ -638,7 +633,7 @@ void rcu_user_enter(void)
  * @irq: Is this call from rcu_irq_exit?
  *
  * If we are returning from the outermost NMI handler that interrupted an
- * RCU-idle period, update rdtp->dynticks and rdp->dynticks_nmi_nesting
+ * RCU-idle period, update rdp->dynticks and rdp->dynticks_nmi_nesting
  * to let the RCU grace-period handling know that the CPU is back to
  * being RCU-idle.
  *
@@ -648,7 +643,6 @@ void rcu_user_enter(void)
 static __always_inline void rcu_nmi_exit_common(bool irq)
 {
 	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-	struct rcu_dynticks __maybe_unused *rdtp = this_cpu_ptr(&rcu_dynticks);
 
 	/*
 	 * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks.
@@ -663,14 +657,14 @@ static __always_inline void rcu_nmi_exit_common(bool irq)
 	 * leave it in non-RCU-idle state.
 	 */
 	if (rdp->dynticks_nmi_nesting != 1) {
-		trace_rcu_dyntick(TPS("--="), rdp->dynticks_nmi_nesting, rdp->dynticks_nmi_nesting - 2, rdtp->dynticks);
+		trace_rcu_dyntick(TPS("--="), rdp->dynticks_nmi_nesting, rdp->dynticks_nmi_nesting - 2, rdp->dynticks);
 		WRITE_ONCE(rdp->dynticks_nmi_nesting, /* No store tearing. */
 			   rdp->dynticks_nmi_nesting - 2);
 		return;
 	}
 
 	/* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */
-	trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, rdtp->dynticks);
+	trace_rcu_dyntick(TPS("Startirq"), rdp->dynticks_nmi_nesting, 0, rdp->dynticks);
 	WRITE_ONCE(rdp->dynticks_nmi_nesting, 0); /* Avoid store tearing. */
 
 	if (irq)
@@ -741,11 +735,9 @@ void rcu_irq_exit_irqson(void)
 static void rcu_eqs_exit(bool user)
 {
 	struct rcu_data *rdp;
-	struct rcu_dynticks *rdtp;
 	long oldval;
 
 	lockdep_assert_irqs_disabled();
-	rdtp = this_cpu_ptr(&rcu_dynticks);
 	rdp = this_cpu_ptr(&rcu_data);
 	oldval = rdp->dynticks_nesting;
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && oldval < 0);
@@ -756,7 +748,7 @@ static void rcu_eqs_exit(bool user)
 	rcu_dynticks_task_exit();
 	rcu_dynticks_eqs_exit();
 	rcu_cleanup_after_idle();
-	trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, rdtp->dynticks);
+	trace_rcu_dyntick(TPS("End"), rdp->dynticks_nesting, 1, rdp->dynticks);
 	WARN_ON_ONCE(IS_ENABLED(CONFIG_RCU_EQS_DEBUG) && !user && !is_idle_task(current));
 	WRITE_ONCE(rdp->dynticks_nesting, 1);
 	WARN_ON_ONCE(rdp->dynticks_nmi_nesting);
@@ -801,7 +793,7 @@ void rcu_user_exit(void)
  * rcu_nmi_enter_common - inform RCU of entry to NMI context
  * @irq: Is this call from rcu_irq_enter?
  *
- * If the CPU was idle from RCU's viewpoint, update rdtp->dynticks and
+ * If the CPU was idle from RCU's viewpoint, update rdp->dynticks and
  * rdp->dynticks_nmi_nesting to let the RCU grace-period handling know
  * that the CPU is active.  This implementation permits nested NMIs, as
  * long as the nesting level does not overflow an int.  (You will probably
@@ -813,7 +805,6 @@ void rcu_user_exit(void)
 static __always_inline void rcu_nmi_enter_common(bool irq)
 {
 	struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-	struct rcu_dynticks __maybe_unused *rdtp = this_cpu_ptr(&rcu_dynticks);
 	long incby = 2;
 
 	/* Complain about underflow. */
@@ -841,7 +832,7 @@ static __always_inline void rcu_nmi_enter_common(bool irq)
 	}
 	trace_rcu_dyntick(incby == 1 ? TPS("Endirq") : TPS("++="),
 			  rdp->dynticks_nmi_nesting,
-			  rdp->dynticks_nmi_nesting + incby, rdtp->dynticks);
+			  rdp->dynticks_nmi_nesting + incby, rdp->dynticks);
 	WRITE_ONCE(rdp->dynticks_nmi_nesting, /* Prevent store tearing. */
 		   rdp->dynticks_nmi_nesting + incby);
 	barrier();
@@ -994,7 +985,7 @@ static void rcu_gpnum_ovf(struct rcu_node *rnp, struct rcu_data *rdp)
  */
 static int dyntick_save_progress_counter(struct rcu_data *rdp)
 {
-	rdp->dynticks_snap = rcu_dynticks_snap(rdp->dynticks);
+	rdp->dynticks_snap = rcu_dynticks_snap(rdp);
 	if (rcu_dynticks_in_eqs(rdp->dynticks_snap)) {
 		trace_rcu_fqs(rcu_state.name, rdp->gp_seq, rdp->cpu, TPS("dti"));
 		rcu_gpnum_ovf(rdp->mynode, rdp);
@@ -1045,7 +1036,7 @@ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp)
 	 * read-side critical section that started before the beginning
 	 * of the current RCU grace period.
 	 */
-	if (rcu_dynticks_in_eqs_since(rdp->dynticks, rdp->dynticks_snap)) {
+	if (rcu_dynticks_in_eqs_since(rdp, rdp->dynticks_snap)) {
 		trace_rcu_fqs(rcu_state.name, rdp->gp_seq, rdp->cpu, TPS("dti"));
 		rdp->dynticks_fqs++;
 		rcu_gpnum_ovf(rnp, rdp);
@@ -3197,9 +3188,8 @@ rcu_boot_init_percpu_data(int cpu)
 
 	/* Set up local state, ensuring consistent view of global state. */
 	rdp->grpmask = leaf_node_cpu_bit(rdp->mynode, cpu);
-	rdp->dynticks = &per_cpu(rcu_dynticks, cpu);
 	WARN_ON_ONCE(rdp->dynticks_nesting != 1);
-	WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp->dynticks)));
+	WARN_ON_ONCE(rcu_dynticks_in_eqs(rcu_dynticks_snap(rdp)));
 	rdp->rcu_ofl_gp_seq = rcu_state.gp_seq;
 	rdp->rcu_ofl_gp_flags = RCU_GP_CLEANED;
 	rdp->rcu_onl_gp_seq = rcu_state.gp_seq;
diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index 2e5eec48a94a..af8681fec23b 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -34,13 +34,6 @@
 
 #include "rcu_segcblist.h"
 
-/*
- * Dynticks per-CPU state.
- */
-struct rcu_dynticks {
-	atomic_t dynticks;	    /* Even value for idle, else odd. */
-};
-
 /* Communicate arguments to a workqueue handler. */
 struct rcu_exp_work {
 	smp_call_func_t rew_func;
@@ -194,24 +187,20 @@ struct rcu_data {
 	long		blimit;		/* Upper limit on a processed batch */
 
 	/* 3) dynticks interface. */
-	struct rcu_dynticks *dynticks;	/* Shared per-CPU dynticks state. */
 	int dynticks_snap;		/* Per-GP tracking for dynticks. */
-	long dynticks_nesting;      /* Track process nesting level. */
-	long dynticks_nmi_nesting;  /* Track irq/NMI nesting level. */
-	// atomic_t dynticks;	    /* Even value for idle, else odd. */
-	bool rcu_need_heavy_qs;     /* GP old, need heavy quiescent state. */
-	bool rcu_urgent_qs;	    /* GP old need light quiescent state. */
+	long dynticks_nesting;		/* Track process nesting level. */
+	long dynticks_nmi_nesting;	/* Track irq/NMI nesting level. */
+	atomic_t dynticks;		/* Even value for idle, else odd. */
+	bool rcu_need_heavy_qs;		/* GP old, so heavy quiescent state! */
+	bool rcu_urgent_qs;		/* GP old need light quiescent state. */
 #ifdef CONFIG_RCU_FAST_NO_HZ
-	bool all_lazy;		    /* Are all CPU's CBs lazy? */
-	unsigned long nonlazy_posted;
-				    /* # times non-lazy CBs posted to CPU. */
+	bool all_lazy;			/* Are all CPU's CBs lazy? */
+	unsigned long nonlazy_posted;	/* # times non-lazy CB posted to CPU. */
 	unsigned long nonlazy_posted_snap;
-				    /* idle-period nonlazy_posted snapshot. */
-	unsigned long last_accelerate;
-				    /* Last jiffy CBs were accelerated. */
-	unsigned long last_advance_all;
-				    /* Last jiffy CBs were all advanced. */
-	int tick_nohz_enabled_snap; /* Previously seen value from sysfs. */
+					/* Nonlazy_posted snapshot. */
+	unsigned long last_accelerate;	/* Last jiffy CBs were accelerated. */
+	unsigned long last_advance_all;	/* Last jiffy CBs were all advanced. */
+	int tick_nohz_enabled_snap;	/* Previously seen value from sysfs. */
 #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 
 	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
@@ -426,7 +415,7 @@ extern struct rcu_state rcu_bh_state;
 extern struct rcu_state rcu_preempt_state;
 #endif /* #ifdef CONFIG_PREEMPT_RCU */
 
-int rcu_dynticks_snap(struct rcu_dynticks *rdtp);
+int rcu_dynticks_snap(struct rcu_data *rdp);
 
 #ifdef CONFIG_RCU_BOOST
 DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status);
diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h
index 11387fcd4d85..8d18c1014e2b 100644
--- a/kernel/rcu/tree_exp.h
+++ b/kernel/rcu/tree_exp.h
@@ -360,14 +360,13 @@ static void sync_rcu_exp_select_node_cpus(struct work_struct *wp)
 	for_each_leaf_node_cpu_mask(rnp, cpu, rnp->expmask) {
 		unsigned long mask = leaf_node_cpu_bit(rnp, cpu);
 		struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu);
-		struct rcu_dynticks *rdtp = per_cpu_ptr(&rcu_dynticks, cpu);
 		int snap;
 
 		if (raw_smp_processor_id() == cpu ||
 		    !(rnp->qsmaskinitnext & mask)) {
 			mask_ofl_test |= mask;
 		} else {
-			snap = rcu_dynticks_snap(rdtp);
+			snap = rcu_dynticks_snap(rdp);
 			if (rcu_dynticks_in_eqs(snap))
 				mask_ofl_test |= mask;
 			else
@@ -393,8 +392,7 @@ static void sync_rcu_exp_select_node_cpus(struct work_struct *wp)
 		if (!(mask_ofl_ipi & mask))
 			continue;
 retry_ipi:
-		if (rcu_dynticks_in_eqs_since(rdp->dynticks,
-					      rdp->exp_dynticks_snap)) {
+		if (rcu_dynticks_in_eqs_since(rdp, rdp->exp_dynticks_snap)) {
 			mask_ofl_test |= mask;
 			continue;
 		}
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 7087ee3e1ea5..05915e536336 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1802,7 +1802,6 @@ static void print_cpu_stall_info(int cpu)
 	unsigned long delta;
 	char fast_no_hz[72];
 	struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu);
-	struct rcu_dynticks *rdtp = rdp->dynticks;
 	char *ticks_title;
 	unsigned long ticks_value;
 
@@ -1830,7 +1829,7 @@ static void print_cpu_stall_info(int cpu)
 			rdp->rcu_iw_pending ? (int)min(delta, 9UL) + '0' :
 				"!."[!delta],
 	       ticks_value, ticks_title,
-	       rcu_dynticks_snap(rdtp) & 0xfff,
+	       rcu_dynticks_snap(rdp) & 0xfff,
 	       rdp->dynticks_nesting, rdp->dynticks_nmi_nesting,
 	       rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu),
 	       READ_ONCE(rcu_state.n_force_qs) - rcu_state.n_force_qs_gpstart,
-- 
2.17.1

[PATCH tip/core/rcu 09/10] rcu: Remove obsolete ->dynticks_fqs and ->cond_resched_completed

[Paul E. McKenney]

The rcu_data structure's ->dynticks_fqs is incremented but never
accesses.  Its ->cond_resched_completed field isn't used at all.
This commit therefore removes both fields.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
---
 kernel/rcu/tree.c |  1 -
 kernel/rcu/tree.h | 12 +++---------
 2 files changed, 3 insertions(+), 10 deletions(-)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index ad09707d99ee..65b6e03ae6f5 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -1038,7 +1038,6 @@ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp)
 	 */
 	if (rcu_dynticks_in_eqs_since(rdp, rdp->dynticks_snap)) {
 		trace_rcu_fqs(rcu_state.name, rdp->gp_seq, rdp->cpu, TPS("dti"));
-		rdp->dynticks_fqs++;
 		rcu_gpnum_ovf(rnp, rdp);
 		return 1;
 	}
diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index af8681fec23b..bfbf97a1c29d 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -203,17 +203,11 @@ struct rcu_data {
 	int tick_nohz_enabled_snap;	/* Previously seen value from sysfs. */
 #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
 
-	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
-	unsigned long dynticks_fqs;	/* Kicked due to dynticks idle. */
-	unsigned long cond_resched_completed;
-					/* Grace period that needs help */
-					/*  from cond_resched(). */
-
-	/* 5) rcu_barrier(), OOM callbacks, and expediting. */
+	/* 4) rcu_barrier(), OOM callbacks, and expediting. */
 	struct rcu_head barrier_head;
 	int exp_dynticks_snap;		/* Double-check need for IPI. */
 
-	/* 6) Callback offloading. */
+	/* 5) Callback offloading. */
 #ifdef CONFIG_RCU_NOCB_CPU
 	struct rcu_head *nocb_head;	/* CBs waiting for kthread. */
 	struct rcu_head **nocb_tail;
@@ -240,7 +234,7 @@ struct rcu_data {
 					/* Leader CPU takes GP-end wakeups. */
 #endif /* #ifdef CONFIG_RCU_NOCB_CPU */
 
-	/* 7) Diagnostic data, including RCU CPU stall warnings. */
+	/* 6) Diagnostic data, including RCU CPU stall warnings. */
 	unsigned int softirq_snap;	/* Snapshot of softirq activity. */
 	/* ->rcu_iw* fields protected by leaf rcu_node ->lock. */
 	struct irq_work rcu_iw;		/* Check for non-irq activity. */
-- 
2.17.1

[PATCH tip/core/rcu 10/10] rcu: Convert rcu_state.ofl_lock to raw_spinlock_t

[Paul E. McKenney]

From: Mike Galbraith <efault@gmx.de>

1e64b15a4b10 ("rcu: Fix grace-period hangs due to race with CPU offline")
added spinlock_t ofl_lock to the rcu_state structure, then takes it with
preemption disabled during CPU offline, which gives the -rt patchset's
sleeping spinlock heartburn.

This commit therefore converts ->ofl_lock to raw_spinlock_t.

Signed-off-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
 kernel/rcu/tree.c | 12 ++++++------
 kernel/rcu/tree.h |  2 +-
 2 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index 65b6e03ae6f5..0ce4702ce09f 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -97,7 +97,7 @@ struct rcu_state rcu_state = {
 	.abbr = RCU_ABBR,
 	.exp_mutex = __MUTEX_INITIALIZER(rcu_state.exp_mutex),
 	.exp_wake_mutex = __MUTEX_INITIALIZER(rcu_state.exp_wake_mutex),
-	.ofl_lock = __SPIN_LOCK_UNLOCKED(rcu_state.ofl_lock),
+	.ofl_lock = __RAW_SPIN_LOCK_UNLOCKED(rcu_state.ofl_lock),
 };
 
 /* Dump rcu_node combining tree at boot to verify correct setup. */
@@ -1775,13 +1775,13 @@ static bool rcu_gp_init(void)
 	 */
 	rcu_state.gp_state = RCU_GP_ONOFF;
 	rcu_for_each_leaf_node(rnp) {
-		spin_lock(&rcu_state.ofl_lock);
+		raw_spin_lock(&rcu_state.ofl_lock);
 		raw_spin_lock_irq_rcu_node(rnp);
 		if (rnp->qsmaskinit == rnp->qsmaskinitnext &&
 		    !rnp->wait_blkd_tasks) {
 			/* Nothing to do on this leaf rcu_node structure. */
 			raw_spin_unlock_irq_rcu_node(rnp);
-			spin_unlock(&rcu_state.ofl_lock);
+			raw_spin_unlock(&rcu_state.ofl_lock);
 			continue;
 		}
 
@@ -1817,7 +1817,7 @@ static bool rcu_gp_init(void)
 		}
 
 		raw_spin_unlock_irq_rcu_node(rnp);
-		spin_unlock(&rcu_state.ofl_lock);
+		raw_spin_unlock(&rcu_state.ofl_lock);
 	}
 	rcu_gp_slow(gp_preinit_delay); /* Races with CPU hotplug. */
 
@@ -3376,7 +3376,7 @@ void rcu_report_dead(unsigned int cpu)
 
 	/* Remove outgoing CPU from mask in the leaf rcu_node structure. */
 	mask = rdp->grpmask;
-	spin_lock(&rcu_state.ofl_lock);
+	raw_spin_lock(&rcu_state.ofl_lock);
 	raw_spin_lock_irqsave_rcu_node(rnp, flags); /* Enforce GP memory-order guarantee. */
 	rdp->rcu_ofl_gp_seq = READ_ONCE(rcu_state.gp_seq);
 	rdp->rcu_ofl_gp_flags = READ_ONCE(rcu_state.gp_flags);
@@ -3387,7 +3387,7 @@ void rcu_report_dead(unsigned int cpu)
 	}
 	rnp->qsmaskinitnext &= ~mask;
 	raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
-	spin_unlock(&rcu_state.ofl_lock);
+	raw_spin_unlock(&rcu_state.ofl_lock);
 
 	per_cpu(rcu_cpu_started, cpu) = 0;
 }
diff --git a/kernel/rcu/tree.h b/kernel/rcu/tree.h
index bfbf97a1c29d..703e19ff532d 100644
--- a/kernel/rcu/tree.h
+++ b/kernel/rcu/tree.h
@@ -343,7 +343,7 @@ struct rcu_state {
 	const char *name;			/* Name of structure. */
 	char abbr;				/* Abbreviated name. */
 
-	spinlock_t ofl_lock ____cacheline_internodealigned_in_smp;
+	raw_spinlock_t ofl_lock ____cacheline_internodealigned_in_smp;
 						/* Synchronize offline with */
 						/*  GP pre-initialization. */
 };
-- 
2.17.1