Preview of RCU changes for 3.2

index 2e138db..920eb50 100644
@@ -1282,8 +1282,9 @@ static int rcutorture_booster_init(int cpu)
 	/* Don't allow time recalculation while creating a new task. */
 	mutex_lock(&boost_mutex);
 	VERBOSE_PRINTK_STRING("Creating rcu_torture_boost task");
-	boost_tasks[cpu] = kthread_create(rcu_torture_boost, NULL,
-					  "rcu_torture_boost");
+	boost_tasks[cpu] = kthread_create_on_node(rcu_torture_boost, NULL,
+						  cpu_to_node(cpu),
+						  "rcu_torture_boost");
 	if (IS_ERR(boost_tasks[cpu])) {
 		retval = PTR_ERR(boost_tasks[cpu]);
 		VERBOSE_PRINTK_STRING("rcu_torture_boost task create failed");

index 8aafbb8..7b850cd 100644
@@ -1560,7 +1560,10 @@ static int __cpuinit rcu_spawn_one_cpu_kthread(int cpu)
 	if (!rcu_scheduler_fully_active ||
 	    per_cpu(rcu_cpu_kthread_task, cpu) != NULL)
 		return 0;
-	t = kthread_create(rcu_cpu_kthread, (void *)(long)cpu, "rcuc%d", cpu);
+	t = kthread_create_on_node(rcu_cpu_kthread,
+				   (void *)(long)cpu,
+				   cpu_to_node(cpu),
+				   "rcuc%d", cpu);
 	if (IS_ERR(t))
 		return PTR_ERR(t);
 	if (cpu_online(cpu))

index 7b850cd..9703298 100644
@@ -1291,11 +1291,9 @@ static void invoke_rcu_callbacks_kthread(void)
 
 	local_irq_save(flags);
 	__this_cpu_write(rcu_cpu_has_work, 1);
-	if (__this_cpu_read(rcu_cpu_kthread_task) == NULL) {
-		local_irq_restore(flags);
-		return;
-	}
-	wake_up_process(__this_cpu_read(rcu_cpu_kthread_task));
+	if (__this_cpu_read(rcu_cpu_kthread_task) != NULL &&
+	    current != __this_cpu_read(rcu_cpu_kthread_task))
+		wake_up_process(__this_cpu_read(rcu_cpu_kthread_task));
 	local_irq_restore(flags);
 }
 

index ef820a3..b6a56e3 100644
@@ -548,7 +548,7 @@ do {									\
 #endif
 
 #ifdef CONFIG_PROVE_RCU
-extern void lockdep_rcu_dereference(const char *file, const int line);
+void lockdep_rcu_suspicious(const char *file, const int line, const char *s);
 #endif
 
 #endif /* __LINUX_LOCKDEP_H */

index 99f9aa7..8be0433 100644
@@ -297,19 +297,31 @@ extern int rcu_my_thread_group_empty(void);
 /**
  * rcu_lockdep_assert - emit lockdep splat if specified condition not met
  * @c: condition to check
+ * @s: informative message
  */
-#define rcu_lockdep_assert(c)						\
+#define rcu_lockdep_assert(c, s)					\
 	do {								\
 		static bool __warned;					\
 		if (debug_lockdep_rcu_enabled() && !__warned && !(c)) {	\
 			__warned = true;				\
-			lockdep_rcu_dereference(__FILE__, __LINE__);	\
+			lockdep_rcu_suspicious(__FILE__, __LINE__, s);	\
 		}							\
 	} while (0)
 
+#define rcu_sleep_check()						\
+	do {								\
+		rcu_lockdep_assert(!lock_is_held(&rcu_bh_lock_map),	\
+				   "Illegal context switch in RCU-bh"	\
+				   " read-side critical section");	\
+		rcu_lockdep_assert(!lock_is_held(&rcu_sched_lock_map),	\
+				   "Illegal context switch in RCU-sched"\
+				   " read-side critical section");	\
+	} while (0)
+
 #else /* #ifdef CONFIG_PROVE_RCU */
 
-#define rcu_lockdep_assert(c) do { } while (0)
+#define rcu_lockdep_assert(c, s) do { } while (0)
+#define rcu_sleep_check() do { } while (0)
 
 #endif /* #else #ifdef CONFIG_PROVE_RCU */
 

@@ -338,14 +350,16 @@ extern int rcu_my_thread_group_empty(void);
 #define __rcu_dereference_check(p, c, space) \
 	({ \
 		typeof(*p) *_________p1 = (typeof(*p)*__force )ACCESS_ONCE(p); \
-		rcu_lockdep_assert(c); \
+		rcu_lockdep_assert(c, "suspicious rcu_dereference_check()" \
+				      " usage"); \
 		rcu_dereference_sparse(p, space); \
 		smp_read_barrier_depends(); \
 		((typeof(*p) __force __kernel *)(_________p1)); \
 	})
 #define __rcu_dereference_protected(p, c, space) \
 	({ \
-		rcu_lockdep_assert(c); \
+		rcu_lockdep_assert(c, "suspicious rcu_dereference_protected()" \
+				      " usage"); \
 		rcu_dereference_sparse(p, space); \
 		((typeof(*p) __force __kernel *)(p)); \
 	})

@@ -359,7 +373,9 @@ extern int rcu_my_thread_group_empty(void);
 #define __rcu_dereference_index_check(p, c) \
 	({ \
 		typeof(p) _________p1 = ACCESS_ONCE(p); \
-		rcu_lockdep_assert(c); \
+		rcu_lockdep_assert(c, \
+				   "suspicious rcu_dereference_index_check()" \
+				   " usage"); \
 		smp_read_barrier_depends(); \
 		(_________p1); \
 	})

165 lines not on this page · raw message

index 57a8346..a7577b3 100644
@@ -419,7 +419,9 @@ EXPORT_SYMBOL(pid_task);
  */
 struct task_struct *find_task_by_pid_ns(pid_t nr, struct pid_namespace *ns)
 {
-	rcu_lockdep_assert(rcu_read_lock_held());
+	rcu_lockdep_assert(rcu_read_lock_held(),
+			   "find_task_by_pid_ns() needs rcu_read_lock()"
+			   " protection");
 	return pid_task(find_pid_ns(nr, ns), PIDTYPE_PID);
 }
 

index fde6ff9..1c87917 100644
@@ -4200,6 +4200,7 @@ static inline void schedule_debug(struct task_struct *prev)
 	 */
 	if (unlikely(in_atomic_preempt_off() && !prev->exit_state))
 		__schedule_bug(prev);
+	rcu_sleep_check();
 
 	profile_hit(SCHED_PROFILING, __builtin_return_address(0));
 

@@ -8198,6 +8199,7 @@ void __might_sleep(const char *file, int line, int preempt_offset)
 #ifdef in_atomic
 	static unsigned long prev_jiffy;	/* ratelimiting */
 
+	rcu_sleep_check(); /* WARN_ON_ONCE() by default, no rate limit reqd. */
 	if ((preempt_count_equals(preempt_offset) && !irqs_disabled()) ||
 	    system_state != SYSTEM_RUNNING || oops_in_progress)
 		return;

index 8be0433..2516555 100644
@@ -33,6 +33,7 @@
 #ifndef __LINUX_RCUPDATE_H
 #define __LINUX_RCUPDATE_H
 
+#include <linux/types.h>
 #include <linux/cache.h>
 #include <linux/spinlock.h>
 #include <linux/threads.h>

@@ -64,16 +65,6 @@ static inline void rcutorture_record_progress(unsigned long vernum)
 #define ULONG_CMP_GE(a, b)	(ULONG_MAX / 2 >= (a) - (b))
 #define ULONG_CMP_LT(a, b)	(ULONG_MAX / 2 < (a) - (b))
 
-/**
- * struct rcu_head - callback structure for use with RCU
- * @next: next update requests in a list
- * @func: actual update function to call after the grace period.
- */
-struct rcu_head {
-	struct rcu_head *next;
-	void (*func)(struct rcu_head *head);
-};
-
 /* Exported common interfaces */
 extern void call_rcu_sched(struct rcu_head *head,
 			   void (*func)(struct rcu_head *rcu));

index 176da8c..57a9723 100644
@@ -238,6 +238,16 @@ struct ustat {
 	char			f_fpack[6];
 };
 
+/**
+ * struct rcu_head - callback structure for use with RCU
+ * @next: next update requests in a list
+ * @func: actual update function to call after the grace period.
+ */
+struct rcu_head {
+	struct rcu_head *next;
+	void (*func)(struct rcu_head *head);
+};
+
 #endif	/* __KERNEL__ */
 #endif /*  __ASSEMBLY__ */
 #endif /* _LINUX_TYPES_H */

index 4e14487..8827b34 100644
@@ -49,7 +49,7 @@
 #ifdef CONFIG_RCU_BOOST
 
 DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status);
-DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_cpu);
+DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu);
 DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops);
 DECLARE_PER_CPU(char, rcu_cpu_has_work);
 

index 2516555..6433a6f 100644
@@ -66,11 +66,73 @@ static inline void rcutorture_record_progress(unsigned long vernum)
 #define ULONG_CMP_LT(a, b)	(ULONG_MAX / 2 < (a) - (b))
 
 /* Exported common interfaces */
+
+#ifdef CONFIG_PREEMPT_RCU
+
+/**
+ * call_rcu() - Queue an RCU callback for invocation after a grace period.
+ * @head: structure to be used for queueing the RCU updates.
+ * @func: actual callback function to be invoked after the grace period
+ *
+ * The callback function will be invoked some time after a full grace
+ * period elapses, in other words after all pre-existing RCU read-side
+ * critical sections have completed.  However, the callback function
+ * might well execute concurrently with RCU read-side critical sections
+ * that started after call_rcu() was invoked.  RCU read-side critical
+ * sections are delimited by rcu_read_lock() and rcu_read_unlock(),
+ * and may be nested.
+ */
+extern void call_rcu(struct rcu_head *head,
+			      void (*func)(struct rcu_head *head));
+
+#else /* #ifdef CONFIG_PREEMPT_RCU */
+
+/* In classic RCU, call_rcu() is just call_rcu_sched(). */
+#define	call_rcu	call_rcu_sched
+
+#endif /* #else #ifdef CONFIG_PREEMPT_RCU */
+
+/**
+ * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period.
+ * @head: structure to be used for queueing the RCU updates.
+ * @func: actual callback function to be invoked after the grace period
+ *
+ * The callback function will be invoked some time after a full grace
+ * period elapses, in other words after all currently executing RCU
+ * read-side critical sections have completed. call_rcu_bh() assumes
+ * that the read-side critical sections end on completion of a softirq
+ * handler. This means that read-side critical sections in process
+ * context must not be interrupted by softirqs. This interface is to be
+ * used when most of the read-side critical sections are in softirq context.
+ * RCU read-side critical sections are delimited by :
+ *  - rcu_read_lock() and  rcu_read_unlock(), if in interrupt context.
+ *  OR
+ *  - rcu_read_lock_bh() and rcu_read_unlock_bh(), if in process context.
+ *  These may be nested.
+ */
+extern void call_rcu_bh(struct rcu_head *head,
+			void (*func)(struct rcu_head *head));
+
+/**
+ * call_rcu_sched() - Queue an RCU for invocation after sched grace period.
+ * @head: structure to be used for queueing the RCU updates.
+ * @func: actual callback function to be invoked after the grace period
+ *
+ * The callback function will be invoked some time after a full grace
+ * period elapses, in other words after all currently executing RCU
+ * read-side critical sections have completed. call_rcu_sched() assumes
+ * that the read-side critical sections end on enabling of preemption
+ * or on voluntary preemption.
+ * RCU read-side critical sections are delimited by :
+ *  - rcu_read_lock_sched() and  rcu_read_unlock_sched(),
+ *  OR
+ *  anything that disables preemption.
+ *  These may be nested.
+ */
 extern void call_rcu_sched(struct rcu_head *head,
 			   void (*func)(struct rcu_head *rcu));
+
 extern void synchronize_sched(void);
-extern void rcu_barrier_bh(void);
-extern void rcu_barrier_sched(void);
 
 static inline void __rcu_read_lock_bh(void)
 {

@@ -143,6 +205,15 @@ static inline void rcu_exit_nohz(void)
 
 #endif /* #else #ifdef CONFIG_NO_HZ */
 
+/*
+ * Infrastructure to implement the synchronize_() primitives in
+ * TREE_RCU and rcu_barrier_() primitives in TINY_RCU.
+ */
+
+typedef void call_rcu_func_t(struct rcu_head *head,
+			     void (*func)(struct rcu_head *head));
+void wait_rcu_gp(call_rcu_func_t crf);
+
 #if defined(CONFIG_TREE_RCU) || defined(CONFIG_TREE_PREEMPT_RCU)
 #include <linux/rcutree.h>
 #elif defined(CONFIG_TINY_RCU) || defined(CONFIG_TINY_PREEMPT_RCU)

@@ -723,61 +794,6 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
 #define RCU_INIT_POINTER(p, v) \
 		p = (typeof(*v) __force __rcu *)(v)
 
-/* Infrastructure to implement the synchronize_() primitives. */
-
-struct rcu_synchronize {
-	struct rcu_head head;
-	struct completion completion;
-};
-
-extern void wakeme_after_rcu(struct rcu_head  *head);
-
-#ifdef CONFIG_PREEMPT_RCU
-
-/**
- * call_rcu() - Queue an RCU callback for invocation after a grace period.
- * @head: structure to be used for queueing the RCU updates.
- * @func: actual callback function to be invoked after the grace period
- *
- * The callback function will be invoked some time after a full grace
- * period elapses, in other words after all pre-existing RCU read-side
- * critical sections have completed.  However, the callback function
- * might well execute concurrently with RCU read-side critical sections
- * that started after call_rcu() was invoked.  RCU read-side critical
- * sections are delimited by rcu_read_lock() and rcu_read_unlock(),
- * and may be nested.
- */
-extern void call_rcu(struct rcu_head *head,
-			      void (*func)(struct rcu_head *head));
-
-#else /* #ifdef CONFIG_PREEMPT_RCU */
-
-/* In classic RCU, call_rcu() is just call_rcu_sched(). */
-#define	call_rcu	call_rcu_sched
-
-#endif /* #else #ifdef CONFIG_PREEMPT_RCU */
-
-/**
- * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period.
- * @head: structure to be used for queueing the RCU updates.
- * @func: actual callback function to be invoked after the grace period
- *
- * The callback function will be invoked some time after a full grace
- * period elapses, in other words after all currently executing RCU
- * read-side critical sections have completed. call_rcu_bh() assumes
- * that the read-side critical sections end on completion of a softirq
- * handler. This means that read-side critical sections in process
- * context must not be interrupted by softirqs. This interface is to be
- * used when most of the read-side critical sections are in softirq context.
- * RCU read-side critical sections are delimited by :
- *  - rcu_read_lock() and  rcu_read_unlock(), if in interrupt context.
- *  OR
- *  - rcu_read_lock_bh() and rcu_read_unlock_bh(), if in process context.
- *  These may be nested.
- */
-extern void call_rcu_bh(struct rcu_head *head,
-			void (*func)(struct rcu_head *head));
-
 /*
  * debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
  * by call_rcu() and rcu callback execution, and are therefore not part of the

index 52b3e02..4eab233 100644
@@ -31,6 +31,16 @@ static inline void rcu_init(void)
 {
 }
 
+static inline void rcu_barrier_bh(void)
+{
+	wait_rcu_gp(call_rcu_bh);
+}
+
+static inline void rcu_barrier_sched(void)
+{
+	wait_rcu_gp(call_rcu_sched);
+}
+
 #ifdef CONFIG_TINY_RCU
 
 static inline void synchronize_rcu_expedited(void)

@@ -45,9 +55,13 @@ static inline void rcu_barrier(void)
 
 #else /* #ifdef CONFIG_TINY_RCU */
 
-void rcu_barrier(void);
 void synchronize_rcu_expedited(void);
 
+static inline void rcu_barrier(void)
+{
+	wait_rcu_gp(call_rcu);
+}
+
 #endif /* #else #ifdef CONFIG_TINY_RCU */
 
 static inline void synchronize_rcu_bh(void)

index e65d066..6745846 100644
@@ -67,6 +67,8 @@ static inline void synchronize_rcu_bh_expedited(void)
 }
 
 extern void rcu_barrier(void);
+extern void rcu_barrier_bh(void);
+extern void rcu_barrier_sched(void);
 
 extern unsigned long rcutorture_testseq;
 extern unsigned long rcutorture_vernum;

index 7784bd2..a088c90 100644
@@ -94,11 +94,16 @@ EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held);
 
 #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */
 
+struct rcu_synchronize {
+	struct rcu_head head;
+	struct completion completion;
+};
+
 /*
  * Awaken the corresponding synchronize_rcu() instance now that a
  * grace period has elapsed.
  */
-void wakeme_after_rcu(struct rcu_head  *head)
+static void wakeme_after_rcu(struct rcu_head  *head)
 {
 	struct rcu_synchronize *rcu;
 

@@ -106,6 +111,20 @@ void wakeme_after_rcu(struct rcu_head  *head)
 	complete(&rcu->completion);
 }
 
+void wait_rcu_gp(call_rcu_func_t crf)
+{
+	struct rcu_synchronize rcu;
+
+	init_rcu_head_on_stack(&rcu.head);
+	init_completion(&rcu.completion);
+	/* Will wake me after RCU finished. */
+	crf(&rcu.head, wakeme_after_rcu);
+	/* Wait for it. */
+	wait_for_completion(&rcu.completion);
+	destroy_rcu_head_on_stack(&rcu.head);
+}
+EXPORT_SYMBOL_GPL(wait_rcu_gp);
+
 #ifdef CONFIG_PROVE_RCU
 /*
  * wrapper function to avoid #include problems.

index 7bbac7d..f544e34 100644
@@ -281,34 +281,6 @@ void call_rcu_bh(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
 }
 EXPORT_SYMBOL_GPL(call_rcu_bh);
 
-void rcu_barrier_bh(void)
-{
-	struct rcu_synchronize rcu;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu_bh(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
-}
-EXPORT_SYMBOL_GPL(rcu_barrier_bh);
-
-void rcu_barrier_sched(void)
-{
-	struct rcu_synchronize rcu;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu_sched(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
-}
-EXPORT_SYMBOL_GPL(rcu_barrier_sched);
-
 /*
  * Spawn the kthread that invokes RCU callbacks.
  */

index f259c67..6b0cedb 100644
@@ -697,20 +697,6 @@ void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
 }
 EXPORT_SYMBOL_GPL(call_rcu);
 
-void rcu_barrier(void)
-{
-	struct rcu_synchronize rcu;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
-}
-EXPORT_SYMBOL_GPL(rcu_barrier);
-
 /*
  * synchronize_rcu - wait until a grace period has elapsed.
  *

index ba06207..a7c6bce 100644
@@ -1613,18 +1613,9 @@ EXPORT_SYMBOL_GPL(call_rcu_bh);
  */
 void synchronize_sched(void)
 {
-	struct rcu_synchronize rcu;
-
 	if (rcu_blocking_is_gp())
 		return;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu_sched(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
+	wait_rcu_gp(call_rcu_sched);
 }
 EXPORT_SYMBOL_GPL(synchronize_sched);
 

@@ -1639,18 +1630,9 @@ EXPORT_SYMBOL_GPL(synchronize_sched);
  */
 void synchronize_rcu_bh(void)
 {
-	struct rcu_synchronize rcu;
-
 	if (rcu_blocking_is_gp())
 		return;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu_bh(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
+	wait_rcu_gp(call_rcu_bh);
 }
 EXPORT_SYMBOL_GPL(synchronize_rcu_bh);
 

index 9703298..43daa46 100644
@@ -656,18 +656,9 @@ EXPORT_SYMBOL_GPL(call_rcu);
  */
 void synchronize_rcu(void)
 {
-	struct rcu_synchronize rcu;
-
 	if (!rcu_scheduler_active)
 		return;
-
-	init_rcu_head_on_stack(&rcu.head);
-	init_completion(&rcu.completion);
-	/* Will wake me after RCU finished. */
-	call_rcu(&rcu.head, wakeme_after_rcu);
-	/* Wait for it. */
-	wait_for_completion(&rcu.completion);
-	destroy_rcu_head_on_stack(&rcu.head);
+	wait_rcu_gp(call_rcu);
 }
 EXPORT_SYMBOL_GPL(synchronize_rcu);
 

103 lines not on this page · raw message

index e0f39ea..1194d6c 100644
@@ -796,11 +796,11 @@ checkwait:	rcu_stutter_wait("rcu_torture_boost");
 
 	/* Clean up and exit. */
 	VERBOSE_PRINTK_STRING("rcu_torture_boost task stopping");
-	destroy_rcu_head_on_stack(&rbi.rcu);
 	rcutorture_shutdown_absorb("rcu_torture_boost");
 	while (!kthread_should_stop() || rbi.inflight)
 		schedule_timeout_uninterruptible(1);
 	smp_mb(); /* order accesses to ->inflight before stack-frame death. */
+	destroy_rcu_head_on_stack(&rbi.rcu);
 	return 0;
 }
 

index 6433a6f..c61a535 100644
@@ -794,44 +794,6 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
 #define RCU_INIT_POINTER(p, v) \
 		p = (typeof(*v) __force __rcu *)(v)
 
-/*
- * debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
- * by call_rcu() and rcu callback execution, and are therefore not part of the
- * RCU API. Leaving in rcupdate.h because they are used by all RCU flavors.
- */
-
-#ifdef CONFIG_DEBUG_OBJECTS_RCU_HEAD
-# define STATE_RCU_HEAD_READY	0
-# define STATE_RCU_HEAD_QUEUED	1
-
-extern struct debug_obj_descr rcuhead_debug_descr;
-
-static inline void debug_rcu_head_queue(struct rcu_head *head)
-{
-	WARN_ON_ONCE((unsigned long)head & 0x3);
-	debug_object_activate(head, &rcuhead_debug_descr);
-	debug_object_active_state(head, &rcuhead_debug_descr,
-				  STATE_RCU_HEAD_READY,
-				  STATE_RCU_HEAD_QUEUED);
-}
-
-static inline void debug_rcu_head_unqueue(struct rcu_head *head)
-{
-	debug_object_active_state(head, &rcuhead_debug_descr,
-				  STATE_RCU_HEAD_QUEUED,
-				  STATE_RCU_HEAD_READY);
-	debug_object_deactivate(head, &rcuhead_debug_descr);
-}
-#else	/* !CONFIG_DEBUG_OBJECTS_RCU_HEAD */
-static inline void debug_rcu_head_queue(struct rcu_head *head)
-{
-}
-
-static inline void debug_rcu_head_unqueue(struct rcu_head *head)
-{
-}
-#endif	/* #else !CONFIG_DEBUG_OBJECTS_RCU_HEAD */
-
 static __always_inline bool __is_kfree_rcu_offset(unsigned long offset)
 {
 	return offset < 4096;

@@ -850,18 +812,6 @@ void __kfree_rcu(struct rcu_head *head, unsigned long offset)
 	call_rcu(head, (rcu_callback)offset);
 }
 
-extern void kfree(const void *);
-
-static inline void __rcu_reclaim(struct rcu_head *head)
-{
-	unsigned long offset = (unsigned long)head->func;
-
-	if (__is_kfree_rcu_offset(offset))
-		kfree((void *)head - offset);
-	else
-		head->func(head);
-}
-
 /**
  * kfree_rcu() - kfree an object after a grace period.
  * @ptr:	pointer to kfree

98 lines not on this page · raw message

79 lines not on this page · raw message

index a088c90..5031caf 100644
@@ -46,6 +46,11 @@
 #include <linux/module.h>
 #include <linux/hardirq.h>
 
+#define CREATE_TRACE_POINTS
+#include <trace/events/rcu.h>
+
+#include "rcu.h"
+
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 static struct lock_class_key rcu_lock_key;
 struct lockdep_map rcu_lock_map =

index f544e34..19453ba 100644
@@ -37,6 +37,25 @@
 #include <linux/cpu.h>
 #include <linux/prefetch.h>
 
+#ifdef CONFIG_RCU_TRACE
+
+#include <trace/events/rcu.h>
+
+#else /* #ifdef CONFIG_RCU_TRACE */
+
+/* No by-default tracing in TINY_RCU: Keep TINY_RCU tiny! */
+static void trace_rcu_invoke_kfree_callback(struct rcu_head *rhp,
+					    unsigned long offset)
+{
+}
+static void trace_rcu_invoke_callback(struct rcu_head *head)
+{
+}
+
+#endif /* #else #ifdef CONFIG_RCU_TRACE */
+
+#include "rcu.h"
+
 /* Controls for rcu_kthread() kthread, replacing RCU_SOFTIRQ used previously. */
 static struct task_struct *rcu_kthread_task;
 static DECLARE_WAIT_QUEUE_HEAD(rcu_kthread_wq);

@@ -161,11 +180,15 @@ static void rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 	RCU_TRACE(int cb_count = 0);
 
 	/* If no RCU callbacks ready to invoke, just return. */
-	if (&rcp->rcucblist == rcp->donetail)
+	if (&rcp->rcucblist == rcp->donetail) {
+		RCU_TRACE(trace_rcu_batch_start(0, -1));
+		RCU_TRACE(trace_rcu_batch_end(0));
 		return;
+	}
 
 	/* Move the ready-to-invoke callbacks to a local list. */
 	local_irq_save(flags);
+	RCU_TRACE(trace_rcu_batch_start(0, -1));
 	list = rcp->rcucblist;
 	rcp->rcucblist = *rcp->donetail;
 	*rcp->donetail = NULL;

@@ -187,6 +210,7 @@ static void rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 		RCU_TRACE(cb_count++);
 	}
 	RCU_TRACE(rcu_trace_sub_qlen(rcp, cb_count));
+	RCU_TRACE(trace_rcu_batch_end(cb_count));
 }
 
 /*

index a7c6bce..45dcc20 100644
@@ -52,6 +52,9 @@
 #include <linux/prefetch.h>
 
 #include "rcutree.h"
+#include <trace/events/rcu.h>
+
+#include "rcu.h"
 
 /* Data structures. */
 

@@ -1190,17 +1193,22 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 {
 	unsigned long flags;
 	struct rcu_head *next, *list, **tail;
-	int count;
+	int bl, count;
 
 	/* If no callbacks are ready, just return.*/
-	if (!cpu_has_callbacks_ready_to_invoke(rdp))
+	if (!cpu_has_callbacks_ready_to_invoke(rdp)) {
+		trace_rcu_batch_start(0, 0);
+		trace_rcu_batch_end(0);
 		return;
+	}
 
 	/*
 	 * Extract the list of ready callbacks, disabling to prevent
 	 * races with call_rcu() from interrupt handlers.
 	 */
 	local_irq_save(flags);
+	bl = rdp->blimit;
+	trace_rcu_batch_start(rdp->qlen, bl);
 	list = rdp->nxtlist;
 	rdp->nxtlist = *rdp->nxttail[RCU_DONE_TAIL];
 	*rdp->nxttail[RCU_DONE_TAIL] = NULL;

@@ -1218,11 +1226,12 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 		debug_rcu_head_unqueue(list);
 		__rcu_reclaim(list);
 		list = next;
-		if (++count >= rdp->blimit)
+		if (++count >= bl)
 			break;
 	}
 
 	local_irq_save(flags);
+	trace_rcu_batch_end(count);
 
 	/* Update count, and requeue any remaining callbacks. */
 	rdp->qlen -= count;

index db3f6e9..ab458eb 100644
@@ -7,29 +7,58 @@
 #include <linux/tracepoint.h>
 
 /*
- * Tracepoint for calling rcu_do_batch, performed to start callback invocation:
+ * Tracepoint for start/end markers used for utilization calculations.
+ * By convention, the string is of the following forms:
+ *
+ * "Start <activity>" -- Mark the start of the specified activity,
+ *			 such as "context switch".  Nesting is permitted.
+ * "End <activity>" -- Mark the end of the specified activity.
+ */
+TRACE_EVENT(rcu_utilization,
+
+	TP_PROTO(char *s),
+
+	TP_ARGS(s),
+
+	TP_STRUCT__entry(
+		__field(char *,	s)
+	),
+
+	TP_fast_assign(
+		__entry->s = s;
+	),
+
+	TP_printk("%s", __entry->s)
+);
+
+/*
+ * Tracepoint for marking the beginning rcu_do_batch, performed to start
+ * RCU callback invocation.  The first argument is the total number of
+ * callbacks (including those that are not yet ready to be invoked),
+ * and the second argument is the current RCU-callback batch limit.
  */
 TRACE_EVENT(rcu_batch_start,
 
-	TP_PROTO(long callbacks_ready, int blimit),
+	TP_PROTO(long qlen, int blimit),
 
-	TP_ARGS(callbacks_ready, blimit),
+	TP_ARGS(qlen, blimit),
 
 	TP_STRUCT__entry(
-		__field(	long,	callbacks_ready		)
-		__field(	int,	blimit			)
+		__field(long, qlen)
+		__field(int, blimit)
 	),
 
 	TP_fast_assign(
-		__entry->callbacks_ready	= callbacks_ready;
-		__entry->blimit			= blimit;
+		__entry->qlen = qlen;
+		__entry->blimit = blimit;
 	),
 
-	TP_printk("CBs=%ld bl=%d", __entry->callbacks_ready, __entry->blimit)
+	TP_printk("CBs=%ld bl=%d", __entry->qlen, __entry->blimit)
 );
 
 /*
- * Tracepoint for the invocation of a single RCU callback
+ * Tracepoint for the invocation of a single RCU callback function.
+ * The argument is a pointer to the RCU callback itself.
  */
 TRACE_EVENT(rcu_invoke_callback,
 

@@ -38,20 +67,23 @@ TRACE_EVENT(rcu_invoke_callback,
 	TP_ARGS(rhp),
 
 	TP_STRUCT__entry(
-		__field(	void *,	rhp	)
-		__field(	void *,	func	)
+		__field(void *,	rhp)
+		__field(void *,	func)
 	),
 
 	TP_fast_assign(
-		__entry->rhp	= rhp;
-		__entry->func	= rhp->func;
+		__entry->rhp = rhp;
+		__entry->func = rhp->func;
 	),
 
 	TP_printk("rhp=%p func=%pf", __entry->rhp, __entry->func)
 );
 
 /*
- * Tracepoint for the invocation of a single RCU kfree callback
+ * Tracepoint for the invocation of a single RCU callback of the special
+ * kfree() form.  The first argument is a pointer to the RCU callback
+ * and the second argument is the offset of the callback within the
+ * enclosing RCU-protected data structure.
  */
 TRACE_EVENT(rcu_invoke_kfree_callback,
 

@@ -60,12 +92,12 @@ TRACE_EVENT(rcu_invoke_kfree_callback,
 	TP_ARGS(rhp, offset),
 
 	TP_STRUCT__entry(
-		__field(void *,	rhp	)
-		__field(unsigned long,	offset	)
+		__field(void *,	rhp)
+		__field(unsigned long, offset)
 	),
 
 	TP_fast_assign(
-		__entry->rhp	= rhp;
+		__entry->rhp = rhp;
 		__entry->offset	= offset;
 	),
 

@@ -73,7 +105,8 @@ TRACE_EVENT(rcu_invoke_kfree_callback,
 );
 
 /*
- * Tracepoint for leaving rcu_do_batch, performed after callback invocation:
+ * Tracepoint for exiting rcu_do_batch after RCU callbacks have been
+ * invoked.  The first argument is the number of callbacks actually invoked.
  */
 TRACE_EVENT(rcu_batch_end,
 

@@ -82,11 +115,11 @@ TRACE_EVENT(rcu_batch_end,
 	TP_ARGS(callbacks_invoked),
 
 	TP_STRUCT__entry(
-		__field(	int,	callbacks_invoked		)
+		__field(int, callbacks_invoked)
 	),
 
 	TP_fast_assign(
-		__entry->callbacks_invoked	= callbacks_invoked;
+		__entry->callbacks_invoked = callbacks_invoked;
 	),
 
 	TP_printk("CBs-invoked=%d", __entry->callbacks_invoked)

index 45dcc20..2a9643b 100644
@@ -184,8 +184,10 @@ void rcu_bh_qs(int cpu)
  */
 void rcu_note_context_switch(int cpu)
 {
+	trace_rcu_utilization("Start context switch");
 	rcu_sched_qs(cpu);
 	rcu_preempt_note_context_switch(cpu);
+	trace_rcu_utilization("End context switch");
 }
 EXPORT_SYMBOL_GPL(rcu_note_context_switch);
 

@@ -1275,6 +1277,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
  */
 void rcu_check_callbacks(int cpu, int user)
 {
+	trace_rcu_utilization("Start scheduler-tick");
 	if (user ||
 	    (idle_cpu(cpu) && rcu_scheduler_active &&
 	     !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) {

@@ -1308,6 +1311,7 @@ void rcu_check_callbacks(int cpu, int user)
 	rcu_preempt_check_callbacks(cpu);
 	if (rcu_pending(cpu))
 		invoke_rcu_core();
+	trace_rcu_utilization("End scheduler-tick");
 }
 
 #ifdef CONFIG_SMP

@@ -1369,10 +1373,14 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
 	unsigned long flags;
 	struct rcu_node *rnp = rcu_get_root(rsp);
 
-	if (!rcu_gp_in_progress(rsp))
+	trace_rcu_utilization("Start fqs");
+	if (!rcu_gp_in_progress(rsp)) {
+		trace_rcu_utilization("End fqs");
 		return;  /* No grace period in progress, nothing to force. */
+	}
 	if (!raw_spin_trylock_irqsave(&rsp->fqslock, flags)) {
 		rsp->n_force_qs_lh++; /* Inexact, can lose counts.  Tough! */
+		trace_rcu_utilization("End fqs");
 		return;	/* Someone else is already on the job. */
 	}
 	if (relaxed && ULONG_CMP_GE(rsp->jiffies_force_qs, jiffies))

@@ -1421,11 +1429,13 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
 		raw_spin_unlock(&rsp->fqslock); /* irqs remain disabled */
 		rsp->fqs_need_gp = 0;
 		rcu_start_gp(rsp, flags); /* releases rnp->lock */
+		trace_rcu_utilization("End fqs");
 		return;
 	}
 	raw_spin_unlock(&rnp->lock);  /* irqs remain disabled */
 unlock_fqs_ret:
 	raw_spin_unlock_irqrestore(&rsp->fqslock, flags);
+	trace_rcu_utilization("End fqs");
 }
 
 #else /* #ifdef CONFIG_SMP */

@@ -1481,6 +1491,7 @@ __rcu_process_callbacks(struct rcu_state *rsp, struct rcu_data *rdp)
  */
 static void rcu_process_callbacks(struct softirq_action *unused)
 {
+	trace_rcu_utilization("Start RCU core");
 	__rcu_process_callbacks(&rcu_sched_state,
 				&__get_cpu_var(rcu_sched_data));
 	__rcu_process_callbacks(&rcu_bh_state, &__get_cpu_var(rcu_bh_data));

@@ -1488,6 +1499,7 @@ static void rcu_process_callbacks(struct softirq_action *unused)
 
 	/* If we are last CPU on way to dyntick-idle mode, accelerate it. */
 	rcu_needs_cpu_flush();
+	trace_rcu_utilization("End RCU core");
 }
 
 /*

@@ -1910,6 +1922,7 @@ static int __cpuinit rcu_cpu_notify(struct notifier_block *self,
 	struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, cpu);
 	struct rcu_node *rnp = rdp->mynode;
 
+	trace_rcu_utilization("Start CPU hotplug");
 	switch (action) {
 	case CPU_UP_PREPARE:
 	case CPU_UP_PREPARE_FROZEN:

@@ -1945,6 +1958,7 @@ static int __cpuinit rcu_cpu_notify(struct notifier_block *self,
 	default:
 		break;
 	}
+	trace_rcu_utilization("End CPU hotplug");
 	return NOTIFY_OK;
 }
 

index 7bc1643..d7f00ec 100644
@@ -23,6 +23,12 @@
 #ifndef __LINUX_RCU_H
 #define __LINUX_RCU_H
 
+#ifdef CONFIG_RCU_TRACE
+#define RCU_TRACE(stmt) stmt
+#else /* #ifdef CONFIG_RCU_TRACE */
+#define RCU_TRACE(stmt)
+#endif /* #else #ifdef CONFIG_RCU_TRACE */
+
 /*
  * debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
  * by call_rcu() and rcu callback execution, and are therefore not part of the

@@ -68,10 +74,10 @@ static inline void __rcu_reclaim(struct rcu_head *head)
 	unsigned long offset = (unsigned long)head->func;
 
 	if (__is_kfree_rcu_offset(offset)) {
-		trace_rcu_invoke_kfree_callback(head, offset);
+		RCU_TRACE(trace_rcu_invoke_kfree_callback(head, offset));
 		kfree((void *)head - offset);
 	} else {
-		trace_rcu_invoke_callback(head);
+		RCU_TRACE(trace_rcu_invoke_callback(head));
 		head->func(head);
 	}
 }

index 19453ba..0d28974 100644
@@ -38,20 +38,7 @@
 #include <linux/prefetch.h>
 
 #ifdef CONFIG_RCU_TRACE
-
 #include <trace/events/rcu.h>
-
-#else /* #ifdef CONFIG_RCU_TRACE */
-
-/* No by-default tracing in TINY_RCU: Keep TINY_RCU tiny! */
-static void trace_rcu_invoke_kfree_callback(struct rcu_head *rhp,
-					    unsigned long offset)
-{
-}
-static void trace_rcu_invoke_callback(struct rcu_head *head)
-{
-}
-
 #endif /* #else #ifdef CONFIG_RCU_TRACE */
 
 #include "rcu.h"

index 6b0cedb..791ddf7 100644
@@ -26,29 +26,26 @@
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
 
-#ifdef CONFIG_RCU_TRACE
-#define RCU_TRACE(stmt)	stmt
-#else /* #ifdef CONFIG_RCU_TRACE */
-#define RCU_TRACE(stmt)
-#endif /* #else #ifdef CONFIG_RCU_TRACE */
-
 /* Global control variables for rcupdate callback mechanism. */
 struct rcu_ctrlblk {
 	struct rcu_head *rcucblist;	/* List of pending callbacks (CBs). */
 	struct rcu_head **donetail;	/* ->next pointer of last "done" CB. */
 	struct rcu_head **curtail;	/* ->next pointer of last CB. */
 	RCU_TRACE(long qlen);		/* Number of pending CBs. */
+	RCU_TRACE(char *name);		/* Name of RCU type. */
 };
 
 /* Definition for rcupdate control block. */
 static struct rcu_ctrlblk rcu_sched_ctrlblk = {
 	.donetail	= &rcu_sched_ctrlblk.rcucblist,
 	.curtail	= &rcu_sched_ctrlblk.rcucblist,
+	RCU_TRACE(.name = "rcu_sched")
 };
 
 static struct rcu_ctrlblk rcu_bh_ctrlblk = {
 	.donetail	= &rcu_bh_ctrlblk.rcucblist,
 	.curtail	= &rcu_bh_ctrlblk.rcucblist,
+	RCU_TRACE(.name = "rcu_bh")
 };
 
 #ifdef CONFIG_DEBUG_LOCK_ALLOC

@@ -131,6 +128,7 @@ static struct rcu_preempt_ctrlblk rcu_preempt_ctrlblk = {
 	.rcb.curtail = &rcu_preempt_ctrlblk.rcb.rcucblist,
 	.nexttail = &rcu_preempt_ctrlblk.rcb.rcucblist,
 	.blkd_tasks = LIST_HEAD_INIT(rcu_preempt_ctrlblk.blkd_tasks),
+	RCU_TRACE(.rcb.name = "rcu_preempt")
 };
 
 static int rcu_preempted_readers_exp(void);

index 2a9643b..b953e2c 100644
@@ -61,7 +61,7 @@
 static struct lock_class_key rcu_node_class[NUM_RCU_LVLS];
 
 #define RCU_STATE_INITIALIZER(structname) { \
-	.level = { &structname.node[0] }, \
+	.level = { &structname##_state.node[0] }, \
 	.levelcnt = { \
 		NUM_RCU_LVL_0,  /* root of hierarchy. */ \
 		NUM_RCU_LVL_1, \

@@ -72,17 +72,17 @@ static struct lock_class_key rcu_node_class[NUM_RCU_LVLS];
 	.signaled = RCU_GP_IDLE, \
 	.gpnum = -300, \
 	.completed = -300, \
-	.onofflock = __RAW_SPIN_LOCK_UNLOCKED(&structname.onofflock), \
-	.fqslock = __RAW_SPIN_LOCK_UNLOCKED(&structname.fqslock), \
+	.onofflock = __RAW_SPIN_LOCK_UNLOCKED(&structname##_state.onofflock), \
+	.fqslock = __RAW_SPIN_LOCK_UNLOCKED(&structname##_state.fqslock), \
 	.n_force_qs = 0, \
 	.n_force_qs_ngp = 0, \
 	.name = #structname, \
 }
 
-struct rcu_state rcu_sched_state = RCU_STATE_INITIALIZER(rcu_sched_state);
+struct rcu_state rcu_sched_state = RCU_STATE_INITIALIZER(rcu_sched);
 DEFINE_PER_CPU(struct rcu_data, rcu_sched_data);
 
-struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state);
+struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh);
 DEFINE_PER_CPU(struct rcu_data, rcu_bh_data);
 
 static struct rcu_state *rcu_state;

index 43daa46..a90bf3c 100644
@@ -64,7 +64,7 @@ static void __init rcu_bootup_announce_oddness(void)
 
 #ifdef CONFIG_TREE_PREEMPT_RCU
 
-struct rcu_state rcu_preempt_state = RCU_STATE_INITIALIZER(rcu_preempt_state);
+struct rcu_state rcu_preempt_state = RCU_STATE_INITIALIZER(rcu_preempt);
 DEFINE_PER_CPU(struct rcu_data, rcu_preempt_data);
 static struct rcu_state *rcu_state = &rcu_preempt_state;
 

index ab458eb..508824e5 100644
@@ -33,27 +33,31 @@ TRACE_EVENT(rcu_utilization,
 
 /*
  * Tracepoint for marking the beginning rcu_do_batch, performed to start
- * RCU callback invocation.  The first argument is the total number of
- * callbacks (including those that are not yet ready to be invoked),
- * and the second argument is the current RCU-callback batch limit.
+ * RCU callback invocation.  The first argument is the RCU flavor,
+ * the second is the total number of callbacks (including those that
+ * are not yet ready to be invoked), and the third argument is the
+ * current RCU-callback batch limit.
  */
 TRACE_EVENT(rcu_batch_start,
 
-	TP_PROTO(long qlen, int blimit),
+	TP_PROTO(char *rcuname, long qlen, int blimit),
 
-	TP_ARGS(qlen, blimit),
+	TP_ARGS(rcuname, qlen, blimit),
 
 	TP_STRUCT__entry(
+		__field(char *, rcuname)
 		__field(long, qlen)
 		__field(int, blimit)
 	),
 
 	TP_fast_assign(
+		__entry->rcuname = rcuname;
 		__entry->qlen = qlen;
 		__entry->blimit = blimit;
 	),
 
-	TP_printk("CBs=%ld bl=%d", __entry->qlen, __entry->blimit)
+	TP_printk("%s CBs=%ld bl=%d",
+		  __entry->rcuname, __entry->qlen, __entry->blimit)
 );
 
 /*

@@ -106,23 +110,27 @@ TRACE_EVENT(rcu_invoke_kfree_callback,
 
 /*
  * Tracepoint for exiting rcu_do_batch after RCU callbacks have been
- * invoked.  The first argument is the number of callbacks actually invoked.
+ * invoked.  The first argument is the name of the RCU flavor and
+ * the second argument is number of callbacks actually invoked.
  */
 TRACE_EVENT(rcu_batch_end,
 
-	TP_PROTO(int callbacks_invoked),
+	TP_PROTO(char *rcuname, int callbacks_invoked),
 
-	TP_ARGS(callbacks_invoked),
+	TP_ARGS(rcuname, callbacks_invoked),
 
 	TP_STRUCT__entry(
+		__field(char *, rcuname)
 		__field(int, callbacks_invoked)
 	),
 
 	TP_fast_assign(
+		__entry->rcuname = rcuname;
 		__entry->callbacks_invoked = callbacks_invoked;
 	),
 
-	TP_printk("CBs-invoked=%d", __entry->callbacks_invoked)
+	TP_printk("%s CBs-invoked=%d",
+		  __entry->rcuname, __entry->callbacks_invoked)
 );
 
 #endif /* _TRACE_RCU_H */

index 0d28974..1c37bdd 100644
@@ -168,14 +168,14 @@ static void rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 
 	/* If no RCU callbacks ready to invoke, just return. */
 	if (&rcp->rcucblist == rcp->donetail) {
-		RCU_TRACE(trace_rcu_batch_start(0, -1));
-		RCU_TRACE(trace_rcu_batch_end(0));
+		RCU_TRACE(trace_rcu_batch_start(rcp->name, 0, -1));
+		RCU_TRACE(trace_rcu_batch_end(rcp->name, 0));
 		return;
 	}
 
 	/* Move the ready-to-invoke callbacks to a local list. */
 	local_irq_save(flags);
-	RCU_TRACE(trace_rcu_batch_start(0, -1));
+	RCU_TRACE(trace_rcu_batch_start(rcp->name, 0, -1));
 	list = rcp->rcucblist;
 	rcp->rcucblist = *rcp->donetail;
 	*rcp->donetail = NULL;

@@ -197,7 +197,7 @@ static void rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 		RCU_TRACE(cb_count++);
 	}
 	RCU_TRACE(rcu_trace_sub_qlen(rcp, cb_count));
-	RCU_TRACE(trace_rcu_batch_end(cb_count));
+	RCU_TRACE(trace_rcu_batch_end(rcp->name, cb_count));
 }
 
 /*

index b953e2c..eb6e731 100644
@@ -1199,8 +1199,8 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 
 	/* If no callbacks are ready, just return.*/
 	if (!cpu_has_callbacks_ready_to_invoke(rdp)) {
-		trace_rcu_batch_start(0, 0);
-		trace_rcu_batch_end(0);
+		trace_rcu_batch_start(rsp->name, 0, 0);
+		trace_rcu_batch_end(rsp->name, 0);
 		return;
 	}
 

@@ -1210,7 +1210,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 	 */
 	local_irq_save(flags);
 	bl = rdp->blimit;
-	trace_rcu_batch_start(rdp->qlen, bl);
+	trace_rcu_batch_start(rsp->name, rdp->qlen, bl);
 	list = rdp->nxtlist;
 	rdp->nxtlist = *rdp->nxttail[RCU_DONE_TAIL];
 	*rdp->nxttail[RCU_DONE_TAIL] = NULL;

@@ -1233,7 +1233,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 	}
 
 	local_irq_save(flags);
-	trace_rcu_batch_end(count);
+	trace_rcu_batch_end(rsp->name, count);
 
 	/* Update count, and requeue any remaining callbacks. */
 	rdp->qlen -= count;

index eb6e731..4e24399 100644
@@ -198,7 +198,7 @@ DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = {
 };
 #endif /* #ifdef CONFIG_NO_HZ */
 
-static int blimit = 10;		/* Maximum callbacks per softirq. */
+static int blimit = 10;		/* Maximum callbacks per rcu_do_batch. */
 static int qhimark = 10000;	/* If this many pending, ignore blimit. */
 static int qlowmark = 100;	/* Once only this many pending, use blimit. */
 

@@ -1261,7 +1261,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 
 	local_irq_restore(flags);
 
-	/* Re-raise the RCU softirq if there are callbacks remaining. */
+	/* Re-invoke RCU core processing if there are callbacks remaining. */
 	if (cpu_has_callbacks_ready_to_invoke(rdp))
 		invoke_rcu_core();
 }

@@ -1269,7 +1269,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp)
 /*
  * Check to see if this CPU is in a non-context-switch quiescent state
  * (user mode or idle loop for rcu, non-softirq execution for rcu_bh).
- * Also schedule the RCU softirq handler.
+ * Also schedule RCU core processing.
  *
  * This function must be called with hardirqs disabled.  It is normally
  * invoked from the scheduling-clock interrupt.  If rcu_pending returns

@@ -1448,9 +1448,9 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
 #endif /* #else #ifdef CONFIG_SMP */
 
 /*
- * This does the RCU processing work from softirq context for the
- * specified rcu_state and rcu_data structures.  This may be called
- * only from the CPU to whom the rdp belongs.
+ * This does the RCU core processing work for the specified rcu_state
+ * and rcu_data structures.  This may be called only from the CPU to
+ * whom the rdp belongs.
  */
 static void
 __rcu_process_callbacks(struct rcu_state *rsp, struct rcu_data *rdp)

@@ -1487,7 +1487,7 @@ __rcu_process_callbacks(struct rcu_state *rsp, struct rcu_data *rdp)
 }
 
 /*
- * Do softirq processing for the current CPU.
+ * Do RCU core processing for the current CPU.
  */
 static void rcu_process_callbacks(struct softirq_action *unused)
 {

@@ -1503,10 +1503,11 @@ static void rcu_process_callbacks(struct softirq_action *unused)
 }
 
 /*
- * Wake up the current CPU's kthread.  This replaces raise_softirq()
- * in earlier versions of RCU.  Note that because we are running on
- * the current CPU with interrupts disabled, the rcu_cpu_kthread_task
- * cannot disappear out from under us.
+ * Schedule RCU callback invocation.  If the specified type of RCU
+ * does not support RCU priority boosting, just do a direct call,
+ * otherwise wake up the per-CPU kernel kthread.  Note that because we
+ * are running on the current CPU with interrupts disabled, the
+ * rcu_cpu_kthread_task cannot disappear out from under us.
  */
 static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp)
 {

index a90bf3c..ecd48a2 100644
@@ -1478,7 +1478,8 @@ static int rcu_cpu_kthread_should_stop(int cpu)
 
 /*
  * Per-CPU kernel thread that invokes RCU callbacks.  This replaces the
- * earlier RCU softirq.
+ * RCU softirq used in flavors and configurations of RCU that do not
+ * support RCU priority boosting.
  */
 static int rcu_cpu_kthread(void *arg)
 {

index 01b2ccd..eee6c94 100644
@@ -417,6 +417,13 @@ extern struct rcu_state rcu_preempt_state;
 DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data);
 #endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */
 
+#ifdef CONFIG_RCU_BOOST
+DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status);
+DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu);
+DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops);
+DECLARE_PER_CPU(char, rcu_cpu_has_work);
+#endif /* #ifdef CONFIG_RCU_BOOST */
+
 #ifndef RCU_TREE_NONCORE
 
 /* Forward declarations for rcutree_plugin.h */

index 8827b34..e623564 100644
@@ -48,11 +48,6 @@
 
 #ifdef CONFIG_RCU_BOOST
 
-DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status);
-DECLARE_PER_CPU(int, rcu_cpu_kthread_cpu);
-DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops);
-DECLARE_PER_CPU(char, rcu_cpu_has_work);
-
 static char convert_kthread_status(unsigned int kthread_status)
 {
 	if (kthread_status > RCU_KTHREAD_MAX)

index 508824e5..ac52aba 100644
@@ -13,6 +13,9 @@
  * "Start <activity>" -- Mark the start of the specified activity,
  *			 such as "context switch".  Nesting is permitted.
  * "End <activity>" -- Mark the end of the specified activity.
+ *
+ * An "@" character within "<activity>" is a comment character: Data
+ * reduction scripts will ignore the "@" and the remainder of the line.
  */
 TRACE_EVENT(rcu_utilization,
 

index ecd48a2..94d9ca1 100644
@@ -1219,9 +1219,12 @@ static int rcu_boost_kthread(void *arg)
 	int spincnt = 0;
 	int more2boost;
 
+	trace_rcu_utilization("Start boost kthread@init");
 	for (;;) {
 		rnp->boost_kthread_status = RCU_KTHREAD_WAITING;
+		trace_rcu_utilization("End boost kthread@rcu_wait");
 		rcu_wait(rnp->boost_tasks || rnp->exp_tasks);
+		trace_rcu_utilization("Start boost kthread@rcu_wait");
 		rnp->boost_kthread_status = RCU_KTHREAD_RUNNING;
 		more2boost = rcu_boost(rnp);
 		if (more2boost)

@@ -1229,11 +1232,14 @@ static int rcu_boost_kthread(void *arg)
 		else
 			spincnt = 0;
 		if (spincnt > 10) {
+			trace_rcu_utilization("End boost kthread@rcu_yield");
 			rcu_yield(rcu_boost_kthread_timer, (unsigned long)rnp);
+			trace_rcu_utilization("Start boost kthread@rcu_yield");
 			spincnt = 0;
 		}
 	}
 	/* NOTREACHED */
+	trace_rcu_utilization("End boost kthread@notreached");
 	return 0;
 }
 

@@ -1490,9 +1496,12 @@ static int rcu_cpu_kthread(void *arg)
 	char work;
 	char *workp = &per_cpu(rcu_cpu_has_work, cpu);
 
+	trace_rcu_utilization("Start CPU kthread@init");
 	for (;;) {
 		*statusp = RCU_KTHREAD_WAITING;
+		trace_rcu_utilization("End CPU kthread@rcu_wait");
 		rcu_wait(*workp != 0 || kthread_should_stop());
+		trace_rcu_utilization("Start CPU kthread@rcu_wait");
 		local_bh_disable();
 		if (rcu_cpu_kthread_should_stop(cpu)) {
 			local_bh_enable();

@@ -1513,11 +1522,14 @@ static int rcu_cpu_kthread(void *arg)
 			spincnt = 0;
 		if (spincnt > 10) {
 			*statusp = RCU_KTHREAD_YIELDING;
+			trace_rcu_utilization("End CPU kthread@rcu_yield");
 			rcu_yield(rcu_cpu_kthread_timer, (unsigned long)cpu);
+			trace_rcu_utilization("Start CPU kthread@rcu_yield");
 			spincnt = 0;
 		}
 	}
 	*statusp = RCU_KTHREAD_STOPPED;
+	trace_rcu_utilization("End CPU kthread@term");
 	return 0;
 }
 

index 4eab233..00b7a5e 100644
@@ -27,9 +27,13 @@
 
 #include <linux/cache.h>
 
+#ifdef CONFIG_RCU_BOOST
 static inline void rcu_init(void)
 {
 }
+#else /* #ifdef CONFIG_RCU_BOOST */
+void rcu_init(void);
+#endif /* #else #ifdef CONFIG_RCU_BOOST */
 
 static inline void rcu_barrier_bh(void)
 {

116 lines not on this page · raw message

164 lines not on this page · raw message

index ac52aba..669fbd6 100644
@@ -24,7 +24,7 @@ TRACE_EVENT(rcu_utilization,
 	TP_ARGS(s),
 
 	TP_STRUCT__entry(
-		__field(char *,	s)
+		__field(char *, s)
 	),
 
 	TP_fast_assign(

@@ -34,6 +34,297 @@ TRACE_EVENT(rcu_utilization,
 	TP_printk("%s", __entry->s)
 );
 
+#ifdef CONFIG_RCU_TRACE
+
+#if defined(CONFIG_TREE_RCU) || defined(CONFIG_TREE_PREEMPT_RCU)
+
+/*
+ * Tracepoint for grace-period events: starting and ending a grace
+ * period ("start" and "end", respectively), a CPU noting the start
+ * of a new grace period or the end of an old grace period ("cpustart"
+ * and "cpuend", respectively), a CPU passing through a quiescent
+ * state ("cpuqs"), a CPU coming online or going offline ("cpuonl"
+ * and "cpuofl", respectively), and a CPU being kicked for being too
+ * long in dyntick-idle mode ("kick").
+ */
+TRACE_EVENT(rcu_grace_period,
+
+	TP_PROTO(char *rcuname, unsigned long gpnum, char *gpevent),
+
+	TP_ARGS(rcuname, gpnum, gpevent),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(char *, gpevent)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->gpevent = gpevent;
+	),
+
+	TP_printk("%s %lu %s",
+		  __entry->rcuname, __entry->gpnum, __entry->gpevent)
+);
+
+/*
+ * Tracepoint for grace-period-initialization events.  These are
+ * distinguished by the type of RCU, the new grace-period number, the
+ * rcu_node structure level, the starting and ending CPU covered by the
+ * rcu_node structure, and the mask of CPUs that will be waited for.
+ * All but the type of RCU are extracted from the rcu_node structure.
+ */
+TRACE_EVENT(rcu_grace_period_init,
+
+	TP_PROTO(char *rcuname, unsigned long gpnum, u8 level,
+		 int grplo, int grphi, unsigned long qsmask),
+
+	TP_ARGS(rcuname, gpnum, level, grplo, grphi, qsmask),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(u8, level)
+		__field(int, grplo)
+		__field(int, grphi)
+		__field(unsigned long, qsmask)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->level = level;
+		__entry->grplo = grplo;
+		__entry->grphi = grphi;
+		__entry->qsmask = qsmask;
+	),
+
+	TP_printk("%s %lu %u %d %d %lx",
+		  __entry->rcuname, __entry->gpnum, __entry->level,
+		  __entry->grplo, __entry->grphi, __entry->qsmask)
+);
+
+/*
+ * Tracepoint for tasks blocking within preemptible-RCU read-side
+ * critical sections.  Track the type of RCU (which one day might
+ * include SRCU), the grace-period number that the task is blocking
+ * (the current or the next), and the task's PID.
+ */
+TRACE_EVENT(rcu_preempt_task,
+
+	TP_PROTO(char *rcuname, int pid, unsigned long gpnum),
+
+	TP_ARGS(rcuname, pid, gpnum),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(int, pid)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->pid = pid;
+	),
+
+	TP_printk("%s %lu %d",
+		  __entry->rcuname, __entry->gpnum, __entry->pid)
+);
+
+/*
+ * Tracepoint for tasks that blocked within a given preemptible-RCU
+ * read-side critical section exiting that critical section.  Track the
+ * type of RCU (which one day might include SRCU) and the task's PID.
+ */
+TRACE_EVENT(rcu_unlock_preempted_task,
+
+	TP_PROTO(char *rcuname, unsigned long gpnum, int pid),
+
+	TP_ARGS(rcuname, gpnum, pid),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(int, pid)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->pid = pid;
+	),
+
+	TP_printk("%s %lu %d", __entry->rcuname, __entry->gpnum, __entry->pid)
+);
+
+/*
+ * Tracepoint for quiescent-state-reporting events.  These are
+ * distinguished by the type of RCU, the grace-period number, the
+ * mask of quiescent lower-level entities, the rcu_node structure level,
+ * the starting and ending CPU covered by the rcu_node structure, and
+ * whether there are any blocked tasks blocking the current grace period.
+ * All but the type of RCU are extracted from the rcu_node structure.
+ */
+TRACE_EVENT(rcu_quiescent_state_report,
+
+	TP_PROTO(char *rcuname, unsigned long gpnum,
+		 unsigned long mask, unsigned long qsmask,
+		 u8 level, int grplo, int grphi, int gp_tasks),
+
+	TP_ARGS(rcuname, gpnum, mask, qsmask, level, grplo, grphi, gp_tasks),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(unsigned long, mask)
+		__field(unsigned long, qsmask)
+		__field(u8, level)
+		__field(int, grplo)
+		__field(int, grphi)
+		__field(u8, gp_tasks)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->mask = mask;
+		__entry->qsmask = qsmask;
+		__entry->level = level;
+		__entry->grplo = grplo;
+		__entry->grphi = grphi;
+		__entry->gp_tasks = gp_tasks;
+	),
+
+	TP_printk("%s %lu %lx>%lx %u %d %d %u",
+		  __entry->rcuname, __entry->gpnum,
+		  __entry->mask, __entry->qsmask, __entry->level,
+		  __entry->grplo, __entry->grphi, __entry->gp_tasks)
+);
+
+/*
+ * Tracepoint for quiescent states detected by force_quiescent_state().
+ * These trace events include the type of RCU, the grace-period number
+ * that was blocked by the CPU, the CPU itself, and the type of quiescent
+ * state, which can be "dti" for dyntick-idle mode, "ofl" for CPU offline,
+ * or "kick" when kicking a CPU that has been in dyntick-idle mode for
+ * too long.
+ */
+TRACE_EVENT(rcu_fqs,
+
+	TP_PROTO(char *rcuname, unsigned long gpnum, int cpu, char *qsevent),
+
+	TP_ARGS(rcuname, gpnum, cpu, qsevent),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(unsigned long, gpnum)
+		__field(int, cpu)
+		__field(char *, qsevent)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpnum = gpnum;
+		__entry->cpu = cpu;
+		__entry->qsevent = qsevent;
+	),
+
+	TP_printk("%s %lu %d %s",
+		  __entry->rcuname, __entry->gpnum,
+		  __entry->cpu, __entry->qsevent)
+);
+
+#endif /* #if defined(CONFIG_TREE_RCU) || defined(CONFIG_TREE_PREEMPT_RCU) */
+
+/*
+ * Tracepoint for dyntick-idle entry/exit events.  These take a string
+ * as argument: "Start" for entering dyntick-idle mode and "End" for
+ * leaving it.
+ */
+TRACE_EVENT(rcu_dyntick,
+
+	TP_PROTO(char *polarity),
+
+	TP_ARGS(polarity),
+
+	TP_STRUCT__entry(
+		__field(char *, polarity)
+	),
+
+	TP_fast_assign(
+		__entry->polarity = polarity;
+	),
+
+	TP_printk("%s", __entry->polarity)
+);
+
+/*
+ * Tracepoint for the registration of a single RCU callback function.
+ * The first argument is the type of RCU, the second argument is
+ * a pointer to the RCU callback itself, and the third element is the
+ * new RCU callback queue length for the current CPU.
+ */
+TRACE_EVENT(rcu_callback,
+
+	TP_PROTO(char *rcuname, struct rcu_head *rhp, long qlen),
+
+	TP_ARGS(rcuname, rhp, qlen),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(void *, rhp)
+		__field(void *, func)
+		__field(long, qlen)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->rhp = rhp;
+		__entry->func = rhp->func;
+		__entry->qlen = qlen;
+	),
+
+	TP_printk("%s rhp=%p func=%pf %ld",
+		  __entry->rcuname, __entry->rhp, __entry->func, __entry->qlen)
+);
+
+/*
+ * Tracepoint for the registration of a single RCU callback of the special
+ * kfree() form.  The first argument is the RCU type, the second argument
+ * is a pointer to the RCU callback, the third argument is the offset
+ * of the callback within the enclosing RCU-protected data structure,
+ * and the fourth argument is the new RCU callback queue length for the
+ * current CPU.
+ */
+TRACE_EVENT(rcu_kfree_callback,
+
+	TP_PROTO(char *rcuname, struct rcu_head *rhp, unsigned long offset,
+		 long qlen),
+
+	TP_ARGS(rcuname, rhp, offset, qlen),
+
+	TP_STRUCT__entry(
+		__field(char *, rcuname)
+		__field(void *, rhp)
+		__field(unsigned long, offset)
+		__field(long, qlen)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->rhp = rhp;
+		__entry->offset = offset;
+		__entry->qlen = qlen;
+	),
+
+	TP_printk("%s rhp=%p func=%ld %ld",
+		  __entry->rcuname, __entry->rhp, __entry->offset,
+		  __entry->qlen)
+);
+
 /*
  * Tracepoint for marking the beginning rcu_do_batch, performed to start
  * RCU callback invocation.  The first argument is the RCU flavor,

@@ -65,50 +356,58 @@ TRACE_EVENT(rcu_batch_start,
 
 /*
  * Tracepoint for the invocation of a single RCU callback function.
- * The argument is a pointer to the RCU callback itself.
+ * The first argument is the type of RCU, and the second argument is
+ * a pointer to the RCU callback itself.
  */
 TRACE_EVENT(rcu_invoke_callback,
 
-	TP_PROTO(struct rcu_head *rhp),
+	TP_PROTO(char *rcuname, struct rcu_head *rhp),
 
-	TP_ARGS(rhp),
+	TP_ARGS(rcuname, rhp),
 
 	TP_STRUCT__entry(
-		__field(void *,	rhp)
-		__field(void *,	func)
+		__field(char *, rcuname)
+		__field(void *, rhp)
+		__field(void *, func)
 	),
 
 	TP_fast_assign(
+		__entry->rcuname = rcuname;
 		__entry->rhp = rhp;
 		__entry->func = rhp->func;
 	),
 
-	TP_printk("rhp=%p func=%pf", __entry->rhp, __entry->func)
+	TP_printk("%s rhp=%p func=%pf",
+		  __entry->rcuname, __entry->rhp, __entry->func)
 );
 
 /*
  * Tracepoint for the invocation of a single RCU callback of the special
- * kfree() form.  The first argument is a pointer to the RCU callback
- * and the second argument is the offset of the callback within the
- * enclosing RCU-protected data structure.
+ * kfree() form.  The first argument is the RCU flavor, the second
+ * argument is a pointer to the RCU callback, and the third argument
+ * is the offset of the callback within the enclosing RCU-protected
+ * data structure.
  */
 TRACE_EVENT(rcu_invoke_kfree_callback,
 
-	TP_PROTO(struct rcu_head *rhp, unsigned long offset),
+	TP_PROTO(char *rcuname, struct rcu_head *rhp, unsigned long offset),
 
-	TP_ARGS(rhp, offset),
+	TP_ARGS(rcuname, rhp, offset),
 
 	TP_STRUCT__entry(
-		__field(void *,	rhp)
+		__field(char *, rcuname)
+		__field(void *, rhp)
 		__field(unsigned long, offset)
 	),
 
 	TP_fast_assign(
+		__entry->rcuname = rcuname;
 		__entry->rhp = rhp;
 		__entry->offset	= offset;
 	),
 
-	TP_printk("rhp=%p func=%ld", __entry->rhp, __entry->offset)
+	TP_printk("%s rhp=%p func=%ld",
+		  __entry->rcuname, __entry->rhp, __entry->offset)
 );
 
 /*

@@ -136,6 +435,24 @@ TRACE_EVENT(rcu_batch_end,
 		  __entry->rcuname, __entry->callbacks_invoked)
 );
 
+#else /* #ifdef CONFIG_RCU_TRACE */
+
+#define trace_rcu_grace_period(rcuname, gpnum, gpevent) do { } while (0)
+#define trace_rcu_grace_period_init(rcuname, gpnum, level, grplo, grphi, qsmask) do { } while (0)
+#define trace_rcu_preempt_task(rcuname, pid, gpnum) do { } while (0)
+#define trace_rcu_unlock_preempted_task(rcuname, gpnum, pid) do { } while (0)
+#define trace_rcu_quiescent_state_report(rcuname, gpnum, mask, qsmask, level, grplo, grphi, gp_tasks) do { } while (0)
+#define trace_rcu_fqs(rcuname, gpnum, cpu, qsevent) do { } while (0)
+#define trace_rcu_dyntick(polarity) do { } while (0)
+#define trace_rcu_callback(rcuname, rhp, qlen) do { } while (0)
+#define trace_rcu_kfree_callback(rcuname, rhp, offset, qlen) do { } while (0)
+#define trace_rcu_batch_start(rcuname, qlen, blimit) do { } while (0)
+#define trace_rcu_invoke_callback(rcuname, rhp) do { } while (0)
+#define trace_rcu_invoke_kfree_callback(rcuname, rhp, offset) do { } while (0)
+#define trace_rcu_batch_end(rcuname, callbacks_invoked) do { } while (0)
+
+#endif /* #else #ifdef CONFIG_RCU_TRACE */
+
 #endif /* _TRACE_RCU_H */
 
 /* This part must be outside protection */

index d7f00ec..f600868 100644
@@ -69,15 +69,15 @@ static inline void debug_rcu_head_unqueue(struct rcu_head *head)
 
 extern void kfree(const void *);
 
-static inline void __rcu_reclaim(struct rcu_head *head)
+static inline void __rcu_reclaim(char *rn, struct rcu_head *head)
 {
 	unsigned long offset = (unsigned long)head->func;
 
 	if (__is_kfree_rcu_offset(offset)) {
-		RCU_TRACE(trace_rcu_invoke_kfree_callback(head, offset));
+		RCU_TRACE(trace_rcu_invoke_kfree_callback(rn, head, offset));
 		kfree((void *)head - offset);
 	} else {
-		RCU_TRACE(trace_rcu_invoke_callback(head));
+		RCU_TRACE(trace_rcu_invoke_callback(rn, head));
 		head->func(head);
 	}
 }

index c9321d8..da775c8 100644
@@ -147,6 +147,7 @@ void rcu_check_callbacks(int cpu, int user)
  */
 static void __rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 {
+	char *rn = NULL;
 	struct rcu_head *next, *list;
 	unsigned long flags;
 	RCU_TRACE(int cb_count = 0);

@@ -171,12 +172,13 @@ static void __rcu_process_callbacks(struct rcu_ctrlblk *rcp)
 	local_irq_restore(flags);
 
 	/* Invoke the callbacks on the local list. */
+	RCU_TRACE(rn = rcp->name);
 	while (list) {
 		next = list->next;
 		prefetch(next);
 		debug_rcu_head_unqueue(list);
 		local_bh_disable();
-		__rcu_reclaim(list);
+		__rcu_reclaim(rn, list);
 		local_bh_enable();
 		list = next;
 		RCU_TRACE(cb_count++);

156 lines not on this page · raw message

index eee6c94..d11a006 100644
@@ -299,6 +299,7 @@ struct rcu_data {
 	unsigned long n_rp_need_nothing;
 
 	int cpu;
+	struct rcu_state *rsp;
 };
 
 /* Values for signaled field in struct rcu_state. */

index 94d9ca1..bdb2e82 100644
@@ -124,6 +124,8 @@ static void rcu_preempt_qs(int cpu)
 
 	rdp->passed_quiesc_completed = rdp->gpnum - 1;
 	barrier();
+	if (rdp->passed_quiesc == 0)
+		trace_rcu_grace_period("rcu_preempt", rdp->gpnum, "cpuqs");
 	rdp->passed_quiesc = 1;
 	current->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_NEED_QS;
 }

@@ -190,6 +192,11 @@ static void rcu_preempt_note_context_switch(int cpu)
 			if (rnp->qsmask & rdp->grpmask)
 				rnp->gp_tasks = &t->rcu_node_entry;
 		}
+		trace_rcu_preempt_task(rdp->rsp->name,
+				       t->pid,
+				       (rnp->qsmask & rdp->grpmask)
+				       ? rnp->gpnum
+				       : rnp->gpnum + 1);
 		raw_spin_unlock_irqrestore(&rnp->lock, flags);
 	} else if (t->rcu_read_lock_nesting < 0 &&
 		   t->rcu_read_unlock_special) {

@@ -344,6 +351,8 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 		smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
 		np = rcu_next_node_entry(t, rnp);
 		list_del_init(&t->rcu_node_entry);
+		trace_rcu_unlock_preempted_task("rcu_preempt",
+						rnp->gpnum, t->pid);
 		if (&t->rcu_node_entry == rnp->gp_tasks)
 			rnp->gp_tasks = np;
 		if (&t->rcu_node_entry == rnp->exp_tasks)

@@ -364,10 +373,17 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 		 * we aren't waiting on any CPUs, report the quiescent state.
 		 * Note that rcu_report_unblock_qs_rnp() releases rnp->lock.
 		 */
-		if (empty)
-			raw_spin_unlock_irqrestore(&rnp->lock, flags);
-		else
+		if (!empty && !rcu_preempt_blocked_readers_cgp(rnp)) {
+			trace_rcu_quiescent_state_report("preempt_rcu",
+							 rnp->gpnum,
+							 0, rnp->qsmask,
+							 rnp->level,
+							 rnp->grplo,
+							 rnp->grphi,
+							 !!rnp->gp_tasks);
 			rcu_report_unblock_qs_rnp(rnp, flags);
+		} else
+			raw_spin_unlock_irqrestore(&rnp->lock, flags);
 
 #ifdef CONFIG_RCU_BOOST
 		/* Unboost if we were boosted. */

127 lines not on this page · raw message

index d11a006..51638b6 100644
@@ -230,9 +230,9 @@ struct rcu_data {
 					/*  in order to detect GP end. */
 	unsigned long	gpnum;		/* Highest gp number that this CPU */
 					/*  is aware of having started. */
-	unsigned long	passed_quiesc_completed;
-					/* Value of completed at time of qs. */
-	bool		passed_quiesc;	/* User-mode/idle loop etc. */
+	unsigned long	passed_quiesce_gpnum;
+					/* gpnum at time of quiescent state. */
+	bool		passed_quiesce;	/* User-mode/idle loop etc. */
 	bool		qs_pending;	/* Core waits for quiesc state. */
 	bool		beenonline;	/* CPU online at least once. */
 	bool		preemptible;	/* Preemptible RCU? */

index bdb2e82..4bac5a2 100644
@@ -122,11 +122,11 @@ static void rcu_preempt_qs(int cpu)
 {
 	struct rcu_data *rdp = &per_cpu(rcu_preempt_data, cpu);
 
-	rdp->passed_quiesc_completed = rdp->gpnum - 1;
+	rdp->passed_quiesce_gpnum = rdp->gpnum;
 	barrier();
-	if (rdp->passed_quiesc == 0)
+	if (rdp->passed_quiesce == 0)
 		trace_rcu_grace_period("rcu_preempt", rdp->gpnum, "cpuqs");
-	rdp->passed_quiesc = 1;
+	rdp->passed_quiesce = 1;
 	current->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_NEED_QS;
 }
 

index e623564..3f739cf3 100644
@@ -61,11 +61,11 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp)
 {
 	if (!rdp->beenonline)
 		return;
-	seq_printf(m, "%3d%cc=%lu g=%lu pq=%d pqc=%lu qp=%d",
+	seq_printf(m, "%3d%cc=%lu g=%lu pq=%d pgp=%lu qp=%d",
 		   rdp->cpu,
 		   cpu_is_offline(rdp->cpu) ? '!' : ' ',
 		   rdp->completed, rdp->gpnum,
-		   rdp->passed_quiesc, rdp->passed_quiesc_completed,
+		   rdp->passed_quiesce, rdp->passed_quiesce_gpnum,
 		   rdp->qs_pending);
 #ifdef CONFIG_NO_HZ
 	seq_printf(m, " dt=%d/%d/%d df=%lu",

@@ -139,7 +139,7 @@ static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp)
 		   rdp->cpu,
 		   cpu_is_offline(rdp->cpu) ? "\"N\"" : "\"Y\"",
 		   rdp->completed, rdp->gpnum,
-		   rdp->passed_quiesc, rdp->passed_quiesc_completed,
+		   rdp->passed_quiesce, rdp->passed_quiesce_gpnum,
 		   rdp->qs_pending);
 #ifdef CONFIG_NO_HZ
 	seq_printf(m, ",%d,%d,%d,%lu",

@@ -170,7 +170,7 @@ static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp)
 
 static int show_rcudata_csv(struct seq_file *m, void *unused)
 {
-	seq_puts(m, "\"CPU\",\"Online?\",\"c\",\"g\",\"pq\",\"pqc\",\"pq\",");
+	seq_puts(m, "\"CPU\",\"Online?\",\"c\",\"g\",\"pq\",\"pgp\",\"pq\",");
 #ifdef CONFIG_NO_HZ
 	seq_puts(m, "\"dt\",\"dt nesting\",\"dt NMI nesting\",\"df\",");
 #endif /* #ifdef CONFIG_NO_HZ */

index c61a535..ea80396 100644
@@ -578,26 +578,6 @@ extern int rcu_my_thread_group_empty(void);
 #define rcu_dereference_protected(p, c) \
 	__rcu_dereference_protected((p), (c), __rcu)
 
-/**
- * rcu_dereference_bh_protected() - fetch RCU-bh pointer when updates prevented
- * @p: The pointer to read, prior to dereferencing
- * @c: The conditions under which the dereference will take place
- *
- * This is the RCU-bh counterpart to rcu_dereference_protected().
- */
-#define rcu_dereference_bh_protected(p, c) \
-	__rcu_dereference_protected((p), (c), __rcu)
-
-/**
- * rcu_dereference_sched_protected() - fetch RCU-sched pointer when updates prevented
- * @p: The pointer to read, prior to dereferencing
- * @c: The conditions under which the dereference will take place
- *
- * This is the RCU-sched counterpart to rcu_dereference_protected().
- */
-#define rcu_dereference_sched_protected(p, c) \
-	__rcu_dereference_protected((p), (c), __rcu)
-
 
 /**
  * rcu_dereference() - fetch RCU-protected pointer for dereferencing

index 1194d6c..6648e00 100644
@@ -73,7 +73,7 @@ module_param(nreaders, int, 0444);
 MODULE_PARM_DESC(nreaders, "Number of RCU reader threads");
 module_param(nfakewriters, int, 0444);
 MODULE_PARM_DESC(nfakewriters, "Number of RCU fake writer threads");
-module_param(stat_interval, int, 0444);
+module_param(stat_interval, int, 0644);
 MODULE_PARM_DESC(stat_interval, "Number of seconds between stats printk()s");
 module_param(verbose, bool, 0444);
 MODULE_PARM_DESC(verbose, "Enable verbose debugging printk()s");

index 14a6c7b..d1a1533 100644
@@ -270,7 +270,6 @@ extern void init_idle_bootup_task(struct task_struct *idle);
 
 extern int runqueue_is_locked(int cpu);
 
-extern cpumask_var_t nohz_cpu_mask;
 #if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ)
 extern void select_nohz_load_balancer(int stop_tick);
 extern int get_nohz_timer_target(void);

index 1c87917..313c0f6 100644
@@ -5940,15 +5940,6 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu)
 }
 
 /*
- * In a system that switches off the HZ timer nohz_cpu_mask
- * indicates which cpus entered this state. This is used
- * in the rcu update to wait only for active cpus. For system
- * which do not switch off the HZ timer nohz_cpu_mask should
- * always be CPU_BITS_NONE.
- */
-cpumask_var_t nohz_cpu_mask;
-
-/*
  * Increase the granularity value when there are more CPUs,
  * because with more CPUs the 'effective latency' as visible
  * to users decreases. But the relationship is not linear,

@@ -8167,8 +8158,6 @@ void __init sched_init(void)
 	 */
 	current->sched_class = &fair_sched_class;
 
-	/* Allocate the nohz_cpu_mask if CONFIG_CPUMASK_OFFSTACK */
-	zalloc_cpumask_var(&nohz_cpu_mask, GFP_NOWAIT);
 #ifdef CONFIG_SMP
 	zalloc_cpumask_var(&sched_domains_tmpmask, GFP_NOWAIT);
 #ifdef CONFIG_NO_HZ

index d5097c4..eb98e55 100644
@@ -139,7 +139,6 @@ static void tick_nohz_update_jiffies(ktime_t now)
 	struct tick_sched *ts = &per_cpu(tick_cpu_sched, cpu);
 	unsigned long flags;
 
-	cpumask_clear_cpu(cpu, nohz_cpu_mask);
 	ts->idle_waketime = now;
 
 	local_irq_save(flags);

@@ -389,9 +388,6 @@ void tick_nohz_stop_sched_tick(int inidle)
 		else
 			expires.tv64 = KTIME_MAX;
 
-		if (delta_jiffies > 1)
-			cpumask_set_cpu(cpu, nohz_cpu_mask);
-
 		/* Skip reprogram of event if its not changed */
 		if (ts->tick_stopped && ktime_equal(expires, dev->next_event))
 			goto out;

@@ -441,7 +437,6 @@ void tick_nohz_stop_sched_tick(int inidle)
 		 * softirq.
 		 */
 		tick_do_update_jiffies64(ktime_get());
-		cpumask_clear_cpu(cpu, nohz_cpu_mask);
 	}
 	raise_softirq_irqoff(TIMER_SOFTIRQ);
 out:

@@ -524,7 +519,6 @@ void tick_nohz_restart_sched_tick(void)
 	/* Update jiffies first */
 	select_nohz_load_balancer(0);
 	tick_do_update_jiffies64(now);
-	cpumask_clear_cpu(cpu, nohz_cpu_mask);
 
 #ifndef CONFIG_VIRT_CPU_ACCOUNTING
 	/*

index 7e2f297..0c6c30d 100644
@@ -373,13 +373,6 @@ void rcu_enter_nohz(void)
 	smp_mb__after_atomic_inc();  /* Force ordering with next sojourn. */
 	WARN_ON_ONCE(atomic_read(&rdtp->dynticks) & 0x1);
 	local_irq_restore(flags);
-
-	/* If the interrupt queued a callback, get out of dyntick mode. */
-	if (in_irq() &&
-	    (__get_cpu_var(rcu_sched_data).nxtlist ||
-	     __get_cpu_var(rcu_bh_data).nxtlist ||
-	     rcu_preempt_needs_cpu(smp_processor_id())))
-		set_need_resched();
 }
 
 /*

index 0c6c30d..ebd18e5 100644
@@ -488,11 +488,11 @@ static int dyntick_save_progress_counter(struct rcu_data *rdp)
  */
 static int rcu_implicit_dynticks_qs(struct rcu_data *rdp)
 {
-	unsigned long curr;
-	unsigned long snap;
+	unsigned int curr;
+	unsigned int snap;
 
-	curr = (unsigned long)atomic_add_return(0, &rdp->dynticks->dynticks);
-	snap = (unsigned long)rdp->dynticks_snap;
+	curr = (unsigned int)atomic_add_return(0, &rdp->dynticks->dynticks);
+	snap = (unsigned int)rdp->dynticks_snap;
 
 	/*
 	 * If the CPU passed through or entered a dynticks idle phase with

@@ -502,7 +502,7 @@ 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 ((curr & 0x1) == 0 || ULONG_CMP_GE(curr, snap + 2)) {
+	if ((curr & 0x1) == 0 || UINT_CMP_GE(curr, snap + 2)) {
 		trace_rcu_fqs(rdp->rsp->name, rdp->gpnum, rdp->cpu, "dti");
 		rdp->dynticks_fqs++;
 		return 1;

index ea80396..b2e5fe8 100644
@@ -443,9 +443,7 @@ extern int rcu_my_thread_group_empty(void);
 	})
 #define __rcu_assign_pointer(p, v, space) \
 	({ \
-		if (!__builtin_constant_p(v) || \
-		    ((v) != NULL)) \
-			smp_wmb(); \
+		smp_wmb(); \
 		(p) = (typeof(*v) __force space *)(v); \
 	})
 

index b2e5fe8..9873040 100644
@@ -754,11 +754,18 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
  * any prior initialization.  Returns the value assigned.
  *
  * Inserts memory barriers on architectures that require them
- * (pretty much all of them other than x86), and also prevents
- * the compiler from reordering the code that initializes the
- * structure after the pointer assignment.  More importantly, this
- * call documents which pointers will be dereferenced by RCU read-side
- * code.
+ * (which is most of them), and also prevents the compiler from
+ * reordering the code that initializes the structure after the pointer
+ * assignment.  More importantly, this call documents which pointers
+ * will be dereferenced by RCU read-side code.
+ *
+ * In some special cases, you may use RCU_INIT_POINTER() instead
+ * of rcu_assign_pointer().  RCU_INIT_POINTER() is a bit faster due
+ * to the fact that it does not constrain either the CPU or the compiler.
+ * That said, using RCU_INIT_POINTER() when you should have used
+ * rcu_assign_pointer() is a very bad thing that results in
+ * impossible-to-diagnose memory corruption.  So please be careful.
+ * See the RCU_INIT_POINTER() comment header for details.
  */
 #define rcu_assign_pointer(p, v) \
 	__rcu_assign_pointer((p), (v), __rcu)

@@ -766,8 +773,34 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
 /**
  * RCU_INIT_POINTER() - initialize an RCU protected pointer
  *
- * Initialize an RCU-protected pointer in such a way to avoid RCU-lockdep
- * splats.
+ * Initialize an RCU-protected pointer in special cases where readers
+ * do not need ordering constraints on the CPU or the compiler.  These
+ * special cases are:
+ *
+ * 1.	This use of RCU_INIT_POINTER() is NULLing out the pointer -or-
+ * 2.	The caller has taken whatever steps are required to prevent
+ *	RCU readers from concurrently accessing this pointer -or-
+ * 3.	The referenced data structure has already been exposed to
+ *	readers either at compile time or via rcu_assign_pointer() -and-
+ *	a.	You have not made -any- reader-visible changes to
+ *		this structure since then -or-
+ *	b.	It is OK for readers accessing this structure from its
+ *		new location to see the old state of the structure.  (For
+ *		example, the changes were to statistical counters or to
+ *		other state where exact synchronization is not required.)
+ *
+ * Failure to follow these rules governing use of RCU_INIT_POINTER() will
+ * result in impossible-to-diagnose memory corruption.  As in the structures
+ * will look OK in crash dumps, but any concurrent RCU readers might
+ * see pre-initialized values of the referenced data structure.  So
+ * please be very careful how you use RCU_INIT_POINTER()!!!
+ *
+ * If you are creating an RCU-protected linked structure that is accessed
+ * by a single external-to-structure RCU-protected pointer, then you may
+ * use RCU_INIT_POINTER() to initialize the internal RCU-protected
+ * pointers, but you must use rcu_assign_pointer() to initialize the
+ * external-to-structure pointer -after- you have completely initialized
+ * the reader-accessible portions of the linked structure.
  */
 #define RCU_INIT_POINTER(p, v) \
 		p = (typeof(*v) __force __rcu *)(v)

index 9873040..e86bc28 100644
@@ -134,16 +134,6 @@ extern void call_rcu_sched(struct rcu_head *head,
 
 extern void synchronize_sched(void);
 
-static inline void __rcu_read_lock_bh(void)
-{
-	local_bh_disable();
-}
-
-static inline void __rcu_read_unlock_bh(void)
-{
-	local_bh_enable();
-}
-
 #ifdef CONFIG_PREEMPT_RCU
 
 extern void __rcu_read_lock(void);

@@ -686,7 +676,7 @@ static inline void rcu_read_unlock(void)
  */
 static inline void rcu_read_lock_bh(void)
 {
-	__rcu_read_lock_bh();
+	local_bh_disable();
 	__acquire(RCU_BH);
 	rcu_read_acquire_bh();
 }

@@ -700,7 +690,7 @@ static inline void rcu_read_unlock_bh(void)
 {
 	rcu_read_release_bh();
 	__release(RCU_BH);
-	__rcu_read_unlock_bh();
+	local_bh_enable();
 }
 
 /**

index 4bac5a2..ed70f6b 100644
@@ -415,10 +415,10 @@ void __rcu_read_unlock(void)
 {
 	struct task_struct *t = current;
 
-	barrier();  /* needed if we ever invoke rcu_read_unlock in rcutree.c */
 	if (t->rcu_read_lock_nesting != 1)
 		--t->rcu_read_lock_nesting;
 	else {
+		barrier();  /* critical section before exit code. */
 		t->rcu_read_lock_nesting = INT_MIN;
 		barrier();  /* assign before ->rcu_read_unlock_special load */
 		if (unlikely(ACCESS_ONCE(t->rcu_read_unlock_special)))

index ebd18e5..a07bf55 100644
@@ -583,7 +583,8 @@ static void print_other_cpu_stall(struct rcu_state *rsp)
 	}
 	printk("} (detected by %d, t=%ld jiffies)\n",
 	       smp_processor_id(), (long)(jiffies - rsp->gp_start));
-	trigger_all_cpu_backtrace();
+	if (!trigger_all_cpu_backtrace())
+		dump_stack();
 
 	/* If so configured, complain about tasks blocking the grace period. */
 

@@ -604,7 +605,8 @@ static void print_cpu_stall(struct rcu_state *rsp)
 	 */
 	printk(KERN_ERR "INFO: %s detected stall on CPU %d (t=%lu jiffies)\n",
 	       rsp->name, smp_processor_id(), jiffies - rsp->gp_start);
-	trigger_all_cpu_backtrace();
+	if (!trigger_all_cpu_backtrace())
+		dump_stack();
 
 	raw_spin_lock_irqsave(&rnp->lock, flags);
 	if (ULONG_CMP_GE(jiffies, rsp->jiffies_stall))

index a07bf55..0051dbf 100644
@@ -1707,7 +1707,8 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp)
 	check_cpu_stall(rsp, rdp);
 
 	/* Is the RCU core waiting for a quiescent state from this CPU? */
-	if (rdp->qs_pending && !rdp->passed_quiesce) {
+	if (rcu_scheduler_fully_active &&
+	    rdp->qs_pending && !rdp->passed_quiesce) {
 
 		/*
 		 * If force_quiescent_state() coming soon and this CPU

index d1a1533..90b02be 100644
@@ -1259,9 +1259,6 @@ struct task_struct {
 #ifdef CONFIG_PREEMPT_RCU
 	int rcu_read_lock_nesting;
 	char rcu_read_unlock_special;
-#if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU)
-	int rcu_boosted;
-#endif /* #if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU) */
 	struct list_head rcu_node_entry;
 #endif /* #ifdef CONFIG_PREEMPT_RCU */
 #ifdef CONFIG_TREE_PREEMPT_RCU

index ed70f6b..eeb38ee 100644
@@ -306,6 +306,9 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 	int empty_exp;
 	unsigned long flags;
 	struct list_head *np;
+#ifdef CONFIG_RCU_BOOST
+	struct rt_mutex *rbmp = NULL;
+#endif /* #ifdef CONFIG_RCU_BOOST */
 	struct rcu_node *rnp;
 	int special;
 

@@ -351,6 +354,7 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 		smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
 		np = rcu_next_node_entry(t, rnp);
 		list_del_init(&t->rcu_node_entry);
+		t->rcu_blocked_node = NULL;
 		trace_rcu_unlock_preempted_task("rcu_preempt",
 						rnp->gpnum, t->pid);
 		if (&t->rcu_node_entry == rnp->gp_tasks)

@@ -360,13 +364,12 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 #ifdef CONFIG_RCU_BOOST
 		if (&t->rcu_node_entry == rnp->boost_tasks)
 			rnp->boost_tasks = np;
-		/* Snapshot and clear ->rcu_boosted with rcu_node lock held. */
-		if (t->rcu_boosted) {
-			special |= RCU_READ_UNLOCK_BOOSTED;
-			t->rcu_boosted = 0;
+		/* Snapshot/clear ->rcu_boost_mutex with rcu_node lock held. */
+		if (t->rcu_boost_mutex) {
+			rbmp = t->rcu_boost_mutex;
+			t->rcu_boost_mutex = NULL;
 		}
 #endif /* #ifdef CONFIG_RCU_BOOST */
-		t->rcu_blocked_node = NULL;
 
 		/*
 		 * If this was the last task on the current list, and if

@@ -387,10 +390,8 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
 
 #ifdef CONFIG_RCU_BOOST
 		/* Unboost if we were boosted. */
-		if (special & RCU_READ_UNLOCK_BOOSTED) {
-			rt_mutex_unlock(t->rcu_boost_mutex);
-			t->rcu_boost_mutex = NULL;
-		}
+		if (rbmp)
+			rt_mutex_unlock(rbmp);
 #endif /* #ifdef CONFIG_RCU_BOOST */
 
 		/*

@@ -1206,7 +1207,6 @@ static int rcu_boost(struct rcu_node *rnp)
 	t = container_of(tb, struct task_struct, rcu_node_entry);
 	rt_mutex_init_proxy_locked(&mtx, t);
 	t->rcu_boost_mutex = &mtx;
-	t->rcu_boosted = 1;
 	raw_spin_unlock_irqrestore(&rnp->lock, flags);
 	rt_mutex_lock(&mtx);  /* Side effect: boosts task t's priority. */
 	rt_mutex_unlock(&mtx);  /* Keep lockdep happy. */

index 0051dbf..9970116 100644
@@ -838,8 +838,11 @@ rcu_start_gp(struct rcu_state *rsp, unsigned long flags)
 	struct rcu_data *rdp = this_cpu_ptr(rsp->rda);
 	struct rcu_node *rnp = rcu_get_root(rsp);
 
-	if (!cpu_needs_another_gp(rsp, rdp) || rsp->fqs_active) {
-		if (cpu_needs_another_gp(rsp, rdp))
+	if (!rcu_scheduler_fully_active ||
+	    !cpu_needs_another_gp(rsp, rdp) ||
+	    rsp->fqs_active) {
+		if (rcu_scheduler_fully_active &&
+		    cpu_needs_another_gp(rsp, rdp))
 			rsp->fqs_need_gp = 1;
 		if (rnp->completed == rsp->completed) {
 			raw_spin_unlock_irqrestore(&rnp->lock, flags);