Merge branches 'doc.2016.04.19a', 'exp.2016.03.31d', 'fixes.2016.03.31d' and 'torture.2016.04.21a' into HEAD

doc.2016.04.19a: Documentation updates
exp.2016.03.31d: Expedited grace-period updates
fixes.2016.03.31d: Miscellaneous fixes
torture.2016.004.21a Torture-test updates

Documentation/RCU/trace.txt

@@ -237,17 +237,17 @@ o	"ktl" is the low-order 16 bits (in hexadecimal) of the count of
 
 The output of "cat rcu/rcu_preempt/rcuexp" looks as follows:
 
-s=21872 wd0=0 wd1=0 wd2=0 wd3=5 n=0 enq=0 sc=21872
+s=21872 wd1=0 wd2=0 wd3=5 n=0 enq=0 sc=21872
 
 These fields are as follows:
 
 o	"s" is the sequence number, with an odd number indicating that
 	an expedited grace period is in progress.
 
-o	"wd0", "wd1", "wd2", and "wd3" are the number of times that an
+o	"wd1", "wd2", and "wd3" are the number of times that an attempt
-	attempt to start an expedited grace period found that someone
+	to start an expedited grace period found that someone else had
-	else had completed an expedited grace period that satisfies the
+	completed an expedited grace period that satisfies the attempted
-	attempted request.  "Our work is done."
+	request.  "Our work is done."
 
 o	"n" is number of times that a concurrent CPU-hotplug operation
 	forced a fallback to a normal grace period.

Documentation/kernel-parameters.txt

@@ -3284,6 +3284,44 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
 			Lazy RCU callbacks are those which RCU can
 			prove do nothing more than free memory.
 
+	rcuperf.gp_exp= [KNL]
+			Measure performance of expedited synchronous
+			grace-period primitives.
+
+	rcuperf.holdoff= [KNL]
+			Set test-start holdoff period.  The purpose of
+			this parameter is to delay the start of the
+			test until boot completes in order to avoid
+			interference.
+
+	rcuperf.nreaders= [KNL]
+			Set number of RCU readers.  The value -1 selects
+			N, where N is the number of CPUs.  A value
+			"n" less than -1 selects N-n+1, where N is again
+			the number of CPUs.  For example, -2 selects N
+			(the number of CPUs), -3 selects N+1, and so on.
+			A value of "n" less than or equal to -N selects
+			a single reader.
+
+	rcuperf.nwriters= [KNL]
+			Set number of RCU writers.  The values operate
+			the same as for rcuperf.nreaders.
+			N, where N is the number of CPUs
+
+	rcuperf.perf_runnable= [BOOT]
+			Start rcuperf running at boot time.
+
+	rcuperf.shutdown= [KNL]
+			Shut the system down after performance tests
+			complete.  This is useful for hands-off automated
+			testing.
+
+	rcuperf.perf_type= [KNL]
+			Specify the RCU implementation to test.
+
+	rcuperf.verbose= [KNL]
+			Enable additional printk() statements.
+
 	rcutorture.cbflood_inter_holdoff= [KNL]
 			Set holdoff time (jiffies) between successive
 			callback-flood tests.

include/linux/rcupdate.h

@@ -508,14 +508,7 @@ int rcu_read_lock_bh_held(void);
  * CONFIG_DEBUG_LOCK_ALLOC, this assumes we are in an RCU-sched read-side
  * critical section unless it can prove otherwise.
  */
-#ifdef CONFIG_PREEMPT_COUNT
 int rcu_read_lock_sched_held(void);
-#else /* #ifdef CONFIG_PREEMPT_COUNT */
-static inline int rcu_read_lock_sched_held(void)
-{
-	return 1;
-}
-#endif /* #else #ifdef CONFIG_PREEMPT_COUNT */
 
 #else /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */
 
@@ -532,18 +525,10 @@ static inline int rcu_read_lock_bh_held(void)
 	return 1;
 }
 
-#ifdef CONFIG_PREEMPT_COUNT
 static inline int rcu_read_lock_sched_held(void)
 {
-	return preempt_count() != 0 || irqs_disabled();
+	return !preemptible();
 }
-#else /* #ifdef CONFIG_PREEMPT_COUNT */
-static inline int rcu_read_lock_sched_held(void)
-{
-	return 1;
-}
-#endif /* #else #ifdef CONFIG_PREEMPT_COUNT */
-
 #endif /* #else #ifdef CONFIG_DEBUG_LOCK_ALLOC */
 
 #ifdef CONFIG_PROVE_RCU

include/linux/rcutiny.h

@@ -149,6 +149,22 @@ static inline unsigned long rcu_batches_completed_sched(void)
 	return 0;
 }
 
+/*
+ * Return the number of expedited grace periods completed.
+ */
+static inline unsigned long rcu_exp_batches_completed(void)
+{
+	return 0;
+}
+
+/*
+ * Return the number of expedited sched grace periods completed.
+ */
+static inline unsigned long rcu_exp_batches_completed_sched(void)
+{
+	return 0;
+}
+
 static inline void rcu_force_quiescent_state(void)
 {
 }

include/linux/rcutree.h

@@ -87,6 +87,8 @@ unsigned long rcu_batches_started_sched(void);
 unsigned long rcu_batches_completed(void);
 unsigned long rcu_batches_completed_bh(void);
 unsigned long rcu_batches_completed_sched(void);
+unsigned long rcu_exp_batches_completed(void);
+unsigned long rcu_exp_batches_completed_sched(void);
 void show_rcu_gp_kthreads(void);
 
 void rcu_force_quiescent_state(void);

include/trace/events/rcu.h

@@ -171,6 +171,77 @@ TRACE_EVENT(rcu_grace_period_init,
 		  __entry->grplo, __entry->grphi, __entry->qsmask)
 );
 
+/*
+ * Tracepoint for expedited grace-period events.  Takes a string identifying
+ * the RCU flavor, the expedited grace-period sequence number, and a string
+ * identifying the grace-period-related event as follows:
+ *
+ *	"snap": Captured snapshot of expedited grace period sequence number.
+ *	"start": Started a real expedited grace period.
+ *	"end": Ended a real expedited grace period.
+ *	"endwake": Woke piggybackers up.
+ *	"done": Someone else did the expedited grace period for us.
+ */
+TRACE_EVENT(rcu_exp_grace_period,
+
+	TP_PROTO(const char *rcuname, unsigned long gpseq, const char *gpevent),
+
+	TP_ARGS(rcuname, gpseq, gpevent),
+
+	TP_STRUCT__entry(
+		__field(const char *, rcuname)
+		__field(unsigned long, gpseq)
+		__field(const char *, gpevent)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->gpseq = gpseq;
+		__entry->gpevent = gpevent;
+	),
+
+	TP_printk("%s %lu %s",
+		  __entry->rcuname, __entry->gpseq, __entry->gpevent)
+);
+
+/*
+ * Tracepoint for expedited grace-period funnel-locking events.  Takes a
+ * string identifying the RCU flavor, an integer identifying the rcu_node
+ * combining-tree level, another pair of integers identifying the lowest-
+ * and highest-numbered CPU associated with the current rcu_node structure,
+ * and a string.  identifying the grace-period-related event as follows:
+ *
+ *	"nxtlvl": Advance to next level of rcu_node funnel
+ *	"wait": Wait for someone else to do expedited GP
+ */
+TRACE_EVENT(rcu_exp_funnel_lock,
+
+	TP_PROTO(const char *rcuname, u8 level, int grplo, int grphi,
+		 const char *gpevent),
+
+	TP_ARGS(rcuname, level, grplo, grphi, gpevent),
+
+	TP_STRUCT__entry(
+		__field(const char *, rcuname)
+		__field(u8, level)
+		__field(int, grplo)
+		__field(int, grphi)
+		__field(const char *, gpevent)
+	),
+
+	TP_fast_assign(
+		__entry->rcuname = rcuname;
+		__entry->level = level;
+		__entry->grplo = grplo;
+		__entry->grphi = grphi;
+		__entry->gpevent = gpevent;
+	),
+
+	TP_printk("%s %d %d %d %s",
+		  __entry->rcuname, __entry->level, __entry->grplo,
+		  __entry->grphi, __entry->gpevent)
+);
+
 /*
  * Tracepoint for RCU no-CBs CPU callback handoffs.  This event is intended
  * to assist debugging of these handoffs.
@@ -704,11 +775,15 @@ TRACE_EVENT(rcu_barrier,
 #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_future_grace_period(rcuname, gpnum, completed, c, \
 				      level, grplo, grphi, event) \
 				      do { } while (0)
+#define trace_rcu_grace_period_init(rcuname, gpnum, level, grplo, grphi, \
+				    qsmask) do { } while (0)
+#define trace_rcu_exp_grace_period(rcuname, gqseq, gpevent) \
+	do { } while (0)
+#define trace_rcu_exp_funnel_lock(rcuname, level, grplo, grphi, gpevent) \
+	do { } while (0)
 #define trace_rcu_nocb_wake(rcuname, cpu, reason) 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)

kernel/rcu/Makefile

@@ -5,6 +5,7 @@ KCOV_INSTRUMENT := n
 obj-y += update.o sync.o
 obj-$(CONFIG_SRCU) += srcu.o
 obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o
+obj-$(CONFIG_RCU_PERF_TEST) += rcuperf.o
 obj-$(CONFIG_TREE_RCU) += tree.o
 obj-$(CONFIG_PREEMPT_RCU) += tree.o
 obj-$(CONFIG_TREE_RCU_TRACE) += tree_trace.o

kernel/rcu/rcutorture.c

@@ -130,8 +130,8 @@ static struct rcu_torture __rcu *rcu_torture_current;
 static unsigned long rcu_torture_current_version;
 static struct rcu_torture rcu_tortures[10 * RCU_TORTURE_PIPE_LEN];
 static DEFINE_SPINLOCK(rcu_torture_lock);
-static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_count) = { 0 };
+static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_count);
-static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_batch) = { 0 };
+static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_batch);
 static atomic_t rcu_torture_wcount[RCU_TORTURE_PIPE_LEN + 1];
 static atomic_t n_rcu_torture_alloc;
 static atomic_t n_rcu_torture_alloc_fail;
@@ -916,7 +916,7 @@ rcu_torture_fqs(void *arg)
 static int
 rcu_torture_writer(void *arg)
 {
-	bool can_expedite = !rcu_gp_is_expedited();
+	bool can_expedite = !rcu_gp_is_expedited() && !rcu_gp_is_normal();
 	int expediting = 0;
 	unsigned long gp_snap;
 	bool gp_cond1 = gp_cond, gp_exp1 = gp_exp, gp_normal1 = gp_normal;
@@ -932,7 +932,7 @@ rcu_torture_writer(void *arg)
 	VERBOSE_TOROUT_STRING("rcu_torture_writer task started");
 	if (!can_expedite) {
 		pr_alert("%s" TORTURE_FLAG
-			 " Grace periods expedited from boot/sysfs for %s,\n",
+			 " GP expediting controlled from boot/sysfs for %s,\n",
 			 torture_type, cur_ops->name);
 		pr_alert("%s" TORTURE_FLAG
 			 " Disabled dynamic grace-period expediting.\n",
@@ -1478,7 +1478,9 @@ static int rcu_torture_barrier_cbs(void *arg)
 		 * The above smp_load_acquire() ensures barrier_phase load
 		 * is ordered before the folloiwng ->call().
 		 */
+		local_irq_disable(); /* Just to test no-irq call_rcu(). */
 		cur_ops->call(&rcu, rcu_torture_barrier_cbf);
+		local_irq_enable();
 		if (atomic_dec_and_test(&barrier_cbs_count))
 			wake_up(&barrier_wq);
 	} while (!torture_must_stop());
@@ -1585,7 +1587,7 @@ static int rcutorture_cpu_notify(struct notifier_block *self,
 {
 	long cpu = (long)hcpu;
 
-	switch (action) {
+	switch (action & ~CPU_TASKS_FROZEN) {
 	case CPU_ONLINE:
 	case CPU_DOWN_FAILED:
 		(void)rcutorture_booster_init(cpu);

kernel/rcu/tree.h

@@ -70,7 +70,6 @@
 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0 }
 #  define RCU_NODE_NAME_INIT  { "rcu_node_0" }
 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0" }
-#  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0" }
 #elif NR_CPUS <= RCU_FANOUT_2
 #  define RCU_NUM_LVLS	      2
 #  define NUM_RCU_LVL_0	      1
@@ -79,7 +78,6 @@
 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1 }
 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1" }
 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1" }
-#  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1" }
 #elif NR_CPUS <= RCU_FANOUT_3
 #  define RCU_NUM_LVLS	      3
 #  define NUM_RCU_LVL_0	      1
@@ -89,7 +87,6 @@
 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2 }
 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1", "rcu_node_2" }
 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2" }
-#  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2" }
 #elif NR_CPUS <= RCU_FANOUT_4
 #  define RCU_NUM_LVLS	      4
 #  define NUM_RCU_LVL_0	      1
@@ -100,7 +97,6 @@
 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2, NUM_RCU_LVL_3 }
 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1", "rcu_node_2", "rcu_node_3" }
 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2", "rcu_node_fqs_3" }
-#  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2", "rcu_node_exp_3" }
 #else
 # error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
 #endif /* #if (NR_CPUS) <= RCU_FANOUT_1 */
@@ -252,7 +248,9 @@ struct rcu_node {
 				/* Counts of upcoming no-CB GP requests. */
 	raw_spinlock_t fqslock ____cacheline_internodealigned_in_smp;
 
-	struct mutex exp_funnel_mutex ____cacheline_internodealigned_in_smp;
+	spinlock_t exp_lock ____cacheline_internodealigned_in_smp;
+	unsigned long exp_seq_rq;
+	wait_queue_head_t exp_wq[4];
 } ____cacheline_internodealigned_in_smp;
 
 /*
@@ -387,11 +385,9 @@ struct rcu_data {
 #ifdef CONFIG_RCU_FAST_NO_HZ
 	struct rcu_head oom_head;
 #endif /* #ifdef CONFIG_RCU_FAST_NO_HZ */
-	struct mutex exp_funnel_mutex;
+	atomic_long_t exp_workdone1;	/* # done by others #1. */
-	atomic_long_t expedited_workdone0;	/* # done by others #0. */
+	atomic_long_t exp_workdone2;	/* # done by others #2. */
-	atomic_long_t expedited_workdone1;	/* # done by others #1. */
+	atomic_long_t exp_workdone3;	/* # done by others #3. */
-	atomic_long_t expedited_workdone2;	/* # done by others #2. */
-	atomic_long_t expedited_workdone3;	/* # done by others #3. */
 
 	/* 7) Callback offloading. */
 #ifdef CONFIG_RCU_NOCB_CPU
@@ -505,6 +501,8 @@ struct rcu_state {
 						/*  _rcu_barrier(). */
 	/* End of fields guarded by barrier_mutex. */
 
+	struct mutex exp_mutex;			/* Serialize expedited GP. */
+	struct mutex exp_wake_mutex;		/* Serialize wakeup. */
 	unsigned long expedited_sequence;	/* Take a ticket. */
 	atomic_long_t expedited_normal;		/* # fallbacks to normal. */
 	atomic_t expedited_need_qs;		/* # CPUs left to check in. */
@@ -513,6 +511,8 @@ struct rcu_state {
 
 	unsigned long jiffies_force_qs;		/* Time at which to invoke */
 						/*  force_quiescent_state(). */
+	unsigned long jiffies_kick_kthreads;	/* Time at which to kick */
+						/*  kthreads, if configured. */
 	unsigned long n_force_qs;		/* Number of calls to */
 						/*  force_quiescent_state(). */
 	unsigned long n_force_qs_lh;		/* ~Number of calls leaving */

kernel/rcu/tree_plugin.h

@@ -722,18 +722,22 @@ static void sync_rcu_exp_handler(void *info)
  * synchronize_rcu_expedited - Brute-force RCU grace period
  *
  * Wait for an RCU-preempt grace period, but expedite it.  The basic
- * idea is to invoke synchronize_sched_expedited() to push all the tasks to
+ * idea is to IPI all non-idle non-nohz online CPUs.  The IPI handler
- * the ->blkd_tasks lists and wait for this list to drain.  This consumes
+ * checks whether the CPU is in an RCU-preempt critical section, and
- * significant time on all CPUs and is unfriendly to real-time workloads,
+ * if so, it sets a flag that causes the outermost rcu_read_unlock()
- * so is thus not recommended for any sort of common-case code.
+ * to report the quiescent state.  On the other hand, if the CPU is
- * In fact, if you are using synchronize_rcu_expedited() in a loop,
+ * not in an RCU read-side critical section, the IPI handler reports
- * please restructure your code to batch your updates, and then Use a
+ * the quiescent state immediately.
- * single synchronize_rcu() instead.
+ *
+ * Although this is a greate improvement over previous expedited
+ * implementations, it is still unfriendly to real-time workloads, so is
+ * thus not recommended for any sort of common-case code.  In fact, if
+ * you are using synchronize_rcu_expedited() in a loop, please restructure
+ * your code to batch your updates, and then Use a single synchronize_rcu()
+ * instead.
  */
 void synchronize_rcu_expedited(void)
 {
-	struct rcu_node *rnp;
-	struct rcu_node *rnp_unlock;
 	struct rcu_state *rsp = rcu_state_p;
 	unsigned long s;
 
@@ -744,23 +748,14 @@ void synchronize_rcu_expedited(void)
 	}
 
 	s = rcu_exp_gp_seq_snap(rsp);
-
+	if (exp_funnel_lock(rsp, s))
-	rnp_unlock = exp_funnel_lock(rsp, s);
-	if (rnp_unlock == NULL)
 		return;  /* Someone else did our work for us. */
 
-	rcu_exp_gp_seq_start(rsp);
-
 	/* Initialize the rcu_node tree in preparation for the wait. */
 	sync_rcu_exp_select_cpus(rsp, sync_rcu_exp_handler);
 
-	/* Wait for snapshotted ->blkd_tasks lists to drain. */
+	/* Wait for ->blkd_tasks lists to drain, then wake everyone up. */
-	rnp = rcu_get_root(rsp);
+	rcu_exp_wait_wake(rsp, s);
-	synchronize_sched_expedited_wait(rsp);
-
-	/* Clean up and exit. */
-	rcu_exp_gp_seq_end(rsp);
-	mutex_unlock(&rnp_unlock->exp_funnel_mutex);
 }
 EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);
 

kernel/rcu/tree_trace.c

@@ -185,17 +185,16 @@ static int show_rcuexp(struct seq_file *m, void *v)
 	int cpu;
 	struct rcu_state *rsp = (struct rcu_state *)m->private;
 	struct rcu_data *rdp;
-	unsigned long s0 = 0, s1 = 0, s2 = 0, s3 = 0;
+	unsigned long s1 = 0, s2 = 0, s3 = 0;
 
 	for_each_possible_cpu(cpu) {
 		rdp = per_cpu_ptr(rsp->rda, cpu);
-		s0 += atomic_long_read(&rdp->expedited_workdone0);
+		s1 += atomic_long_read(&rdp->exp_workdone1);
-		s1 += atomic_long_read(&rdp->expedited_workdone1);
+		s2 += atomic_long_read(&rdp->exp_workdone2);
-		s2 += atomic_long_read(&rdp->expedited_workdone2);
+		s3 += atomic_long_read(&rdp->exp_workdone3);
-		s3 += atomic_long_read(&rdp->expedited_workdone3);
 	}
-	seq_printf(m, "s=%lu wd0=%lu wd1=%lu wd2=%lu wd3=%lu n=%lu enq=%d sc=%lu\n",
+	seq_printf(m, "s=%lu wd1=%lu wd2=%lu wd3=%lu n=%lu enq=%d sc=%lu\n",
-		   rsp->expedited_sequence, s0, s1, s2, s3,
+		   rsp->expedited_sequence, s1, s2, s3,
 		   atomic_long_read(&rsp->expedited_normal),
 		   atomic_read(&rsp->expedited_need_qs),
 		   rsp->expedited_sequence / 2);

kernel/rcu/update.c

@@ -67,7 +67,7 @@ static int rcu_normal_after_boot;
 module_param(rcu_normal_after_boot, int, 0);
 #endif /* #ifndef CONFIG_TINY_RCU */
 
-#if defined(CONFIG_DEBUG_LOCK_ALLOC) && defined(CONFIG_PREEMPT_COUNT)
+#ifdef CONFIG_DEBUG_LOCK_ALLOC
 /**
  * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section?
  *
@@ -111,7 +111,7 @@ int rcu_read_lock_sched_held(void)
 		return 0;
 	if (debug_locks)
 		lockdep_opinion = lock_is_held(&rcu_sched_lock_map);
-	return lockdep_opinion || preempt_count() != 0 || irqs_disabled();
+	return lockdep_opinion || !preemptible();
 }
 EXPORT_SYMBOL(rcu_read_lock_sched_held);
 #endif

kernel/torture.c

@@ -451,6 +451,7 @@ static int torture_shutdown(void *arg)
 		torture_shutdown_hook();
 	else
 		VERBOSE_TOROUT_STRING("No torture_shutdown_hook(), skipping.");
+	ftrace_dump(DUMP_ALL);
 	kernel_power_off();	/* Shut down the system. */
 	return 0;
 }
@@ -602,8 +603,9 @@ bool torture_init_begin(char *ttype, bool v, int *runnable)
 {
 	mutex_lock(&fullstop_mutex);
 	if (torture_type != NULL) {
-		pr_alert("torture_init_begin: refusing %s init: %s running",
+		pr_alert("torture_init_begin: Refusing %s init: %s running.\n",
 			 ttype, torture_type);
+		pr_alert("torture_init_begin: One torture test at a time!\n");
 		mutex_unlock(&fullstop_mutex);
 		return false;
 	}

lib/Kconfig.debug

@@ -1289,6 +1289,39 @@ config TORTURE_TEST
 	tristate
 	default n
 
+config RCU_PERF_TEST
+	tristate "performance tests for RCU"
+	depends on DEBUG_KERNEL
+	select TORTURE_TEST
+	select SRCU
+	select TASKS_RCU
+	default n
+	help
+	  This option provides a kernel module that runs performance
+	  tests on the RCU infrastructure.  The kernel module may be built
+	  after the fact on the running kernel to be tested, if desired.
+
+	  Say Y here if you want RCU performance tests to be built into
+	  the kernel.
+	  Say M if you want the RCU performance tests to build as a module.
+	  Say N if you are unsure.
+
+config RCU_PERF_TEST_RUNNABLE
+	bool "performance tests for RCU runnable by default"
+	depends on RCU_PERF_TEST = y
+	default n
+	help
+	  This option provides a way to build the RCU performance tests
+	  directly into the kernel without them starting up at boot time.
+	  You can use /sys/module to manually override this setting.
+	  This /proc file is available only when the RCU performance
+	  tests have been built into the kernel.
+
+	  Say Y here if you want the RCU performance tests to start during
+	  boot (you probably don't).
+	  Say N here if you want the RCU performance tests to start only
+	  after being manually enabled via /sys/module.
+
 config RCU_TORTURE_TEST
 	tristate "torture tests for RCU"
 	depends on DEBUG_KERNEL

tools/testing/selftests/rcutorture/bin/jitter.sh

@@ -0,0 +1,90 @@
+#!/bin/bash
+#
+# Alternate sleeping and spinning on randomly selected CPUs.  The purpose
+# of this script is to inflict random OS jitter on a concurrently running
+# test.
+#
+# Usage: jitter.sh me duration [ sleepmax [ spinmax ] ]
+#
+# me: Random-number-generator seed salt.
+# duration: Time to run in seconds.
+# sleepmax: Maximum microseconds to sleep, defaults to one second.
+# spinmax: Maximum microseconds to spin, defaults to one millisecond.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, you can access it online at
+# http://www.gnu.org/licenses/gpl-2.0.html.
+#
+# Copyright (C) IBM Corporation, 2016
+#
+# Authors: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+
+me=$(($1 * 1000))
+duration=$2
+sleepmax=${3-1000000}
+spinmax=${4-1000}
+
+n=1
+
+starttime=`awk 'BEGIN { print systime(); }' < /dev/null`
+
+while :
+do
+	# Check for done.
+	t=`awk -v s=$starttime 'BEGIN { print systime() - s; }' < /dev/null`
+	if test "$t" -gt "$duration"
+	then
+		exit 0;
+	fi
+
+	# Set affinity to randomly selected CPU
+	cpus=`ls /sys/devices/system/cpu/*/online |
+		sed -e 's,/[^/]*$,,' -e 's/^[^0-9]*//' |
+		grep -v '^0*$'`
+	cpumask=`awk -v cpus="$cpus" -v me=$me -v n=$n 'BEGIN {
+		srand(n + me + systime());
+		ncpus = split(cpus, ca);
+		curcpu = ca[int(rand() * ncpus + 1)];
+		mask = lshift(1, curcpu);
+		if (mask + 0 <= 0)
+			mask = 1;
+		printf("%#x\n", mask);
+	}' < /dev/null`
+	n=$(($n+1))
+	if ! taskset -p $cpumask $$ > /dev/null 2>&1
+	then
+		echo taskset failure: '"taskset -p ' $cpumask $$ '"'
+		exit 1
+	fi
+
+	# Sleep a random duration
+	sleeptime=`awk -v me=$me -v n=$n -v sleepmax=$sleepmax 'BEGIN {
+		srand(n + me + systime());
+		printf("%06d", int(rand() * sleepmax));
+	}' < /dev/null`
+	n=$(($n+1))
+	sleep .$sleeptime
+
+	# Spin a random duration
+	limit=`awk -v me=$me -v n=$n -v spinmax=$spinmax 'BEGIN {
+		srand(n + me + systime());
+		printf("%06d", int(rand() * spinmax));
+	}' < /dev/null`
+	n=$(($n+1))
+	for i in {1..$limit}
+	do
+		echo > /dev/null
+	done
+done
+
+exit 1

tools/testing/selftests/rcutorture/bin/kvm-recheck-rcuperf-ftrace.sh

@@ -0,0 +1,121 @@
+#!/bin/bash
+#
+# Analyze a given results directory for rcuperf performance measurements,
+# looking for ftrace data.  Exits with 0 if data was found, analyzed, and
+# printed.  Intended to be invoked from kvm-recheck-rcuperf.sh after
+# argument checking.
+#
+# Usage: kvm-recheck-rcuperf-ftrace.sh resdir
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, you can access it online at
+# http://www.gnu.org/licenses/gpl-2.0.html.
+#
+# Copyright (C) IBM Corporation, 2016
+#
+# Authors: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+
+i="$1"
+. tools/testing/selftests/rcutorture/bin/functions.sh
+
+if test "`grep -c 'rcu_exp_grace_period.*start' < $i/console.log`" -lt 100
+then
+	exit 10
+fi
+
+sed -e 's/^\[[^]]*]//' < $i/console.log |
+grep 'us : rcu_exp_grace_period' |
+sed -e 's/us : / : /' |
+tr -d '\015' |
+awk '
+$8 == "start" {
+	if (starttask != "")
+		nlost++;
+	starttask = $1;
+	starttime = $3;
+	startseq = $7;
+}
+
+$8 == "end" {
+	if (starttask == $1 && startseq == $7) {
+		curgpdur = $3 - starttime;
+		gptimes[++n] = curgpdur;
+		gptaskcnt[starttask]++;
+		sum += curgpdur;
+		if (curgpdur > 1000)
+			print "Long GP " starttime "us to " $3 "us (" curgpdur "us)";
+		starttask = "";
+	} else {
+		# Lost a message or some such, reset.
+		starttask = "";
+		nlost++;
+	}
+}
+
+$8 == "done" {
+	piggybackcnt[$1]++;
+}
+
+END {
+	newNR = asort(gptimes);
+	if (newNR <= 0) {
+		print "No ftrace records found???"
+		exit 10;
+	}
+	pct50 = int(newNR * 50 / 100);
+	if (pct50 < 1)
+		pct50 = 1;
+	pct90 = int(newNR * 90 / 100);
+	if (pct90 < 1)
+		pct90 = 1;
+	pct99 = int(newNR * 99 / 100);
+	if (pct99 < 1)
+		pct99 = 1;
+	div = 10 ** int(log(gptimes[pct90]) / log(10) + .5) / 100;
+	print "Histogram bucket size: " div;
+	last = gptimes[1] - 10;
+	count = 0;
+	for (i = 1; i <= newNR; i++) {
+		current = div * int(gptimes[i] / div);
+		if (last == current) {
+			count++;
+		} else {
+			if (count > 0)
+				print last, count;
+			count = 1;
+			last = current;
+		}
+	}
+	if (count > 0)
+		print last, count;
+	print "Distribution of grace periods across tasks:";
+	for (i in gptaskcnt) {
+		print "\t" i, gptaskcnt[i];
+		nbatches += gptaskcnt[i];
+	}
+	ngps = nbatches;
+	print "Distribution of piggybacking across tasks:";
+	for (i in piggybackcnt) {
+		print "\t" i, piggybackcnt[i];
+		ngps += piggybackcnt[i];
+	}
+	print "Average grace-period duration: " sum / newNR " microseconds";
+	print "Minimum grace-period duration: " gptimes[1];
+	print "50th percentile grace-period duration: " gptimes[pct50];
+	print "90th percentile grace-period duration: " gptimes[pct90];
+	print "99th percentile grace-period duration: " gptimes[pct99];
+	print "Maximum grace-period duration: " gptimes[newNR];
+	print "Grace periods: " ngps + 0 " Batches: " nbatches + 0 " Ratio: " ngps / nbatches " Lost: " nlost + 0;
+	print "Computed from ftrace data.";
+}'
+exit 0

tools/testing/selftests/rcutorture/bin/kvm-recheck-rcuperf.sh

@@ -0,0 +1,96 @@
+#!/bin/bash
+#
+# Analyze a given results directory for rcuperf performance measurements.
+#
+# Usage: kvm-recheck-rcuperf.sh resdir
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, you can access it online at
+# http://www.gnu.org/licenses/gpl-2.0.html.
+#
+# Copyright (C) IBM Corporation, 2016
+#
+# Authors: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+
+i="$1"
+if test -d $i
+then
+	:
+else
+	echo Unreadable results directory: $i
+	exit 1
+fi
+PATH=`pwd`/tools/testing/selftests/rcutorture/bin:$PATH; export PATH
+. tools/testing/selftests/rcutorture/bin/functions.sh
+
+if kvm-recheck-rcuperf-ftrace.sh $i
+then
+	# ftrace data was successfully analyzed, call it good!
+	exit 0
+fi
+
+configfile=`echo $i | sed -e 's/^.*\///'`
+
+sed -e 's/^\[[^]]*]//' < $i/console.log |
+awk '
+/-perf: .* gps: .* batches:/ {
+	ngps = $9;
+	nbatches = $11;
+}
+
+/-perf: .*writer-duration/ {
+	gptimes[++n] = $5 / 1000.;
+	sum += $5 / 1000.;
+}
+
+END {
+	newNR = asort(gptimes);
+	if (newNR <= 0) {
+		print "No rcuperf records found???"
+		exit;
+	}
+	pct50 = int(newNR * 50 / 100);
+	if (pct50 < 1)
+		pct50 = 1;
+	pct90 = int(newNR * 90 / 100);
+	if (pct90 < 1)
+		pct90 = 1;
+	pct99 = int(newNR * 99 / 100);
+	if (pct99 < 1)
+		pct99 = 1;
+	div = 10 ** int(log(gptimes[pct90]) / log(10) + .5) / 100;
+	print "Histogram bucket size: " div;
+	last = gptimes[1] - 10;
+	count = 0;
+	for (i = 1; i <= newNR; i++) {
+		current = div * int(gptimes[i] / div);
+		if (last == current) {
+			count++;
+		} else {
+			if (count > 0)
+				print last, count;
+			count = 1;
+			last = current;
+		}
+	}
+	if (count > 0)
+		print last, count;
+	print "Average grace-period duration: " sum / newNR " microseconds";
+	print "Minimum grace-period duration: " gptimes[1];
+	print "50th percentile grace-period duration: " gptimes[pct50];
+	print "90th percentile grace-period duration: " gptimes[pct90];
+	print "99th percentile grace-period duration: " gptimes[pct99];
+	print "Maximum grace-period duration: " gptimes[newNR];
+	print "Grace periods: " ngps + 0 " Batches: " nbatches + 0 " Ratio: " ngps / nbatches;
+	print "Computed from rcuperf printk output.";
+}'

tools/testing/selftests/rcutorture/bin/kvm-recheck.sh

@@ -48,7 +48,10 @@ do
 				cat $i/Make.oldconfig.err
 			fi
 			parse-build.sh $i/Make.out $configfile
-			parse-torture.sh $i/console.log $configfile
+			if test "$TORTURE_SUITE" != rcuperf
+			then
+				parse-torture.sh $i/console.log $configfile
+			fi
 			parse-console.sh $i/console.log $configfile
 			if test -r $i/Warnings
 			then