Adding upstream version 6.7.7.upstream/6.7.7

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

11 files changed, 269 insertions, 86 deletions

diff --git a/kernel/locking/lock_events.c b/kernel/locking/lock_events.c
index fa2c2f951..e68d82099 100644
--- a/kernel/locking/lock_events.c
+++ b/kernel/locking/lock_events.c
@@ -146,7 +146,7 @@ static int __init init_lockevent_counts(void)
 	struct dentry *d_counts = debugfs_create_dir(LOCK_EVENTS_DIR, NULL);
 	int i;
 
-	if (!d_counts)
+	if (IS_ERR(d_counts))
 		goto out;
 
 	/*
@@ -159,14 +159,14 @@ static int __init init_lockevent_counts(void)
 	for (i = 0; i < lockevent_num; i++) {
 		if (skip_lockevent(lockevent_names[i]))
 			continue;
-		if (!debugfs_create_file(lockevent_names[i], 0400, d_counts,
-					 (void *)(long)i, &fops_lockevent))
+		if (IS_ERR(debugfs_create_file(lockevent_names[i], 0400, d_counts,
+					 (void *)(long)i, &fops_lockevent)))
 			goto fail_undo;
 	}
 
-	if (!debugfs_create_file(lockevent_names[LOCKEVENT_reset_cnts], 0200,
+	if (IS_ERR(debugfs_create_file(lockevent_names[LOCKEVENT_reset_cnts], 0200,
 				 d_counts, (void *)(long)LOCKEVENT_reset_cnts,
-				 &fops_lockevent))
+				 &fops_lockevent)))
 		goto fail_undo;
 
 	return 0;
diff --git a/kernel/locking/lockdep_proc.c b/kernel/locking/lockdep_proc.c
index 15fdc7fa5..e2bfb1db5 100644
--- a/kernel/locking/lockdep_proc.c
+++ b/kernel/locking/lockdep_proc.c
@@ -440,7 +440,7 @@ static void snprint_time(char *buf, size_t bufsiz, s64 nr)
 
 static void seq_time(struct seq_file *m, s64 time)
 {
-	char num[15];
+	char num[22];
 
 	snprint_time(num, sizeof(num), time);
 	seq_printf(m, " %14s", num);
diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c
index 270c7f80c..69d3cd2cf 100644
--- a/kernel/locking/locktorture.c
+++ b/kernel/locking/locktorture.c
@@ -33,21 +33,23 @@
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Paul E. McKenney <paulmck@linux.ibm.com>");
 
-torture_param(int, nwriters_stress, -1, "Number of write-locking stress-test threads");
-torture_param(int, nreaders_stress, -1, "Number of read-locking stress-test threads");
+torture_param(int, acq_writer_lim, 0, "Write_acquisition time limit (jiffies).");
+torture_param(int, call_rcu_chains, 0, "Self-propagate call_rcu() chains during test (0=disable).");
 torture_param(int, long_hold, 100, "Do occasional long hold of lock (ms), 0=disable");
+torture_param(int, nested_locks, 0, "Number of nested locks (max = 8)");
+torture_param(int, nreaders_stress, -1, "Number of read-locking stress-test threads");
+torture_param(int, nwriters_stress, -1, "Number of write-locking stress-test threads");
 torture_param(int, onoff_holdoff, 0, "Time after boot before CPU hotplugs (s)");
 torture_param(int, onoff_interval, 0, "Time between CPU hotplugs (s), 0=disable");
+torture_param(int, rt_boost, 2,
+		   "Do periodic rt-boost. 0=Disable, 1=Only for rt_mutex, 2=For all lock types.");
+torture_param(int, rt_boost_factor, 50, "A factor determining how often rt-boost happens.");
 torture_param(int, shuffle_interval, 3, "Number of jiffies between shuffles, 0=disable");
 torture_param(int, shutdown_secs, 0, "Shutdown time (j), <= zero to disable.");
 torture_param(int, stat_interval, 60, "Number of seconds between stats printk()s");
 torture_param(int, stutter, 5, "Number of jiffies to run/halt test, 0=disable");
-torture_param(int, rt_boost, 2,
-		   "Do periodic rt-boost. 0=Disable, 1=Only for rt_mutex, 2=For all lock types.");
-torture_param(int, rt_boost_factor, 50, "A factor determining how often rt-boost happens.");
-torture_param(int, writer_fifo, 0, "Run writers at sched_set_fifo() priority");
 torture_param(int, verbose, 1, "Enable verbose debugging printk()s");
-torture_param(int, nested_locks, 0, "Number of nested locks (max = 8)");
+torture_param(int, writer_fifo, 0, "Run writers at sched_set_fifo() priority");
 /* Going much higher trips "BUG: MAX_LOCKDEP_CHAIN_HLOCKS too low!" errors */
 #define MAX_NESTED_LOCKS 8
 
@@ -56,6 +58,55 @@ module_param(torture_type, charp, 0444);
 MODULE_PARM_DESC(torture_type,
 		 "Type of lock to torture (spin_lock, spin_lock_irq, mutex_lock, ...)");
 
+static cpumask_var_t bind_readers; // Bind the readers to the specified set of CPUs.
+static cpumask_var_t bind_writers; // Bind the writers to the specified set of CPUs.
+
+// Parse a cpumask kernel parameter.  If there are more users later on,
+// this might need to got to a more central location.
+static int param_set_cpumask(const char *val, const struct kernel_param *kp)
+{
+	cpumask_var_t *cm_bind = kp->arg;
+	int ret;
+	char *s;
+
+	if (!alloc_cpumask_var(cm_bind, GFP_KERNEL)) {
+		s = "Out of memory";
+		ret = -ENOMEM;
+		goto out_err;
+	}
+	ret = cpulist_parse(val, *cm_bind);
+	if (!ret)
+		return ret;
+	s = "Bad CPU range";
+out_err:
+	pr_warn("%s: %s, all CPUs set\n", kp->name, s);
+	cpumask_setall(*cm_bind);
+	return ret;
+}
+
+// Output a cpumask kernel parameter.
+static int param_get_cpumask(char *buffer, const struct kernel_param *kp)
+{
+	cpumask_var_t *cm_bind = kp->arg;
+
+	return sprintf(buffer, "%*pbl", cpumask_pr_args(*cm_bind));
+}
+
+static bool cpumask_nonempty(cpumask_var_t mask)
+{
+	return cpumask_available(mask) && !cpumask_empty(mask);
+}
+
+static const struct kernel_param_ops lt_bind_ops = {
+	.set = param_set_cpumask,
+	.get = param_get_cpumask,
+};
+
+module_param_cb(bind_readers, &lt_bind_ops, &bind_readers, 0644);
+module_param_cb(bind_writers, &lt_bind_ops, &bind_writers, 0644);
+
+long torture_sched_setaffinity(pid_t pid, const struct cpumask *in_mask);
+
 static struct task_struct *stats_task;
 static struct task_struct **writer_tasks;
 static struct task_struct **reader_tasks;
@@ -69,6 +120,12 @@ struct lock_stress_stats {
 	long n_lock_acquired;
 };
 
+struct call_rcu_chain {
+	struct rcu_head crc_rh;
+	bool crc_stop;
+};
+struct call_rcu_chain *call_rcu_chain;
+
 /* Forward reference. */
 static void lock_torture_cleanup(void);
 
@@ -116,12 +173,9 @@ static int torture_lock_busted_write_lock(int tid __maybe_unused)
 
 static void torture_lock_busted_write_delay(struct torture_random_state *trsp)
 {
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
-
 	/* We want a long delay occasionally to force massive contention.  */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealwriters_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold)))
+		mdelay(long_hold);
 	if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000)))
 		torture_preempt_schedule();  /* Allow test to be preempted. */
 }
@@ -194,15 +248,14 @@ __acquires(torture_spinlock)
 static void torture_spin_lock_write_delay(struct torture_random_state *trsp)
 {
 	const unsigned long shortdelay_us = 2;
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
 	unsigned long j;
 
 	/* We want a short delay mostly to emulate likely code, and
 	 * we want a long delay occasionally to force massive contention.
 	 */
-	if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * longdelay_ms))) {
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold))) {
 		j = jiffies;
-		mdelay(longdelay_ms);
+		mdelay(long_hold);
 		pr_alert("%s: delay = %lu jiffies.\n", __func__, jiffies - j);
 	}
 	if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 200 * shortdelay_us)))
@@ -320,14 +373,12 @@ __acquires(torture_rwlock)
 static void torture_rwlock_write_delay(struct torture_random_state *trsp)
 {
 	const unsigned long shortdelay_us = 2;
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
 
 	/* We want a short delay mostly to emulate likely code, and
 	 * we want a long delay occasionally to force massive contention.
 	 */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealwriters_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold)))
+		mdelay(long_hold);
 	else
 		udelay(shortdelay_us);
 }
@@ -348,14 +399,12 @@ __acquires(torture_rwlock)
 static void torture_rwlock_read_delay(struct torture_random_state *trsp)
 {
 	const unsigned long shortdelay_us = 10;
-	const unsigned long longdelay_ms = 100;
 
 	/* We want a short delay mostly to emulate likely code, and
 	 * we want a long delay occasionally to force massive contention.
 	 */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealreaders_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealreaders_stress * 2000 * long_hold)))
+		mdelay(long_hold);
 	else
 		udelay(shortdelay_us);
 }
@@ -453,12 +502,9 @@ __acquires(torture_mutex)
 
 static void torture_mutex_delay(struct torture_random_state *trsp)
 {
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
-
 	/* We want a long delay occasionally to force massive contention.  */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealwriters_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms * 5);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold)))
+		mdelay(long_hold * 5);
 	if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000)))
 		torture_preempt_schedule();  /* Allow test to be preempted. */
 }
@@ -626,15 +672,13 @@ __acquires(torture_rtmutex)
 static void torture_rtmutex_delay(struct torture_random_state *trsp)
 {
 	const unsigned long shortdelay_us = 2;
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
 
 	/*
 	 * We want a short delay mostly to emulate likely code, and
 	 * we want a long delay occasionally to force massive contention.
 	 */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealwriters_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold)))
+		mdelay(long_hold);
 	if (!(torture_random(trsp) %
 	      (cxt.nrealwriters_stress * 200 * shortdelay_us)))
 		udelay(shortdelay_us);
@@ -691,12 +735,9 @@ __acquires(torture_rwsem)
 
 static void torture_rwsem_write_delay(struct torture_random_state *trsp)
 {
-	const unsigned long longdelay_ms = long_hold ? long_hold : ULONG_MAX;
-
 	/* We want a long delay occasionally to force massive contention.  */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealwriters_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms * 10);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealwriters_stress * 2000 * long_hold)))
+		mdelay(long_hold * 10);
 	if (!(torture_random(trsp) % (cxt.nrealwriters_stress * 20000)))
 		torture_preempt_schedule();  /* Allow test to be preempted. */
 }
@@ -716,14 +757,11 @@ __acquires(torture_rwsem)
 
 static void torture_rwsem_read_delay(struct torture_random_state *trsp)
 {
-	const unsigned long longdelay_ms = 100;
-
 	/* We want a long delay occasionally to force massive contention.  */
-	if (!(torture_random(trsp) %
-	      (cxt.nrealreaders_stress * 2000 * longdelay_ms)))
-		mdelay(longdelay_ms * 2);
+	if (long_hold && !(torture_random(trsp) % (cxt.nrealreaders_stress * 2000 * long_hold)))
+		mdelay(long_hold * 2);
 	else
-		mdelay(longdelay_ms / 2);
+		mdelay(long_hold / 2);
 	if (!(torture_random(trsp) % (cxt.nrealreaders_stress * 20000)))
 		torture_preempt_schedule();  /* Allow test to be preempted. */
 }
@@ -803,11 +841,13 @@ static struct lock_torture_ops percpu_rwsem_lock_ops = {
  */
 static int lock_torture_writer(void *arg)
 {
+	unsigned long j;
+	unsigned long j1;
+	u32 lockset_mask;
 	struct lock_stress_stats *lwsp = arg;
-	int tid = lwsp - cxt.lwsa;
 	DEFINE_TORTURE_RANDOM(rand);
-	u32 lockset_mask;
 	bool skip_main_lock;
+	int tid = lwsp - cxt.lwsa;
 
 	VERBOSE_TOROUT_STRING("lock_torture_writer task started");
 	if (!rt_task(current))
@@ -834,17 +874,24 @@ static int lock_torture_writer(void *arg)
 			cxt.cur_ops->nested_lock(tid, lockset_mask);
 
 		if (!skip_main_lock) {
+			if (acq_writer_lim > 0)
+				j = jiffies;
 			cxt.cur_ops->writelock(tid);
 			if (WARN_ON_ONCE(lock_is_write_held))
 				lwsp->n_lock_fail++;
 			lock_is_write_held = true;
 			if (WARN_ON_ONCE(atomic_read(&lock_is_read_held)))
 				lwsp->n_lock_fail++; /* rare, but... */
-
+			if (acq_writer_lim > 0) {
+				j1 = jiffies;
+				WARN_ONCE(time_after(j1, j + acq_writer_lim),
+					  "%s: Lock acquisition took %lu jiffies.\n",
+					  __func__, j1 - j);
+			}
 			lwsp->n_lock_acquired++;
-		}
-		if (!skip_main_lock) {
+
 			cxt.cur_ops->write_delay(&rand);
+
 			lock_is_write_held = false;
 			WRITE_ONCE(last_lock_release, jiffies);
 			cxt.cur_ops->writeunlock(tid);
@@ -986,16 +1033,69 @@ static int lock_torture_stats(void *arg)
 	return 0;
 }
 
+
 static inline void
 lock_torture_print_module_parms(struct lock_torture_ops *cur_ops,
 				const char *tag)
 {
+	static cpumask_t cpumask_all;
+	cpumask_t *rcmp = cpumask_nonempty(bind_readers) ? bind_readers : &cpumask_all;
+	cpumask_t *wcmp = cpumask_nonempty(bind_writers) ? bind_writers : &cpumask_all;
+
+	cpumask_setall(&cpumask_all);
 	pr_alert("%s" TORTURE_FLAG
-		 "--- %s%s: nwriters_stress=%d nreaders_stress=%d nested_locks=%d stat_interval=%d verbose=%d shuffle_interval=%d stutter=%d shutdown_secs=%d onoff_interval=%d onoff_holdoff=%d\n",
+		 "--- %s%s: acq_writer_lim=%d bind_readers=%*pbl bind_writers=%*pbl call_rcu_chains=%d long_hold=%d nested_locks=%d nreaders_stress=%d nwriters_stress=%d onoff_holdoff=%d onoff_interval=%d rt_boost=%d rt_boost_factor=%d shuffle_interval=%d shutdown_secs=%d stat_interval=%d stutter=%d verbose=%d writer_fifo=%d\n",
 		 torture_type, tag, cxt.debug_lock ? " [debug]": "",
-		 cxt.nrealwriters_stress, cxt.nrealreaders_stress,
-		 nested_locks, stat_interval, verbose, shuffle_interval,
-		 stutter, shutdown_secs, onoff_interval, onoff_holdoff);
+		 acq_writer_lim, cpumask_pr_args(rcmp), cpumask_pr_args(wcmp),
+		 call_rcu_chains, long_hold, nested_locks, cxt.nrealreaders_stress,
+		 cxt.nrealwriters_stress, onoff_holdoff, onoff_interval, rt_boost,
+		 rt_boost_factor, shuffle_interval, shutdown_secs, stat_interval, stutter,
+		 verbose, writer_fifo);
+}
+
+// If requested, maintain call_rcu() chains to keep a grace period always
+// in flight.  These increase the probability of getting an RCU CPU stall
+// warning and associated diagnostics when a locking primitive stalls.
+
+static void call_rcu_chain_cb(struct rcu_head *rhp)
+{
+	struct call_rcu_chain *crcp = container_of(rhp, struct call_rcu_chain, crc_rh);
+
+	if (!smp_load_acquire(&crcp->crc_stop)) {
+		(void)start_poll_synchronize_rcu(); // Start one grace period...
+		call_rcu(&crcp->crc_rh, call_rcu_chain_cb); // ... and later start another.
+	}
+}
+
+// Start the requested number of call_rcu() chains.
+static int call_rcu_chain_init(void)
+{
+	int i;
+
+	if (call_rcu_chains <= 0)
+		return 0;
+	call_rcu_chain = kcalloc(call_rcu_chains, sizeof(*call_rcu_chain), GFP_KERNEL);
+	if (!call_rcu_chain)
+		return -ENOMEM;
+	for (i = 0; i < call_rcu_chains; i++) {
+		call_rcu_chain[i].crc_stop = false;
+		call_rcu(&call_rcu_chain[i].crc_rh, call_rcu_chain_cb);
+	}
+	return 0;
+}
+
+// Stop all of the call_rcu() chains.
+static void call_rcu_chain_cleanup(void)
+{
+	int i;
+
+	if (!call_rcu_chain)
+		return;
+	for (i = 0; i < call_rcu_chains; i++)
+		smp_store_release(&call_rcu_chain[i].crc_stop, true);
+	rcu_barrier();
+	kfree(call_rcu_chain);
+	call_rcu_chain = NULL;
 }
 
 static void lock_torture_cleanup(void)
@@ -1048,6 +1148,8 @@ static void lock_torture_cleanup(void)
 	kfree(cxt.lrsa);
 	cxt.lrsa = NULL;
 
+	call_rcu_chain_cleanup();
+
 end:
 	if (cxt.init_called) {
 		if (cxt.cur_ops->exit)
@@ -1177,6 +1279,10 @@ static int __init lock_torture_init(void)
 		}
 	}
 
+	firsterr = call_rcu_chain_init();
+	if (torture_init_error(firsterr))
+		goto unwind;
+
 	lock_torture_print_module_parms(cxt.cur_ops, "Start of test");
 
 	/* Prepare torture context. */
@@ -1250,6 +1356,8 @@ static int __init lock_torture_init(void)
 						     writer_fifo ? sched_set_fifo : NULL);
 		if (torture_init_error(firsterr))
 			goto unwind;
+		if (cpumask_nonempty(bind_writers))
+			torture_sched_setaffinity(writer_tasks[i]->pid, bind_writers);
 
 	create_reader:
 		if (cxt.cur_ops->readlock == NULL || (j >= cxt.nrealreaders_stress))
@@ -1259,6 +1367,8 @@ static int __init lock_torture_init(void)
 						  reader_tasks[j]);
 		if (torture_init_error(firsterr))
 			goto unwind;
+		if (cpumask_nonempty(bind_readers))
+			torture_sched_setaffinity(reader_tasks[j]->pid, bind_readers);
 	}
 	if (stat_interval > 0) {
 		firsterr = torture_create_kthread(lock_torture_stats, NULL,
diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
index d973fe604..2deeeca3e 100644
--- a/kernel/locking/mutex.c
+++ b/kernel/locking/mutex.c
@@ -1126,6 +1126,9 @@ EXPORT_SYMBOL(ww_mutex_lock_interruptible);
 #endif /* !CONFIG_DEBUG_LOCK_ALLOC */
 #endif /* !CONFIG_PREEMPT_RT */
 
+EXPORT_TRACEPOINT_SYMBOL_GPL(contention_begin);
+EXPORT_TRACEPOINT_SYMBOL_GPL(contention_end);
+
 /**
  * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0
  * @cnt: the atomic which we are to dec
diff --git a/kernel/locking/osq_lock.c b/kernel/locking/osq_lock.c
index d5610ad52..75a6f6133 100644
--- a/kernel/locking/osq_lock.c
+++ b/kernel/locking/osq_lock.c
@@ -11,6 +11,13 @@
  * called from interrupt context and we have preemption disabled while
  * spinning.
  */
+
+struct optimistic_spin_node {
+	struct optimistic_spin_node *next, *prev;
+	int locked; /* 1 if lock acquired */
+	int cpu; /* encoded CPU # + 1 value */
+};
+
 static DEFINE_PER_CPU_SHARED_ALIGNED(struct optimistic_spin_node, osq_node);
 
 /*
@@ -37,32 +44,28 @@ static inline struct optimistic_spin_node *decode_cpu(int encoded_cpu_val)
 /*
  * Get a stable @node->next pointer, either for unlock() or unqueue() purposes.
  * Can return NULL in case we were the last queued and we updated @lock instead.
+ *
+ * If osq_lock() is being cancelled there must be a previous node
+ * and 'old_cpu' is its CPU #.
+ * For osq_unlock() there is never a previous node and old_cpu is
+ * set to OSQ_UNLOCKED_VAL.
  */
 static inline struct optimistic_spin_node *
 osq_wait_next(struct optimistic_spin_queue *lock,
 	      struct optimistic_spin_node *node,
-	      struct optimistic_spin_node *prev)
+	      int old_cpu)
 {
-	struct optimistic_spin_node *next = NULL;
 	int curr = encode_cpu(smp_processor_id());
-	int old;
-
-	/*
-	 * If there is a prev node in queue, then the 'old' value will be
-	 * the prev node's CPU #, else it's set to OSQ_UNLOCKED_VAL since if
-	 * we're currently last in queue, then the queue will then become empty.
-	 */
-	old = prev ? prev->cpu : OSQ_UNLOCKED_VAL;
 
 	for (;;) {
 		if (atomic_read(&lock->tail) == curr &&
-		    atomic_cmpxchg_acquire(&lock->tail, curr, old) == curr) {
+		    atomic_cmpxchg_acquire(&lock->tail, curr, old_cpu) == curr) {
 			/*
 			 * We were the last queued, we moved @lock back. @prev
 			 * will now observe @lock and will complete its
 			 * unlock()/unqueue().
 			 */
-			break;
+			return NULL;
 		}
 
 		/*
@@ -76,15 +79,15 @@ osq_wait_next(struct optimistic_spin_queue *lock,
 		 * wait for a new @node->next from its Step-C.
 		 */
 		if (node->next) {
+			struct optimistic_spin_node *next;
+
 			next = xchg(&node->next, NULL);
 			if (next)
-				break;
+				return next;
 		}
 
 		cpu_relax();
 	}
-
-	return next;
 }
 
 bool osq_lock(struct optimistic_spin_queue *lock)
@@ -186,7 +189,7 @@ bool osq_lock(struct optimistic_spin_queue *lock)
 	 * back to @prev.
 	 */
 
-	next = osq_wait_next(lock, node, prev);
+	next = osq_wait_next(lock, node, prev->cpu);
 	if (!next)
 		return false;
 
@@ -226,7 +229,7 @@ void osq_unlock(struct optimistic_spin_queue *lock)
 		return;
 	}
 
-	next = osq_wait_next(lock, node, NULL);
+	next = osq_wait_next(lock, node, OSQ_UNLOCKED_VAL);
 	if (next)
 		WRITE_ONCE(next->locked, 1);
 }
diff --git a/kernel/locking/rtmutex.c b/kernel/locking/rtmutex.c
index 21db0df0e..4a10e8c16 100644
--- a/kernel/locking/rtmutex.c
+++ b/kernel/locking/rtmutex.c
@@ -218,6 +218,11 @@ static __always_inline bool rt_mutex_cmpxchg_acquire(struct rt_mutex_base *lock,
 	return try_cmpxchg_acquire(&lock->owner, &old, new);
 }
 
+static __always_inline bool rt_mutex_try_acquire(struct rt_mutex_base *lock)
+{
+	return rt_mutex_cmpxchg_acquire(lock, NULL, current);
+}
+
 static __always_inline bool rt_mutex_cmpxchg_release(struct rt_mutex_base *lock,
 						     struct task_struct *old,
 						     struct task_struct *new)
@@ -297,6 +302,20 @@ static __always_inline bool rt_mutex_cmpxchg_acquire(struct rt_mutex_base *lock,
 
 }
 
+static int __sched rt_mutex_slowtrylock(struct rt_mutex_base *lock);
+
+static __always_inline bool rt_mutex_try_acquire(struct rt_mutex_base *lock)
+{
+	/*
+	 * With debug enabled rt_mutex_cmpxchg trylock() will always fail.
+	 *
+	 * Avoid unconditionally taking the slow path by using
+	 * rt_mutex_slow_trylock() which is covered by the debug code and can
+	 * acquire a non-contended rtmutex.
+	 */
+	return rt_mutex_slowtrylock(lock);
+}
+
 static __always_inline bool rt_mutex_cmpxchg_release(struct rt_mutex_base *lock,
 						     struct task_struct *old,
 						     struct task_struct *new)
@@ -1613,7 +1632,7 @@ static int __sched rt_mutex_slowlock_block(struct rt_mutex_base *lock,
 		raw_spin_unlock_irq(&lock->wait_lock);
 
 		if (!owner || !rtmutex_spin_on_owner(lock, waiter, owner))
-			schedule();
+			rt_mutex_schedule();
 
 		raw_spin_lock_irq(&lock->wait_lock);
 		set_current_state(state);
@@ -1642,7 +1661,7 @@ static void __sched rt_mutex_handle_deadlock(int res, int detect_deadlock,
 	WARN(1, "rtmutex deadlock detected\n");
 	while (1) {
 		set_current_state(TASK_INTERRUPTIBLE);
-		schedule();
+		rt_mutex_schedule();
 	}
 }
 
@@ -1738,6 +1757,15 @@ static int __sched rt_mutex_slowlock(struct rt_mutex_base *lock,
 	int ret;
 
 	/*
+	 * Do all pre-schedule work here, before we queue a waiter and invoke
+	 * PI -- any such work that trips on rtlock (PREEMPT_RT spinlock) would
+	 * otherwise recurse back into task_blocks_on_rt_mutex() through
+	 * rtlock_slowlock() and will then enqueue a second waiter for this
+	 * same task and things get really confusing real fast.
+	 */
+	rt_mutex_pre_schedule();
+
+	/*
 	 * Technically we could use raw_spin_[un]lock_irq() here, but this can
 	 * be called in early boot if the cmpxchg() fast path is disabled
 	 * (debug, no architecture support). In this case we will acquire the
@@ -1748,6 +1776,7 @@ static int __sched rt_mutex_slowlock(struct rt_mutex_base *lock,
 	raw_spin_lock_irqsave(&lock->wait_lock, flags);
 	ret = __rt_mutex_slowlock_locked(lock, ww_ctx, state);
 	raw_spin_unlock_irqrestore(&lock->wait_lock, flags);
+	rt_mutex_post_schedule();
 
 	return ret;
 }
@@ -1755,7 +1784,9 @@ static int __sched rt_mutex_slowlock(struct rt_mutex_base *lock,
 static __always_inline int __rt_mutex_lock(struct rt_mutex_base *lock,
 					   unsigned int state)
 {
-	if (likely(rt_mutex_cmpxchg_acquire(lock, NULL, current)))
+	lockdep_assert(!current->pi_blocked_on);
+
+	if (likely(rt_mutex_try_acquire(lock)))
 		return 0;
 
 	return rt_mutex_slowlock(lock, NULL, state);
diff --git a/kernel/locking/rwbase_rt.c b/kernel/locking/rwbase_rt.c
index 25ec02394..34a59569d 100644
--- a/kernel/locking/rwbase_rt.c
+++ b/kernel/locking/rwbase_rt.c
@@ -71,6 +71,7 @@ static int __sched __rwbase_read_lock(struct rwbase_rt *rwb,
 	struct rt_mutex_base *rtm = &rwb->rtmutex;
 	int ret;
 
+	rwbase_pre_schedule();
 	raw_spin_lock_irq(&rtm->wait_lock);
 
 	/*
@@ -125,12 +126,15 @@ static int __sched __rwbase_read_lock(struct rwbase_rt *rwb,
 		rwbase_rtmutex_unlock(rtm);
 
 	trace_contention_end(rwb, ret);
+	rwbase_post_schedule();
 	return ret;
 }
 
 static __always_inline int rwbase_read_lock(struct rwbase_rt *rwb,
 					    unsigned int state)
 {
+	lockdep_assert(!current->pi_blocked_on);
+
 	if (rwbase_read_trylock(rwb))
 		return 0;
 
@@ -237,6 +241,8 @@ static int __sched rwbase_write_lock(struct rwbase_rt *rwb,
 	/* Force readers into slow path */
 	atomic_sub(READER_BIAS, &rwb->readers);
 
+	rwbase_pre_schedule();
+
 	raw_spin_lock_irqsave(&rtm->wait_lock, flags);
 	if (__rwbase_write_trylock(rwb))
 		goto out_unlock;
@@ -248,6 +254,7 @@ static int __sched rwbase_write_lock(struct rwbase_rt *rwb,
 		if (rwbase_signal_pending_state(state, current)) {
 			rwbase_restore_current_state();
 			__rwbase_write_unlock(rwb, 0, flags);
+			rwbase_post_schedule();
 			trace_contention_end(rwb, -EINTR);
 			return -EINTR;
 		}
@@ -266,6 +273,7 @@ static int __sched rwbase_write_lock(struct rwbase_rt *rwb,
 
 out_unlock:
 	raw_spin_unlock_irqrestore(&rtm->wait_lock, flags);
+	rwbase_post_schedule();
 	return 0;
 }
 
diff --git a/kernel/locking/rwsem.c b/kernel/locking/rwsem.c
index 9eabd585c..2340b6d90 100644
--- a/kernel/locking/rwsem.c
+++ b/kernel/locking/rwsem.c
@@ -1427,8 +1427,14 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
 #define rwbase_signal_pending_state(state, current)	\
 	signal_pending_state(state, current)
 
+#define rwbase_pre_schedule()				\
+	rt_mutex_pre_schedule()
+
 #define rwbase_schedule()				\
-	schedule()
+	rt_mutex_schedule()
+
+#define rwbase_post_schedule()				\
+	rt_mutex_post_schedule()
 
 #include "rwbase_rt.c"
 
diff --git a/kernel/locking/spinlock_rt.c b/kernel/locking/spinlock_rt.c
index 48a19ed84..38e292454 100644
--- a/kernel/locking/spinlock_rt.c
+++ b/kernel/locking/spinlock_rt.c
@@ -37,6 +37,8 @@
 
 static __always_inline void rtlock_lock(struct rt_mutex_base *rtm)
 {
+	lockdep_assert(!current->pi_blocked_on);
+
 	if (unlikely(!rt_mutex_cmpxchg_acquire(rtm, NULL, current)))
 		rtlock_slowlock(rtm);
 }
@@ -184,9 +186,13 @@ static __always_inline int  rwbase_rtmutex_trylock(struct rt_mutex_base *rtm)
 
 #define rwbase_signal_pending_state(state, current)	(0)
 
+#define rwbase_pre_schedule()
+
 #define rwbase_schedule()				\
 	schedule_rtlock()
 
+#define rwbase_post_schedule()
+
 #include "rwbase_rt.c"
 /*
  * The common functions which get wrapped into the rwlock API.
diff --git a/kernel/locking/test-ww_mutex.c b/kernel/locking/test-ww_mutex.c
index 7c5a8f054..78719e1ef 100644
--- a/kernel/locking/test-ww_mutex.c
+++ b/kernel/locking/test-ww_mutex.c
@@ -9,7 +9,7 @@
 #include <linux/delay.h>
 #include <linux/kthread.h>
 #include <linux/module.h>
-#include <linux/random.h>
+#include <linux/prandom.h>
 #include <linux/slab.h>
 #include <linux/ww_mutex.h>
 
@@ -386,6 +386,19 @@ struct stress {
 	int nlocks;
 };
 
+struct rnd_state rng;
+DEFINE_SPINLOCK(rng_lock);
+
+static inline u32 prandom_u32_below(u32 ceil)
+{
+	u32 ret;
+
+	spin_lock(&rng_lock);
+	ret = prandom_u32_state(&rng) % ceil;
+	spin_unlock(&rng_lock);
+	return ret;
+}
+
 static int *get_random_order(int count)
 {
 	int *order;
@@ -399,7 +412,7 @@ static int *get_random_order(int count)
 		order[n] = n;
 
 	for (n = count - 1; n > 1; n--) {
-		r = get_random_u32_below(n + 1);
+		r = prandom_u32_below(n + 1);
 		if (r != n) {
 			tmp = order[n];
 			order[n] = order[r];
@@ -452,17 +465,18 @@ retry:
 			ww_mutex_unlock(&locks[order[n]]);
 
 		if (err == -EDEADLK) {
-			ww_mutex_lock_slow(&locks[order[contended]], &ctx);
-			goto retry;
+			if (!time_after(jiffies, stress->timeout)) {
+				ww_mutex_lock_slow(&locks[order[contended]], &ctx);
+				goto retry;
+			}
 		}
 
+		ww_acquire_fini(&ctx);
 		if (err) {
 			pr_err_once("stress (%s) failed with %d\n",
 				    __func__, err);
 			break;
 		}
-
-		ww_acquire_fini(&ctx);
 	} while (!time_after(jiffies, stress->timeout));
 
 	kfree(order);
@@ -629,6 +643,8 @@ static int __init test_ww_mutex_init(void)
 
 	printk(KERN_INFO "Beginning ww mutex selftests\n");
 
+	prandom_seed_state(&rng, get_random_u64());
+
 	wq = alloc_workqueue("test-ww_mutex", WQ_UNBOUND, 0);
 	if (!wq)
 		return -ENOMEM;
diff --git a/kernel/locking/ww_rt_mutex.c b/kernel/locking/ww_rt_mutex.c
index d1473c624..c7196de83 100644
--- a/kernel/locking/ww_rt_mutex.c
+++ b/kernel/locking/ww_rt_mutex.c
@@ -62,7 +62,7 @@ __ww_rt_mutex_lock(struct ww_mutex *lock, struct ww_acquire_ctx *ww_ctx,
 	}
 	mutex_acquire_nest(&rtm->dep_map, 0, 0, nest_lock, ip);
 
-	if (likely(rt_mutex_cmpxchg_acquire(&rtm->rtmutex, NULL, current))) {
+	if (likely(rt_mutex_try_acquire(&rtm->rtmutex))) {
 		if (ww_ctx)
 			ww_mutex_set_context_fastpath(lock, ww_ctx);
 		return 0;