diff --git a/include/linux/rcupdate_trace.h b/include/linux/rcupdate_trace.h
index 86c8f6c98412..6f9c35817398 100644
--- a/include/linux/rcupdate_trace.h
+++ b/include/linux/rcupdate_trace.h
@@ -31,7 +31,7 @@ static inline int rcu_read_lock_trace_held(void)
 
 #ifdef CONFIG_TASKS_TRACE_RCU
 
-void rcu_read_unlock_trace_special(struct task_struct *t, int nesting);
+void rcu_read_unlock_trace_special(struct task_struct *t);
 
 /**
  * rcu_read_lock_trace - mark beginning of RCU-trace read-side critical section
@@ -80,7 +80,8 @@ static inline void rcu_read_unlock_trace(void)
 		WRITE_ONCE(t->trc_reader_nesting, nesting);
 		return;  // We assume shallow reader nesting.
 	}
-	rcu_read_unlock_trace_special(t, nesting);
+	WARN_ON_ONCE(nesting != 0);
+	rcu_read_unlock_trace_special(t);
 }
 
 void call_rcu_tasks_trace(struct rcu_head *rhp, rcu_callback_t func);
diff --git a/include/linux/torture.h b/include/linux/torture.h
index 0910c5803f35..24f58e50a94b 100644
--- a/include/linux/torture.h
+++ b/include/linux/torture.h
@@ -47,6 +47,14 @@ do {										\
 } while (0)
 void verbose_torout_sleep(void);
 
+#define torture_init_error(firsterr)						\
+({										\
+	int ___firsterr = (firsterr);						\
+										\
+	WARN_ONCE(!IS_MODULE(CONFIG_RCU_TORTURE_TEST) && ___firsterr < 0, "Torture-test initialization failed with error code %d\n", ___firsterr); \
+	___firsterr < 0;								\
+})
+
 /* Definitions for online/offline exerciser. */
 #ifdef CONFIG_HOTPLUG_CPU
 int torture_num_online_cpus(void);
diff --git a/kernel/locking/locktorture.c b/kernel/locking/locktorture.c
index 7c5a4a087cc7..397ac13d2ef7 100644
--- a/kernel/locking/locktorture.c
+++ b/kernel/locking/locktorture.c
@@ -1022,23 +1022,23 @@ static int __init lock_torture_init(void)
 	if (onoff_interval > 0) {
 		firsterr = torture_onoff_init(onoff_holdoff * HZ,
 					      onoff_interval * HZ, NULL);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (shuffle_interval > 0) {
 		firsterr = torture_shuffle_init(shuffle_interval);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (shutdown_secs > 0) {
 		firsterr = torture_shutdown_init(shutdown_secs,
 						 lock_torture_cleanup);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stutter > 0) {
 		firsterr = torture_stutter_init(stutter, stutter);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 
@@ -1082,7 +1082,7 @@ static int __init lock_torture_init(void)
 		/* Create writer. */
 		firsterr = torture_create_kthread(lock_torture_writer, &cxt.lwsa[i],
 						  writer_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 
 	create_reader:
@@ -1091,13 +1091,13 @@ static int __init lock_torture_init(void)
 		/* Create reader. */
 		firsterr = torture_create_kthread(lock_torture_reader, &cxt.lrsa[j],
 						  reader_tasks[j]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stat_interval > 0) {
 		firsterr = torture_create_kthread(lock_torture_stats, NULL,
 						  stats_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	torture_init_end();
diff --git a/kernel/rcu/rcuscale.c b/kernel/rcu/rcuscale.c
index 2cc34a22a506..228f143bf935 100644
--- a/kernel/rcu/rcuscale.c
+++ b/kernel/rcu/rcuscale.c
@@ -758,7 +758,7 @@ kfree_scale_init(void)
 		init_waitqueue_head(&shutdown_wq);
 		firsterr = torture_create_kthread(kfree_scale_shutdown, NULL,
 						  shutdown_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 		schedule_timeout_uninterruptible(1);
 	}
@@ -775,7 +775,7 @@ kfree_scale_init(void)
 	for (i = 0; i < kfree_nrealthreads; i++) {
 		firsterr = torture_create_kthread(kfree_scale_thread, (void *)i,
 						  kfree_reader_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 
@@ -838,7 +838,7 @@ rcu_scale_init(void)
 		init_waitqueue_head(&shutdown_wq);
 		firsterr = torture_create_kthread(rcu_scale_shutdown, NULL,
 						  shutdown_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 		schedule_timeout_uninterruptible(1);
 	}
@@ -852,7 +852,7 @@ rcu_scale_init(void)
 	for (i = 0; i < nrealreaders; i++) {
 		firsterr = torture_create_kthread(rcu_scale_reader, (void *)i,
 						  reader_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	while (atomic_read(&n_rcu_scale_reader_started) < nrealreaders)
@@ -879,7 +879,7 @@ rcu_scale_init(void)
 		}
 		firsterr = torture_create_kthread(rcu_scale_writer, (void *)i,
 						  writer_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	torture_init_end();
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index f640cbd9262c..8b410d982990 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -1432,28 +1432,34 @@ static void rcutorture_one_extend(int *readstate, int newstate,
 	/* First, put new protection in place to avoid critical-section gap. */
 	if (statesnew & RCUTORTURE_RDR_BH)
 		local_bh_disable();
+	if (statesnew & RCUTORTURE_RDR_RBH)
+		rcu_read_lock_bh();
 	if (statesnew & RCUTORTURE_RDR_IRQ)
 		local_irq_disable();
 	if (statesnew & RCUTORTURE_RDR_PREEMPT)
 		preempt_disable();
-	if (statesnew & RCUTORTURE_RDR_RBH)
-		rcu_read_lock_bh();
 	if (statesnew & RCUTORTURE_RDR_SCHED)
 		rcu_read_lock_sched();
 	if (statesnew & RCUTORTURE_RDR_RCU)
 		idxnew = cur_ops->readlock() << RCUTORTURE_RDR_SHIFT;
 
-	/* Next, remove old protection, irq first due to bh conflict. */
+	/*
+	 * Next, remove old protection, in decreasing order of strength
+	 * to avoid unlock paths that aren't safe in the stronger
+	 * context. Namely: BH can not be enabled with disabled interrupts.
+	 * Additionally PREEMPT_RT requires that BH is enabled in preemptible
+	 * context.
+	 */
 	if (statesold & RCUTORTURE_RDR_IRQ)
 		local_irq_enable();
-	if (statesold & RCUTORTURE_RDR_BH)
-		local_bh_enable();
 	if (statesold & RCUTORTURE_RDR_PREEMPT)
 		preempt_enable();
-	if (statesold & RCUTORTURE_RDR_RBH)
-		rcu_read_unlock_bh();
 	if (statesold & RCUTORTURE_RDR_SCHED)
 		rcu_read_unlock_sched();
+	if (statesold & RCUTORTURE_RDR_BH)
+		local_bh_enable();
+	if (statesold & RCUTORTURE_RDR_RBH)
+		rcu_read_unlock_bh();
 	if (statesold & RCUTORTURE_RDR_RCU) {
 		bool lockit = !statesnew && !(torture_random(trsp) & 0xffff);
 
@@ -1496,6 +1502,9 @@ rcutorture_extend_mask(int oldmask, struct torture_random_state *trsp)
 	int mask = rcutorture_extend_mask_max();
 	unsigned long randmask1 = torture_random(trsp) >> 8;
 	unsigned long randmask2 = randmask1 >> 3;
+	unsigned long preempts = RCUTORTURE_RDR_PREEMPT | RCUTORTURE_RDR_SCHED;
+	unsigned long preempts_irq = preempts | RCUTORTURE_RDR_IRQ;
+	unsigned long bhs = RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH;
 
 	WARN_ON_ONCE(mask >> RCUTORTURE_RDR_SHIFT);
 	/* Mostly only one bit (need preemption!), sometimes lots of bits. */
@@ -1503,11 +1512,26 @@ rcutorture_extend_mask(int oldmask, struct torture_random_state *trsp)
 		mask = mask & randmask2;
 	else
 		mask = mask & (1 << (randmask2 % RCUTORTURE_RDR_NBITS));
-	/* Can't enable bh w/irq disabled. */
-	if ((mask & RCUTORTURE_RDR_IRQ) &&
-	    ((!(mask & RCUTORTURE_RDR_BH) && (oldmask & RCUTORTURE_RDR_BH)) ||
-	     (!(mask & RCUTORTURE_RDR_RBH) && (oldmask & RCUTORTURE_RDR_RBH))))
-		mask |= RCUTORTURE_RDR_BH | RCUTORTURE_RDR_RBH;
+
+	/*
+	 * Can't enable bh w/irq disabled.
+	 */
+	if (mask & RCUTORTURE_RDR_IRQ)
+		mask |= oldmask & bhs;
+
+	/*
+	 * Ideally these sequences would be detected in debug builds
+	 * (regardless of RT), but until then don't stop testing
+	 * them on non-RT.
+	 */
+	if (IS_ENABLED(CONFIG_PREEMPT_RT)) {
+		/* Can't modify BH in atomic context */
+		if (oldmask & preempts_irq)
+			mask &= ~bhs;
+		if ((oldmask | mask) & preempts_irq)
+			mask |= oldmask & bhs;
+	}
+
 	return mask ?: RCUTORTURE_RDR_RCU;
 }
 
@@ -2741,7 +2765,7 @@ static int rcu_torture_read_exit(void *unused)
 static int rcu_torture_read_exit_init(void)
 {
 	if (read_exit_burst <= 0)
-		return -EINVAL;
+		return 0;
 	init_waitqueue_head(&read_exit_wq);
 	read_exit_child_stop = false;
 	read_exit_child_stopped = false;
@@ -2819,7 +2843,7 @@ rcu_torture_cleanup(void)
 		 rcutorture_seq_diff(gp_seq, start_gp_seq));
 	torture_stop_kthread(rcu_torture_stats, stats_task);
 	torture_stop_kthread(rcu_torture_fqs, fqs_task);
-	if (rcu_torture_can_boost())
+	if (rcu_torture_can_boost() && rcutor_hp >= 0)
 		cpuhp_remove_state(rcutor_hp);
 
 	/*
@@ -3037,7 +3061,7 @@ rcu_torture_init(void)
 	rcu_torture_write_types();
 	firsterr = torture_create_kthread(rcu_torture_writer, NULL,
 					  writer_task);
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	if (nfakewriters > 0) {
 		fakewriter_tasks = kcalloc(nfakewriters,
@@ -3052,7 +3076,7 @@ rcu_torture_init(void)
 	for (i = 0; i < nfakewriters; i++) {
 		firsterr = torture_create_kthread(rcu_torture_fakewriter,
 						  NULL, fakewriter_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	reader_tasks = kcalloc(nrealreaders, sizeof(reader_tasks[0]),
@@ -3068,7 +3092,7 @@ rcu_torture_init(void)
 		rcu_torture_reader_mbchk[i].rtc_chkrdr = -1;
 		firsterr = torture_create_kthread(rcu_torture_reader, (void *)i,
 						  reader_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	nrealnocbers = nocbs_nthreads;
@@ -3088,18 +3112,18 @@ rcu_torture_init(void)
 	}
 	for (i = 0; i < nrealnocbers; i++) {
 		firsterr = torture_create_kthread(rcu_nocb_toggle, NULL, nocb_tasks[i]);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stat_interval > 0) {
 		firsterr = torture_create_kthread(rcu_torture_stats, NULL,
 						  stats_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (test_no_idle_hz && shuffle_interval > 0) {
 		firsterr = torture_shuffle_init(shuffle_interval * HZ);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stutter < 0)
@@ -3109,7 +3133,7 @@ rcu_torture_init(void)
 
 		t = cur_ops->stall_dur ? cur_ops->stall_dur() : stutter * HZ;
 		firsterr = torture_stutter_init(stutter * HZ, t);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (fqs_duration < 0)
@@ -3118,7 +3142,7 @@ rcu_torture_init(void)
 		/* Create the fqs thread */
 		firsterr = torture_create_kthread(rcu_torture_fqs, NULL,
 						  fqs_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (test_boost_interval < 1)
@@ -3132,9 +3156,9 @@ rcu_torture_init(void)
 		firsterr = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "RCU_TORTURE",
 					     rcutorture_booster_init,
 					     rcutorture_booster_cleanup);
-		if (firsterr < 0)
-			goto unwind;
 		rcutor_hp = firsterr;
+		if (torture_init_error(firsterr))
+			goto unwind;
 
 		// Testing RCU priority boosting requires rcutorture do
 		// some serious abuse.  Counter this by running ksoftirqd
@@ -3153,23 +3177,23 @@ rcu_torture_init(void)
 	}
 	shutdown_jiffies = jiffies + shutdown_secs * HZ;
 	firsterr = torture_shutdown_init(shutdown_secs, rcu_torture_cleanup);
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	firsterr = torture_onoff_init(onoff_holdoff * HZ, onoff_interval,
 				      rcutorture_sync);
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	firsterr = rcu_torture_stall_init();
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	firsterr = rcu_torture_fwd_prog_init();
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	firsterr = rcu_torture_barrier_init();
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	firsterr = rcu_torture_read_exit_init();
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 	if (object_debug)
 		rcu_test_debug_objects();
diff --git a/kernel/rcu/refscale.c b/kernel/rcu/refscale.c
index 66dc14cf5687..1631ef8a138d 100644
--- a/kernel/rcu/refscale.c
+++ b/kernel/rcu/refscale.c
@@ -824,7 +824,7 @@ ref_scale_init(void)
 		init_waitqueue_head(&shutdown_wq);
 		firsterr = torture_create_kthread(ref_scale_shutdown, NULL,
 						  shutdown_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 		schedule_timeout_uninterruptible(1);
 	}
@@ -851,7 +851,7 @@ ref_scale_init(void)
 	for (i = 0; i < nreaders; i++) {
 		firsterr = torture_create_kthread(ref_scale_reader, (void *)i,
 						  reader_tasks[i].task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 
 		init_waitqueue_head(&(reader_tasks[i].wq));
@@ -860,7 +860,7 @@ ref_scale_init(void)
 	// Main Task
 	init_waitqueue_head(&main_wq);
 	firsterr = torture_create_kthread(main_func, NULL, main_task);
-	if (firsterr)
+	if (torture_init_error(firsterr))
 		goto unwind;
 
 	torture_init_end();
diff --git a/kernel/rcu/tasks.h b/kernel/rcu/tasks.h
index 806160c44b17..66e7586a33e9 100644
--- a/kernel/rcu/tasks.h
+++ b/kernel/rcu/tasks.h
@@ -197,6 +197,7 @@ static int __noreturn rcu_tasks_kthread(void *arg)
 	 * This loop is terminated by the system going down.  ;-)
 	 */
 	for (;;) {
+		set_tasks_gp_state(rtp, RTGS_WAIT_CBS);
 
 		/* Pick up any new callbacks. */
 		raw_spin_lock_irqsave(&rtp->cbs_lock, flags);
@@ -236,8 +237,6 @@ static int __noreturn rcu_tasks_kthread(void *arg)
 		}
 		/* Paranoid sleep to keep this from entering a tight loop */
 		schedule_timeout_idle(rtp->gp_sleep);
-
-		set_tasks_gp_state(rtp, RTGS_WAIT_CBS);
 	}
 }
 
@@ -369,7 +368,7 @@ static void rcu_tasks_wait_gp(struct rcu_tasks *rtp)
 ////////////////////////////////////////////////////////////////////////
 //
 // Simple variant of RCU whose quiescent states are voluntary context
-// switch, cond_resched_rcu_qs(), user-space execution, and idle.
+// switch, cond_resched_tasks_rcu_qs(), user-space execution, and idle.
 // As such, grace periods can take one good long time.  There are no
 // read-side primitives similar to rcu_read_lock() and rcu_read_unlock()
 // because this implementation is intended to get the system into a safe
@@ -540,7 +539,7 @@ DEFINE_RCU_TASKS(rcu_tasks, rcu_tasks_wait_gp, call_rcu_tasks, "RCU Tasks");
  * period elapses, in other words after all currently executing RCU
  * read-side critical sections have completed. call_rcu_tasks() assumes
  * that the read-side critical sections end at a voluntary context
- * switch (not a preemption!), cond_resched_rcu_qs(), entry into idle,
+ * switch (not a preemption!), cond_resched_tasks_rcu_qs(), entry into idle,
  * or transition to usermode execution.  As such, there are no read-side
  * primitives analogous to rcu_read_lock() and rcu_read_unlock() because
  * this primitive is intended to determine that all tasks have passed
@@ -678,11 +677,11 @@ DEFINE_RCU_TASKS(rcu_tasks_rude, rcu_tasks_rude_wait_gp, call_rcu_tasks_rude,
  * period elapses, in other words after all currently executing RCU
  * read-side critical sections have completed. call_rcu_tasks_rude()
  * assumes that the read-side critical sections end at context switch,
- * cond_resched_rcu_qs(), or transition to usermode execution.  As such,
- * there are no read-side primitives analogous to rcu_read_lock() and
- * rcu_read_unlock() because this primitive is intended to determine
- * that all tasks have passed through a safe state, not so much for
- * data-structure synchronization.
+ * cond_resched_tasks_rcu_qs(), or transition to usermode execution (as
+ * usermode execution is schedulable). As such, there are no read-side
+ * primitives analogous to rcu_read_lock() and rcu_read_unlock() because
+ * this primitive is intended to determine that all tasks have passed
+ * through a safe state, not so much for data-structure synchronization.
  *
  * See the description of call_rcu() for more detailed information on
  * memory ordering guarantees.
@@ -700,8 +699,8 @@ EXPORT_SYMBOL_GPL(call_rcu_tasks_rude);
  * grace period has elapsed, in other words after all currently
  * executing rcu-tasks read-side critical sections have elapsed.  These
  * read-side critical sections are delimited by calls to schedule(),
- * cond_resched_tasks_rcu_qs(), userspace execution, and (in theory,
- * anyway) cond_resched().
+ * cond_resched_tasks_rcu_qs(), userspace execution (which is a schedulable
+ * context), and (in theory, anyway) cond_resched().
  *
  * This is a very specialized primitive, intended only for a few uses in
  * tracing and other situations requiring manipulation of function preambles
@@ -758,7 +757,7 @@ EXPORT_SYMBOL_GPL(show_rcu_tasks_rude_gp_kthread);
 // 2.	Protects code in the idle loop, exception entry/exit, and
 //	CPU-hotplug code paths, similar to the capabilities of SRCU.
 //
-// 3.	Avoids expensive read-side instruction, having overhead similar
+// 3.	Avoids expensive read-side instructions, having overhead similar
 //	to that of Preemptible RCU.
 //
 // There are of course downsides.  The grace-period code can send IPIs to
@@ -848,7 +847,7 @@ static void rcu_read_unlock_iw(struct irq_work *iwp)
 static DEFINE_IRQ_WORK(rcu_tasks_trace_iw, rcu_read_unlock_iw);
 
 /* If we are the last reader, wake up the grace-period kthread. */
-void rcu_read_unlock_trace_special(struct task_struct *t, int nesting)
+void rcu_read_unlock_trace_special(struct task_struct *t)
 {
 	int nq = READ_ONCE(t->trc_reader_special.b.need_qs);
 
@@ -858,7 +857,7 @@ void rcu_read_unlock_trace_special(struct task_struct *t, int nesting)
 	// Update .need_qs before ->trc_reader_nesting for irq/NMI handlers.
 	if (nq)
 		WRITE_ONCE(t->trc_reader_special.b.need_qs, false);
-	WRITE_ONCE(t->trc_reader_nesting, nesting);
+	WRITE_ONCE(t->trc_reader_nesting, 0);
 	if (nq && atomic_dec_and_test(&trc_n_readers_need_end))
 		irq_work_queue(&rcu_tasks_trace_iw);
 }
@@ -890,32 +889,24 @@ static void trc_read_check_handler(void *t_in)
 
 	// If the task is no longer running on this CPU, leave.
 	if (unlikely(texp != t)) {
-		if (WARN_ON_ONCE(atomic_dec_and_test(&trc_n_readers_need_end)))
-			wake_up(&trc_wait);
 		goto reset_ipi; // Already on holdout list, so will check later.
 	}
 
 	// If the task is not in a read-side critical section, and
 	// if this is the last reader, awaken the grace-period kthread.
 	if (likely(!READ_ONCE(t->trc_reader_nesting))) {
-		if (WARN_ON_ONCE(atomic_dec_and_test(&trc_n_readers_need_end)))
-			wake_up(&trc_wait);
-		// Mark as checked after decrement to avoid false
-		// positives on the above WARN_ON_ONCE().
 		WRITE_ONCE(t->trc_reader_checked, true);
 		goto reset_ipi;
 	}
 	// If we are racing with an rcu_read_unlock_trace(), try again later.
-	if (unlikely(READ_ONCE(t->trc_reader_nesting) < 0)) {
-		if (WARN_ON_ONCE(atomic_dec_and_test(&trc_n_readers_need_end)))
-			wake_up(&trc_wait);
+	if (unlikely(READ_ONCE(t->trc_reader_nesting) < 0))
 		goto reset_ipi;
-	}
 	WRITE_ONCE(t->trc_reader_checked, true);
 
 	// Get here if the task is in a read-side critical section.  Set
 	// its state so that it will awaken the grace-period kthread upon
 	// exit from that critical section.
+	atomic_inc(&trc_n_readers_need_end); // One more to wait on.
 	WARN_ON_ONCE(READ_ONCE(t->trc_reader_special.b.need_qs));
 	WRITE_ONCE(t->trc_reader_special.b.need_qs, true);
 
@@ -931,7 +922,7 @@ reset_ipi:
 static bool trc_inspect_reader(struct task_struct *t, void *arg)
 {
 	int cpu = task_cpu(t);
-	bool in_qs = false;
+	int nesting;
 	bool ofl = cpu_is_offline(cpu);
 
 	if (task_curr(t)) {
@@ -951,18 +942,18 @@ static bool trc_inspect_reader(struct task_struct *t, void *arg)
 		n_heavy_reader_updates++;
 		if (ofl)
 			n_heavy_reader_ofl_updates++;
-		in_qs = true;
+		nesting = 0;
 	} else {
 		// The task is not running, so C-language access is safe.
-		in_qs = likely(!t->trc_reader_nesting);
+		nesting = t->trc_reader_nesting;
 	}
 
-	// Mark as checked so that the grace-period kthread will
-	// remove it from the holdout list.
-	t->trc_reader_checked = true;
-
-	if (in_qs)
-		return true;  // Already in quiescent state, done!!!
+	// If not exiting a read-side critical section, mark as checked
+	// so that the grace-period kthread will remove it from the
+	// holdout list.
+	t->trc_reader_checked = nesting >= 0;
+	if (nesting <= 0)
+		return !nesting;  // If in QS, done, otherwise try again later.
 
 	// The task is in a read-side critical section, so set up its
 	// state so that it will awaken the grace-period kthread upon exit
@@ -1000,7 +991,7 @@ static void trc_wait_for_one_reader(struct task_struct *t,
 
 	// If this task is not yet on the holdout list, then we are in
 	// an RCU read-side critical section.  Otherwise, the invocation of
-	// rcu_add_holdout() that added it to the list did the necessary
+	// trc_add_holdout() that added it to the list did the necessary
 	// get_task_struct().  Either way, the task cannot be freed out
 	// from under this code.
 
@@ -1015,21 +1006,17 @@ static void trc_wait_for_one_reader(struct task_struct *t,
 		if (per_cpu(trc_ipi_to_cpu, cpu) || t->trc_ipi_to_cpu >= 0)
 			return;
 
-		atomic_inc(&trc_n_readers_need_end);
 		per_cpu(trc_ipi_to_cpu, cpu) = true;
 		t->trc_ipi_to_cpu = cpu;
 		rcu_tasks_trace.n_ipis++;
-		if (smp_call_function_single(cpu,
-					     trc_read_check_handler, t, 0)) {
+		if (smp_call_function_single(cpu, trc_read_check_handler, t, 0)) {
 			// Just in case there is some other reason for
 			// failure than the target CPU being offline.
+			WARN_ONCE(1, "%s():  smp_call_function_single() failed for CPU: %d\n",
+				  __func__, cpu);
 			rcu_tasks_trace.n_ipis_fails++;
 			per_cpu(trc_ipi_to_cpu, cpu) = false;
-			t->trc_ipi_to_cpu = cpu;
-			if (atomic_dec_and_test(&trc_n_readers_need_end)) {
-				WARN_ON_ONCE(1);
-				wake_up(&trc_wait);
-			}
+			t->trc_ipi_to_cpu = -1;
 		}
 	}
 }
@@ -1099,9 +1086,9 @@ static void show_stalled_task_trace(struct task_struct *t, bool *firstreport)
 	cpu = task_cpu(t);
 	pr_alert("P%d: %c%c%c nesting: %d%c cpu: %d\n",
 		 t->pid,
-		 ".I"[READ_ONCE(t->trc_ipi_to_cpu) > 0],
+		 ".I"[READ_ONCE(t->trc_ipi_to_cpu) >= 0],
 		 ".i"[is_idle_task(t)],
-		 ".N"[cpu > 0 && tick_nohz_full_cpu(cpu)],
+		 ".N"[cpu >= 0 && tick_nohz_full_cpu(cpu)],
 		 READ_ONCE(t->trc_reader_nesting),
 		 " N"[!!READ_ONCE(t->trc_reader_special.b.need_qs)],
 		 cpu);
@@ -1144,20 +1131,34 @@ static void check_all_holdout_tasks_trace(struct list_head *hop,
 	cpus_read_unlock();
 
 	if (needreport) {
-		if (firstreport)
+		if (*firstreport)
 			pr_err("INFO: rcu_tasks_trace detected stalls? (Late IPI?)\n");
 		show_stalled_ipi_trace();
 	}
 }
 
+static void rcu_tasks_trace_empty_fn(void *unused)
+{
+}
+
 /* Wait for grace period to complete and provide ordering. */
 static void rcu_tasks_trace_postgp(struct rcu_tasks *rtp)
 {
+	int cpu;
 	bool firstreport;
 	struct task_struct *g, *t;
 	LIST_HEAD(holdouts);
 	long ret;
 
+	// Wait for any lingering IPI handlers to complete.  Note that
+	// if a CPU has gone offline or transitioned to userspace in the
+	// meantime, all IPI handlers should have been drained beforehand.
+	// Yes, this assumes that CPUs process IPIs in order.  If that ever
+	// changes, there will need to be a recheck and/or timed wait.
+	for_each_online_cpu(cpu)
+		if (smp_load_acquire(per_cpu_ptr(&trc_ipi_to_cpu, cpu)))
+			smp_call_function_single(cpu, rcu_tasks_trace_empty_fn, NULL, 1);
+
 	// Remove the safety count.
 	smp_mb__before_atomic();  // Order vs. earlier atomics
 	atomic_dec(&trc_n_readers_need_end);
@@ -1200,7 +1201,7 @@ static void exit_tasks_rcu_finish_trace(struct task_struct *t)
 	WARN_ON_ONCE(READ_ONCE(t->trc_reader_nesting));
 	WRITE_ONCE(t->trc_reader_nesting, 0);
 	if (WARN_ON_ONCE(READ_ONCE(t->trc_reader_special.b.need_qs)))
-		rcu_read_unlock_trace_special(t, 0);
+		rcu_read_unlock_trace_special(t);
 }
 
 /**
@@ -1208,15 +1209,11 @@ static void exit_tasks_rcu_finish_trace(struct task_struct *t)
  * @rhp: 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_tasks_trace()
- * assumes that the read-side critical sections end at context switch,
- * cond_resched_rcu_qs(), or transition to usermode execution.  As such,
- * there are no read-side primitives analogous to rcu_read_lock() and
- * rcu_read_unlock() because this primitive is intended to determine
- * that all tasks have passed through a safe state, not so much for
- * data-structure synchronization.
+ * The callback function will be invoked some time after a trace rcu-tasks
+ * grace period elapses, in other words after all currently executing
+ * trace rcu-tasks read-side critical sections have completed. These
+ * read-side critical sections are delimited by calls to rcu_read_lock_trace()
+ * and rcu_read_unlock_trace().
  *
  * See the description of call_rcu() for more detailed information on
  * memory ordering guarantees.
@@ -1232,7 +1229,7 @@ EXPORT_SYMBOL_GPL(call_rcu_tasks_trace);
  *
  * Control will return to the caller some time after a trace rcu-tasks
  * grace period has elapsed, in other words after all currently executing
- * rcu-tasks read-side critical sections have elapsed.  These read-side
+ * trace rcu-tasks read-side critical sections have elapsed. These read-side
  * critical sections are delimited by calls to rcu_read_lock_trace()
  * and rcu_read_unlock_trace().
  *
diff --git a/kernel/scftorture.c b/kernel/scftorture.c
index 64a08288b1a6..5d42f44e3e1a 100644
--- a/kernel/scftorture.c
+++ b/kernel/scftorture.c
@@ -341,6 +341,7 @@ static void scftorture_invoke_one(struct scf_statistics *scfp, struct torture_ra
 			cpu = torture_random(trsp) % nr_cpu_ids;
 			scfp->n_resched++;
 			resched_cpu(cpu);
+			this_cpu_inc(scf_invoked_count);
 		}
 		break;
 	case SCF_PRIM_SINGLE:
@@ -553,18 +554,18 @@ static int __init scf_torture_init(void)
 
 	scftorture_print_module_parms("Start of test");
 
-	if (weight_resched == -1 &&
-	    weight_single == -1 && weight_single_rpc == -1 && weight_single_wait == -1 &&
-	    weight_many == -1 && weight_many_wait == -1 &&
-	    weight_all == -1 && weight_all_wait == -1) {
-		weight_resched1 = 2 * nr_cpu_ids;
-		weight_single1 = 2 * nr_cpu_ids;
-		weight_single_rpc1 = 2 * nr_cpu_ids;
-		weight_single_wait1 = 2 * nr_cpu_ids;
-		weight_many1 = 2;
-		weight_many_wait1 = 2;
-		weight_all1 = 1;
-		weight_all_wait1 = 1;
+	if (weight_resched <= 0 &&
+	    weight_single <= 0 && weight_single_rpc <= 0 && weight_single_wait <= 0 &&
+	    weight_many <= 0 && weight_many_wait <= 0 &&
+	    weight_all <= 0 && weight_all_wait <= 0) {
+		weight_resched1 = weight_resched == 0 ? 0 : 2 * nr_cpu_ids;
+		weight_single1 = weight_single == 0 ? 0 : 2 * nr_cpu_ids;
+		weight_single_rpc1 = weight_single_rpc == 0 ? 0 : 2 * nr_cpu_ids;
+		weight_single_wait1 = weight_single_wait == 0 ? 0 : 2 * nr_cpu_ids;
+		weight_many1 = weight_many == 0 ? 0 : 2;
+		weight_many_wait1 = weight_many_wait == 0 ? 0 : 2;
+		weight_all1 = weight_all == 0 ? 0 : 1;
+		weight_all_wait1 = weight_all_wait == 0 ? 0 : 1;
 	} else {
 		if (weight_resched == -1)
 			weight_resched1 = 0;
@@ -583,8 +584,8 @@ static int __init scf_torture_init(void)
 		if (weight_all_wait == -1)
 			weight_all_wait1 = 0;
 	}
-	if (weight_single1 == 0 && weight_single_rpc1 == 0 && weight_single_wait1 == 0 &&
-	    weight_many1 == 0 && weight_many_wait1 == 0 &&
+	if (weight_resched1 == 0 && weight_single1 == 0 && weight_single_rpc1 == 0 &&
+	    weight_single_wait1 == 0 && weight_many1 == 0 && weight_many_wait1 == 0 &&
 	    weight_all1 == 0 && weight_all_wait1 == 0) {
 		VERBOSE_SCFTORTOUT_ERRSTRING("all zero weights makes no sense");
 		firsterr = -EINVAL;
@@ -605,17 +606,17 @@ static int __init scf_torture_init(void)
 
 	if (onoff_interval > 0) {
 		firsterr = torture_onoff_init(onoff_holdoff * HZ, onoff_interval, NULL);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (shutdown_secs > 0) {
 		firsterr = torture_shutdown_init(shutdown_secs, scf_torture_cleanup);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stutter > 0) {
 		firsterr = torture_stutter_init(stutter, stutter);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 
@@ -636,12 +637,12 @@ static int __init scf_torture_init(void)
 		scf_stats_p[i].cpu = i;
 		firsterr = torture_create_kthread(scftorture_invoker, (void *)&scf_stats_p[i],
 						  scf_stats_p[i].task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 	if (stat_interval > 0) {
 		firsterr = torture_create_kthread(scf_torture_stats, NULL, scf_torture_stats_task);
-		if (firsterr)
+		if (torture_init_error(firsterr))
 			goto unwind;
 	}
 
@@ -651,6 +652,10 @@ static int __init scf_torture_init(void)
 unwind:
 	torture_init_end();
 	scf_torture_cleanup();
+	if (shutdown_secs) {
+		WARN_ON(!IS_MODULE(CONFIG_SCF_TORTURE_TEST));
+		kernel_power_off();
+	}
 	return firsterr;
 }
 
diff --git a/tools/rcu/extract-stall.sh b/tools/rcu/extract-stall.sh
new file mode 100644
index 000000000000..e565697c9f90
--- /dev/null
+++ b/tools/rcu/extract-stall.sh
@@ -0,0 +1,34 @@
+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0+
+#
+# Extract any RCU CPU stall warnings present in specified file.
+# Filter out clocksource lines.  Note that preceding-lines excludes the
+# initial line of the stall warning but trailing-lines includes it.
+#
+# Usage: extract-stall.sh dmesg-file [ preceding-lines [ trailing-lines ] ]
+
+echo $1
+preceding_lines="${2-3}"
+trailing_lines="${3-10}"
+
+awk -v preceding_lines="$preceding_lines" -v trailing_lines="$trailing_lines" '
+suffix <= 0 {
+	for (i = preceding_lines; i > 0; i--)
+		last[i] = last[i - 1];
+	last[0] = $0;
+}
+
+suffix > 0 {
+	print $0;
+	suffix--;
+	if (suffix <= 0)
+		print "";
+}
+
+suffix <= 0 && /detected stall/ {
+	for (i = preceding_lines; i >= 0; i--)
+		if (last[i] != "")
+			print last[i];
+	suffix = trailing_lines;
+}' < "$1" | tr -d '\015' | grep -v clocksource
+
diff --git a/tools/testing/selftests/rcutorture/bin/kvm-remote.sh b/tools/testing/selftests/rcutorture/bin/kvm-remote.sh
index 03126eb6ec5a..c7d42ef80c53 100755
--- a/tools/testing/selftests/rcutorture/bin/kvm-remote.sh
+++ b/tools/testing/selftests/rcutorture/bin/kvm-remote.sh
@@ -149,6 +149,7 @@ do
 done
 
 # Download and expand the tarball on all systems.
+echo Build-products tarball: `du -h $T/binres.tgz` | tee -a "$oldrun/remote-log"
 for i in $systems
 do
 	echo Downloading tarball to $i `date` | tee -a "$oldrun/remote-log"
diff --git a/tools/testing/selftests/rcutorture/bin/kvm.sh b/tools/testing/selftests/rcutorture/bin/kvm.sh
index f442d84fb2a3..6cf9ec6a3d1c 100755
--- a/tools/testing/selftests/rcutorture/bin/kvm.sh
+++ b/tools/testing/selftests/rcutorture/bin/kvm.sh
@@ -184,7 +184,7 @@ do
 		TORTURE_KCONFIG_KASAN_ARG="CONFIG_DEBUG_INFO=y CONFIG_KASAN=y"; export TORTURE_KCONFIG_KASAN_ARG
 		;;
 	--kcsan)
-		TORTURE_KCONFIG_KCSAN_ARG="CONFIG_DEBUG_INFO=y CONFIG_KCSAN=y CONFIG_KCSAN_ASSUME_PLAIN_WRITES_ATOMIC=n CONFIG_KCSAN_REPORT_VALUE_CHANGE_ONLY=n CONFIG_KCSAN_REPORT_ONCE_IN_MS=100000 CONFIG_KCSAN_INTERRUPT_WATCHER=y CONFIG_KCSAN_VERBOSE=y CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_PROVE_LOCKING=y"; export TORTURE_KCONFIG_KCSAN_ARG
+		TORTURE_KCONFIG_KCSAN_ARG="CONFIG_DEBUG_INFO=y CONFIG_KCSAN=y CONFIG_KCSAN_STRICT=y CONFIG_KCSAN_REPORT_ONCE_IN_MS=100000 CONFIG_KCSAN_VERBOSE=y CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_PROVE_LOCKING=y"; export TORTURE_KCONFIG_KCSAN_ARG
 		;;
 	--kmake-arg|--kmake-args)
 		checkarg --kmake-arg "(kernel make arguments)" $# "$2" '.*' '^error$'
diff --git a/tools/testing/selftests/rcutorture/bin/torture.sh b/tools/testing/selftests/rcutorture/bin/torture.sh
index 363f56081eff..eae88aacca2a 100755
--- a/tools/testing/selftests/rcutorture/bin/torture.sh
+++ b/tools/testing/selftests/rcutorture/bin/torture.sh
@@ -351,7 +351,7 @@ fi
 if test "$do_scftorture" = "yes"
 then
 	torture_bootargs="scftorture.nthreads=$HALF_ALLOTED_CPUS torture.disable_onoff_at_boot"
-	torture_set "scftorture" tools/testing/selftests/rcutorture/bin/kvm.sh --torture scf --allcpus --duration "$duration_scftorture" --configs "$configs_scftorture" --kconfig "CONFIG_NR_CPUS=$HALF_ALLOTED_CPUS" --trust-make
+	torture_set "scftorture" tools/testing/selftests/rcutorture/bin/kvm.sh --torture scf --allcpus --duration "$duration_scftorture" --configs "$configs_scftorture" --kconfig "CONFIG_NR_CPUS=$HALF_ALLOTED_CPUS" --memory 1G --trust-make
 fi
 
 if test "$do_refscale" = yes
@@ -434,7 +434,12 @@ then
 	batchno=1
 	if test -s $T/xz-todo
 	then
-		echo Size before compressing: `du -sh $tdir | awk '{ print $1 }'` `date` 2>&1 | tee -a "$tdir/log-xz" | tee -a $T/log
+		for i in `cat $T/xz-todo`
+		do
+			find $i -name 'vmlinux*' -print
+		done | wc -l | awk '{ print $1 }' > $T/xz-todo-count
+		n2compress="`cat $T/xz-todo-count`"
+		echo Size before compressing $n2compress files: `du -sh $tdir | awk '{ print $1 }'` `date` 2>&1 | tee -a "$tdir/log-xz" | tee -a $T/log
 		for i in `cat $T/xz-todo`
 		do
 			echo Compressing vmlinux files in ${i}: `date` >> "$tdir/log-xz" 2>&1
@@ -456,7 +461,7 @@ then
 			echo Waiting for final batch $batchno of $ncompresses compressions `date` | tee -a "$tdir/log-xz" | tee -a $T/log
 		fi
 		wait
-		echo Size after compressing: `du -sh $tdir | awk '{ print $1 }'` `date` 2>&1 | tee -a "$tdir/log-xz" | tee -a $T/log
+		echo Size after compressing $n2compress files: `du -sh $tdir | awk '{ print $1 }'` `date` 2>&1 | tee -a "$tdir/log-xz" | tee -a $T/log
 		echo Total duration `get_starttime_duration $starttime`. | tee -a $T/log
 	else
 		echo No compression needed: `date` >> "$tdir/log-xz" 2>&1