Merge branches 'doc.2021.07.20c', 'fixes.2021.08.06a', 'nocb.2021.07.…

…20c', 'nolibc.2021.07.20c', 'tasks.2021.07.20c', 'torture.2021.07.27a' and 'torturescript.2021.07.27a' into HEAD

doc.2021.07.20c: Documentation updates.
fixes.2021.08.06a: Miscellaneous fixes.
nocb.2021.07.20c: Callback-offloading (NOCB CPU) updates.
nolibc.2021.07.20c: Tiny userspace library updates.
tasks.2021.07.20c: Tasks RCU updates.
torture.2021.07.27a: In-kernel torture-test updates.
torturescript.2021.07.27a: Torture-test scripting updates.

include/linux/rculist.h

@@ -10,15 +10,6 @@
 #include <linux/list.h>
 #include <linux/rcupdate.h>
 
-/*
- * Why is there no list_empty_rcu()?  Because list_empty() serves this
- * purpose.  The list_empty() function fetches the RCU-protected pointer
- * and compares it to the address of the list head, but neither dereferences
- * this pointer itself nor provides this pointer to the caller.  Therefore,
- * it is not necessary to use rcu_dereference(), so that list_empty() can
- * be used anywhere you would want to use a list_empty_rcu().
- */
-
 /*
  * INIT_LIST_HEAD_RCU - Initialize a list_head visible to RCU readers
  * @list: list to be initialized
@@ -318,21 +309,29 @@ static inline void list_splice_tail_init_rcu(struct list_head *list,
 /*
  * Where are list_empty_rcu() and list_first_entry_rcu()?
  *
- * Implementing those functions following their counterparts list_empty() and
- * list_first_entry() is not advisable because they lead to subtle race
- * conditions as the following snippet shows:
+ * They do not exist because they would lead to subtle race conditions:
  *
  * if (!list_empty_rcu(mylist)) {
  *	struct foo *bar = list_first_entry_rcu(mylist, struct foo, list_member);
  *	do_something(bar);
  * }
  *
- * The list may not be empty when list_empty_rcu checks it, but it may be when
- * list_first_entry_rcu rereads the ->next pointer.
- *
- * Rereading the ->next pointer is not a problem for list_empty() and
- * list_first_entry() because they would be protected by a lock that blocks
- * writers.
+ * The list might be non-empty when list_empty_rcu() checks it, but it
+ * might have become empty by the time that list_first_entry_rcu() rereads
+ * the ->next pointer, which would result in a SEGV.
+ *
+ * When not using RCU, it is OK for list_first_entry() to re-read that
+ * pointer because both functions should be protected by some lock that
+ * blocks writers.
+ *
+ * When using RCU, list_empty() uses READ_ONCE() to fetch the
+ * RCU-protected ->next pointer and then compares it to the address of the
+ * list head.  However, it neither dereferences this pointer nor provides
+ * this pointer to its caller.  Thus, READ_ONCE() suffices (that is,
+ * rcu_dereference() is not needed), which means that list_empty() can be
+ * used anywhere you would want to use list_empty_rcu().  Just don't
+ * expect anything useful to happen if you do a subsequent lockless
+ * call to list_first_entry_rcu()!!!
  *
  * See list_first_or_null_rcu for an alternative.
  */

include/linux/rcupdate.h

@@ -53,7 +53,7 @@ void __rcu_read_unlock(void);
  * nesting depth, but makes sense only if CONFIG_PREEMPT_RCU -- in other
  * types of kernel builds, the rcu_read_lock() nesting depth is unknowable.
  */
-#define rcu_preempt_depth() (current->rcu_read_lock_nesting)
+#define rcu_preempt_depth() READ_ONCE(current->rcu_read_lock_nesting)
 
 #else /* #ifdef CONFIG_PREEMPT_RCU */
 
@@ -167,7 +167,7 @@ void synchronize_rcu_tasks(void);
 # define synchronize_rcu_tasks synchronize_rcu
 # endif
 
-# ifdef CONFIG_TASKS_RCU_TRACE
+# ifdef CONFIG_TASKS_TRACE_RCU
 # define rcu_tasks_trace_qs(t)						\
 	do {								\
 		if (!likely(READ_ONCE((t)->trc_reader_checked)) &&	\

include/linux/rcutiny.h

@@ -14,9 +14,6 @@
 
 #include <asm/param.h> /* for HZ */
 
-/* Never flag non-existent other CPUs! */
-static inline bool rcu_eqs_special_set(int cpu) { return false; }
-
 unsigned long get_state_synchronize_rcu(void);
 unsigned long start_poll_synchronize_rcu(void);
 bool poll_state_synchronize_rcu(unsigned long oldstate);

include/linux/srcutiny.h

@@ -61,7 +61,7 @@ static inline int __srcu_read_lock(struct srcu_struct *ssp)
 	int idx;
 
 	idx = ((READ_ONCE(ssp->srcu_idx) + 1) & 0x2) >> 1;
-	WRITE_ONCE(ssp->srcu_lock_nesting[idx], ssp->srcu_lock_nesting[idx] + 1);
+	WRITE_ONCE(ssp->srcu_lock_nesting[idx], READ_ONCE(ssp->srcu_lock_nesting[idx]) + 1);
 	return idx;
 }
 
@@ -81,11 +81,11 @@ static inline void srcu_torture_stats_print(struct srcu_struct *ssp,
 {
 	int idx;
 
-	idx = ((READ_ONCE(ssp->srcu_idx) + 1) & 0x2) >> 1;
+	idx = ((data_race(READ_ONCE(ssp->srcu_idx)) + 1) & 0x2) >> 1;
 	pr_alert("%s%s Tiny SRCU per-CPU(idx=%d): (%hd,%hd)\n",
 		 tt, tf, idx,
-		 READ_ONCE(ssp->srcu_lock_nesting[!idx]),
-		 READ_ONCE(ssp->srcu_lock_nesting[idx]));
+		 data_race(READ_ONCE(ssp->srcu_lock_nesting[!idx])),
+		 data_race(READ_ONCE(ssp->srcu_lock_nesting[idx])));
 }
 
 #endif

kernel/locking/locktorture.c

@@ -59,7 +59,7 @@ static struct task_struct **writer_tasks;
 static struct task_struct **reader_tasks;
 
 static bool lock_is_write_held;
-static bool lock_is_read_held;
+static atomic_t lock_is_read_held;
 static unsigned long last_lock_release;
 
 struct lock_stress_stats {
@@ -682,7 +682,7 @@ static int lock_torture_writer(void *arg)
 		if (WARN_ON_ONCE(lock_is_write_held))
 			lwsp->n_lock_fail++;
 		lock_is_write_held = true;
-		if (WARN_ON_ONCE(lock_is_read_held))
+		if (WARN_ON_ONCE(atomic_read(&lock_is_read_held)))
 			lwsp->n_lock_fail++; /* rare, but... */
 
 		lwsp->n_lock_acquired++;
@@ -717,13 +717,13 @@ static int lock_torture_reader(void *arg)
 			schedule_timeout_uninterruptible(1);
 
 		cxt.cur_ops->readlock(tid);
-		lock_is_read_held = true;
+		atomic_inc(&lock_is_read_held);
 		if (WARN_ON_ONCE(lock_is_write_held))
 			lrsp->n_lock_fail++; /* rare, but... */
 
 		lrsp->n_lock_acquired++;
 		cxt.cur_ops->read_delay(&rand);
-		lock_is_read_held = false;
+		atomic_dec(&lock_is_read_held);
 		cxt.cur_ops->readunlock(tid);
 
 		stutter_wait("lock_torture_reader");
@@ -738,20 +738,22 @@ static int lock_torture_reader(void *arg)
 static void __torture_print_stats(char *page,
 				  struct lock_stress_stats *statp, bool write)
 {
+	long cur;
 	bool fail = false;
 	int i, n_stress;
-	long max = 0, min = statp ? statp[0].n_lock_acquired : 0;
+	long max = 0, min = statp ? data_race(statp[0].n_lock_acquired) : 0;
 	long long sum = 0;
 
 	n_stress = write ? cxt.nrealwriters_stress : cxt.nrealreaders_stress;
 	for (i = 0; i < n_stress; i++) {
-		if (statp[i].n_lock_fail)
+		if (data_race(statp[i].n_lock_fail))
 			fail = true;
-		sum += statp[i].n_lock_acquired;
-		if (max < statp[i].n_lock_acquired)
-			max = statp[i].n_lock_acquired;
-		if (min > statp[i].n_lock_acquired)
-			min = statp[i].n_lock_acquired;
+		cur = data_race(statp[i].n_lock_acquired);
+		sum += cur;
+		if (max < cur)
+			max = cur;
+		if (min > cur)
+			min = cur;
 	}
 	page += sprintf(page,
 			"%s:  Total: %lld  Max/Min: %ld/%ld %s  Fail: %d %s\n",
@@ -996,7 +998,6 @@ static int __init lock_torture_init(void)
 		}
 
 		if (nreaders_stress) {
-			lock_is_read_held = false;
 			cxt.lrsa = kmalloc_array(cxt.nrealreaders_stress,
 						 sizeof(*cxt.lrsa),
 						 GFP_KERNEL);

kernel/rcu/rcuscale.c

@@ -487,7 +487,7 @@ rcu_scale_writer(void *arg)
 	if (gp_async) {
 		cur_ops->gp_barrier();
 	}
-	writer_n_durations[me] = i_max;
+	writer_n_durations[me] = i_max + 1;
 	torture_kthread_stopping("rcu_scale_writer");
 	return 0;
 }
@@ -561,7 +561,7 @@ rcu_scale_cleanup(void)
 			wdpp = writer_durations[i];
 			if (!wdpp)
 				continue;
-			for (j = 0; j <= writer_n_durations[i]; j++) {
+			for (j = 0; j < writer_n_durations[i]; j++) {
 				wdp = &wdpp[j];
 				pr_alert("%s%s %4d writer-duration: %5d %llu\n",
 					scale_type, SCALE_FLAG,

kernel/rcu/rcutorture.c

@@ -2022,8 +2022,13 @@ static int rcu_torture_stall(void *args)
 			  __func__, raw_smp_processor_id());
 		while (ULONG_CMP_LT((unsigned long)ktime_get_seconds(),
 				    stop_at))
-			if (stall_cpu_block)
+			if (stall_cpu_block) {
+#ifdef CONFIG_PREEMPTION
+				preempt_schedule();
+#else
 				schedule_timeout_uninterruptible(HZ);
+#endif
+			}
 		if (stall_cpu_irqsoff)
 			local_irq_enable();
 		else if (!stall_cpu_block)

kernel/rcu/refscale.c

@@ -467,6 +467,40 @@ static struct ref_scale_ops acqrel_ops = {
 	.name		= "acqrel"
 };
 
+static volatile u64 stopopts;
+
+static void ref_clock_section(const int nloops)
+{
+	u64 x = 0;
+	int i;
+
+	preempt_disable();
+	for (i = nloops; i >= 0; i--)
+		x += ktime_get_real_fast_ns();
+	preempt_enable();
+	stopopts = x;
+}
+
+static void ref_clock_delay_section(const int nloops, const int udl, const int ndl)
+{
+	u64 x = 0;
+	int i;
+
+	preempt_disable();
+	for (i = nloops; i >= 0; i--) {
+		x += ktime_get_real_fast_ns();
+		un_delay(udl, ndl);
+	}
+	preempt_enable();
+	stopopts = x;
+}
+
+static struct ref_scale_ops clock_ops = {
+	.readsection	= ref_clock_section,
+	.delaysection	= ref_clock_delay_section,
+	.name		= "clock"
+};
+
 static void rcu_scale_one_reader(void)
 {
 	if (readdelay <= 0)
@@ -759,7 +793,7 @@ ref_scale_init(void)
 	int firsterr = 0;
 	static struct ref_scale_ops *scale_ops[] = {
 		&rcu_ops, &srcu_ops, &rcu_trace_ops, &rcu_tasks_ops, &refcnt_ops, &rwlock_ops,
-		&rwsem_ops, &lock_ops, &lock_irq_ops, &acqrel_ops,
+		&rwsem_ops, &lock_ops, &lock_irq_ops, &acqrel_ops, &clock_ops,
 	};
 
 	if (!torture_init_begin(scale_type, verbose))

kernel/rcu/srcutiny.c

@@ -96,7 +96,7 @@ EXPORT_SYMBOL_GPL(cleanup_srcu_struct);
  */
 void __srcu_read_unlock(struct srcu_struct *ssp, int idx)
 {
-	int newval = ssp->srcu_lock_nesting[idx] - 1;
+	int newval = READ_ONCE(ssp->srcu_lock_nesting[idx]) - 1;
 
 	WRITE_ONCE(ssp->srcu_lock_nesting[idx], newval);
 	if (!newval && READ_ONCE(ssp->srcu_gp_waiting))

kernel/rcu/tasks.h

@@ -643,8 +643,8 @@ void exit_tasks_rcu_finish(void) { exit_tasks_rcu_finish_trace(current); }
 //
 // "Rude" variant of Tasks RCU, inspired by Steve Rostedt's trick of
 // passing an empty function to schedule_on_each_cpu().  This approach
-// provides an asynchronous call_rcu_tasks_rude() API and batching
-// of concurrent calls to the synchronous synchronize_rcu_rude() API.
+// provides an asynchronous call_rcu_tasks_rude() API and batching of
+// concurrent calls to the synchronous synchronize_rcu_tasks_rude() API.
 // This invokes schedule_on_each_cpu() in order to send IPIs far and wide
 // and induces otherwise unnecessary context switches on all online CPUs,
 // whether idle or not.
@@ -785,7 +785,10 @@ EXPORT_SYMBOL_GPL(show_rcu_tasks_rude_gp_kthread);
 //	set that task's .need_qs flag so that task's next outermost
 //	rcu_read_unlock_trace() will report the quiescent state (in which
 //	case the count of readers is incremented).  If both attempts fail,
-//	the task is added to a "holdout" list.
+//	the task is added to a "holdout" list.  Note that IPIs are used
+//	to invoke trc_read_check_handler() in the context of running tasks
+//	in order to avoid ordering overhead on common-case shared-variable
+//	accessses.
 // rcu_tasks_trace_postscan():
 //	Initialize state and attempt to identify an immediate quiescent
 //	state as above (but only for idle tasks), unblock CPU-hotplug
@@ -847,7 +850,7 @@ 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)
 {
-	int nq = t->trc_reader_special.b.need_qs;
+	int nq = READ_ONCE(t->trc_reader_special.b.need_qs);
 
 	if (IS_ENABLED(CONFIG_TASKS_TRACE_RCU_READ_MB) &&
 	    t->trc_reader_special.b.need_mb)
@@ -894,7 +897,7 @@ static void trc_read_check_handler(void *t_in)
 
 	// 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(!t->trc_reader_nesting)) {
+	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
@@ -903,7 +906,7 @@ static void trc_read_check_handler(void *t_in)
 		goto reset_ipi;
 	}
 	// If we are racing with an rcu_read_unlock_trace(), try again later.
-	if (unlikely(t->trc_reader_nesting < 0)) {
+	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);
 		goto reset_ipi;
@@ -913,14 +916,14 @@ static void trc_read_check_handler(void *t_in)
 	// 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.
-	WARN_ON_ONCE(t->trc_reader_special.b.need_qs);
+	WARN_ON_ONCE(READ_ONCE(t->trc_reader_special.b.need_qs));
 	WRITE_ONCE(t->trc_reader_special.b.need_qs, true);
 
 reset_ipi:
 	// Allow future IPIs to be sent on CPU and for task.
 	// Also order this IPI handler against any later manipulations of
 	// the intended task.
-	smp_store_release(&per_cpu(trc_ipi_to_cpu, smp_processor_id()), false); // ^^^
+	smp_store_release(per_cpu_ptr(&trc_ipi_to_cpu, smp_processor_id()), false); // ^^^
 	smp_store_release(&texp->trc_ipi_to_cpu, -1); // ^^^
 }
 
@@ -950,6 +953,7 @@ static bool trc_inspect_reader(struct task_struct *t, void *arg)
 			n_heavy_reader_ofl_updates++;
 		in_qs = true;
 	} else {
+		// The task is not running, so C-language access is safe.
 		in_qs = likely(!t->trc_reader_nesting);
 	}
 
@@ -964,7 +968,7 @@ static bool trc_inspect_reader(struct task_struct *t, void *arg)
 	// 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(t->trc_reader_special.b.need_qs);
+	WARN_ON_ONCE(READ_ONCE(t->trc_reader_special.b.need_qs));
 	WRITE_ONCE(t->trc_reader_special.b.need_qs, true);
 	return true;
 }
@@ -982,7 +986,7 @@ static void trc_wait_for_one_reader(struct task_struct *t,
 	// The current task had better be in a quiescent state.
 	if (t == current) {
 		t->trc_reader_checked = true;
-		WARN_ON_ONCE(t->trc_reader_nesting);
+		WARN_ON_ONCE(READ_ONCE(t->trc_reader_nesting));
 		return;
 	}
 
@@ -994,6 +998,12 @@ static void trc_wait_for_one_reader(struct task_struct *t,
 	}
 	put_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
+	// get_task_struct().  Either way, the task cannot be freed out
+	// from under this code.
+
 	// If currently running, send an IPI, either way, add to list.
 	trc_add_holdout(t, bhp);
 	if (task_curr(t) &&
@@ -1092,8 +1102,8 @@ static void show_stalled_task_trace(struct task_struct *t, bool *firstreport)
 		 ".I"[READ_ONCE(t->trc_ipi_to_cpu) > 0],
 		 ".i"[is_idle_task(t)],
 		 ".N"[cpu > 0 && tick_nohz_full_cpu(cpu)],
-		 t->trc_reader_nesting,
-		 " N"[!!t->trc_reader_special.b.need_qs],
+		 READ_ONCE(t->trc_reader_nesting),
+		 " N"[!!READ_ONCE(t->trc_reader_special.b.need_qs)],
 		 cpu);
 	sched_show_task(t);
 }
@@ -1187,7 +1197,7 @@ static void rcu_tasks_trace_postgp(struct rcu_tasks *rtp)
 static void exit_tasks_rcu_finish_trace(struct task_struct *t)
 {
 	WRITE_ONCE(t->trc_reader_checked, true);
-	WARN_ON_ONCE(t->trc_reader_nesting);
+	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);