Merge tag 'rcu.2022.09.30a' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu

Pull RCU updates from Paul McKenney:

 - Documentation updates.

   This is the first in a series from an ongoing review of the RCU
   documentation. "Why are people thinking -that- about RCU? Oh. Because
   that is an entirely reasonable interpretation of its documentation."

 - Miscellaneous fixes.

 - Improved memory allocation and heuristics.

 - Improve rcu_nocbs diagnostic output.

 - Add full-sized polled RCU grace period state values.

   These are the same size as an rcu_head structure, which is double
   that of the traditional unsigned long state values that may still be
   obtained from et_state_synchronize_rcu(). The added size avoids
   missing overlapping grace periods. This benefit is that call_rcu()
   can be replaced by polling, which can be attractive in situations
   where RCU-protected data is aged out of memory.

   Early in the series, the size of this state value is three unsigned
   longs. Later in the series, the fastpaths in synchronize_rcu() and
   synchronize_rcu_expedited() are reworked to permit the full state to
   be represented by only two unsigned longs. This reworking slows these
   two functions down in SMP kernels running either on single-CPU
   systems or on systems with all but one CPU offlined, but this should
   not be a significant problem. And if it somehow becomes a problem in
   some yet-as-unforeseen situations, three-value state values can be
   provided for only those situations.

   Finally, a pair of functions named same_state_synchronize_rcu() and
   same_state_synchronize_rcu_full() allow grace-period state values to
   be compared for equality. This permits users to maintain lists of
   data structures having the same state value, removing the need for
   per-data-structure grace-period state values, thus decreasing memory
   footprint.

 - Polled SRCU grace-period updates, including adding tests to
   rcutorture and reducing the incidence of Tiny SRCU grace-period-state
   counter wrap.

 - Improve Tasks RCU diagnostics and quiescent-state detection.

* tag 'rcu.2022.09.30a' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu: (55 commits)
  rcutorture: Use the barrier operation specified by cur_ops
  rcu-tasks: Make RCU Tasks Trace check for userspace execution
  rcu-tasks: Ensure RCU Tasks Trace loops have quiescent states
  rcu-tasks: Convert RCU_LOCKDEP_WARN() to WARN_ONCE()
  srcu: Make Tiny SRCU use full-sized grace-period counters
  srcu: Make Tiny SRCU poll_state_synchronize_srcu() more precise
  srcu: Add GP and maximum requested GP to Tiny SRCU rcutorture output
  rcutorture: Make "srcud" option also test polled grace-period API
  rcutorture: Limit read-side polling-API testing
  rcu: Add functions to compare grace-period state values
  rcutorture: Expand rcu_torture_write_types() first "if" statement
  rcutorture: Use 1-suffixed variable in rcu_torture_write_types() check
  rcu: Make synchronize_rcu() fastpath update only boot-CPU counters
  rcutorture: Adjust rcu_poll_need_2gp() for rcu_gp_oldstate field removal
  rcu: Remove ->rgos_polled field from rcu_gp_oldstate structure
  rcu: Make synchronize_rcu_expedited() fast path update .expedited_sequence
  rcu: Remove expedited grace-period fast-path forward-progress helper
  rcu: Make synchronize_rcu() fast path update ->gp_seq counters
  rcu-tasks: Remove grace-period fast-path rcu-tasks helper
  rcu: Set rcu_data structures' initial ->gpwrap value to true
  ...

Author: torvalds

Documentation/RCU/checklist.rst

@@ -66,8 +66,13 @@ over a rather long period of time, but improvements are always welcome!
 	As a rough rule of thumb, any dereference of an RCU-protected
 	pointer must be covered by rcu_read_lock(), rcu_read_lock_bh(),
 	rcu_read_lock_sched(), or by the appropriate update-side lock.
-	Disabling of preemption can serve as rcu_read_lock_sched(), but
-	is less readable and prevents lockdep from detecting locking issues.
+	Explicit disabling of preemption (preempt_disable(), for example)
+	can serve as rcu_read_lock_sched(), but is less readable and
+	prevents lockdep from detecting locking issues.
+
+	Please not that you *cannot* rely on code known to be built
+	only in non-preemptible kernels.  Such code can and will break,
+	especially in kernels built with CONFIG_PREEMPT_COUNT=y.
 
 	Letting RCU-protected pointers "leak" out of an RCU read-side
 	critical section is every bit as bad as letting them leak out
@@ -185,6 +190,9 @@ over a rather long period of time, but improvements are always welcome!
 
 5.	If call_rcu() or call_srcu() is used, the callback function will
 	be called from softirq context.  In particular, it cannot block.
+	If you need the callback to block, run that code in a workqueue
+	handler scheduled from the callback.  The queue_rcu_work()
+	function does this for you in the case of call_rcu().
 
 6.	Since synchronize_rcu() can block, it cannot be called
 	from any sort of irq context.  The same rule applies
@@ -297,7 +305,8 @@ over a rather long period of time, but improvements are always welcome!
 		the machine.
 
 	d.	Periodically invoke synchronize_rcu(), permitting a limited
-		number of updates per grace period.
+		number of updates per grace period.  Better yet, periodically
+		invoke rcu_barrier() to wait for all outstanding callbacks.
 
 	The same cautions apply to call_srcu() and kfree_rcu().
 

Documentation/RCU/rcu_dereference.rst

@@ -128,10 +128,16 @@ Follow these rules to keep your RCU code working properly:
 		This sort of comparison occurs frequently when scanning
 		RCU-protected circular linked lists.
 
-		Note that if checks for being within an RCU read-side
-		critical section are not required and the pointer is never
-		dereferenced, rcu_access_pointer() should be used in place
-		of rcu_dereference().
+		Note that if the pointer comparison is done outside
+		of an RCU read-side critical section, and the pointer
+		is never dereferenced, rcu_access_pointer() should be
+		used in place of rcu_dereference().  In most cases,
+		it is best to avoid accidental dereferences by testing
+		the rcu_access_pointer() return value directly, without
+		assigning it to a variable.
+
+		Within an RCU read-side critical section, there is little
+		reason to use rcu_access_pointer().
 
 	-	The comparison is against a pointer that references memory
 		that was initialized "a long time ago."  The reason

Documentation/RCU/whatisRCU.rst

@@ -6,13 +6,15 @@ What is RCU?  --  "Read, Copy, Update"
 Please note that the "What is RCU?" LWN series is an excellent place
 to start learning about RCU:
 
-| 1.	What is RCU, Fundamentally?  http://lwn.net/Articles/262464/
-| 2.	What is RCU? Part 2: Usage   http://lwn.net/Articles/263130/
-| 3.	RCU part 3: the RCU API      http://lwn.net/Articles/264090/
-| 4.	The RCU API, 2010 Edition    http://lwn.net/Articles/418853/
-| 	2010 Big API Table           http://lwn.net/Articles/419086/
-| 5.	The RCU API, 2014 Edition    http://lwn.net/Articles/609904/
-|	2014 Big API Table           http://lwn.net/Articles/609973/
+| 1.	What is RCU, Fundamentally?  https://lwn.net/Articles/262464/
+| 2.	What is RCU? Part 2: Usage   https://lwn.net/Articles/263130/
+| 3.	RCU part 3: the RCU API      https://lwn.net/Articles/264090/
+| 4.	The RCU API, 2010 Edition    https://lwn.net/Articles/418853/
+| 	2010 Big API Table           https://lwn.net/Articles/419086/
+| 5.	The RCU API, 2014 Edition    https://lwn.net/Articles/609904/
+|	2014 Big API Table           https://lwn.net/Articles/609973/
+| 6.	The RCU API, 2019 Edition    https://lwn.net/Articles/777036/
+|	2019 Big API Table           https://lwn.net/Articles/777165/
 
 
 What is RCU?
@@ -915,13 +917,18 @@ which an RCU reference is held include:
 The understanding that RCU provides a reference that only prevents a
 change of type is particularly visible with objects allocated from a
 slab cache marked ``SLAB_TYPESAFE_BY_RCU``.  RCU operations may yield a
-reference to an object from such a cache that has been concurrently
-freed and the memory reallocated to a completely different object,
-though of the same type.  In this case RCU doesn't even protect the
-identity of the object from changing, only its type.  So the object
-found may not be the one expected, but it will be one where it is safe
-to take a reference or spinlock and then confirm that the identity
-matches the expectations.
+reference to an object from such a cache that has been concurrently freed
+and the memory reallocated to a completely different object, though of
+the same type.  In this case RCU doesn't even protect the identity of the
+object from changing, only its type.  So the object found may not be the
+one expected, but it will be one where it is safe to take a reference
+(and then potentially acquiring a spinlock), allowing subsequent code
+to check whether the identity matches expectations.  It is tempting
+to simply acquire the spinlock without first taking the reference, but
+unfortunately any spinlock in a ``SLAB_TYPESAFE_BY_RCU`` object must be
+initialized after each and every call to kmem_cache_alloc(), which renders
+reference-free spinlock acquisition completely unsafe.  Therefore, when
+using ``SLAB_TYPESAFE_BY_RCU``, make proper use of a reference counter.
 
 With traditional reference counting -- such as that implemented by the
 kref library in Linux -- there is typically code that runs when the last
@@ -1057,14 +1064,20 @@ SRCU: Initialization/cleanup::
 	init_srcu_struct
 	cleanup_srcu_struct
 
-All: lockdep-checked RCU-protected pointer access::
+All: lockdep-checked RCU utility APIs::
 
-	rcu_access_pointer
-	rcu_dereference_raw
 	RCU_LOCKDEP_WARN
 	rcu_sleep_check
 	RCU_NONIDLE
 
+All: Unchecked RCU-protected pointer access::
+
+	rcu_dereference_raw
+
+All: Unchecked RCU-protected pointer access with dereferencing prohibited::
+
+	rcu_access_pointer
+
 See the comment headers in the source code (or the docbook generated
 from them) for more information.
 

include/linux/rcupdate.h

@@ -42,7 +42,31 @@ void call_rcu(struct rcu_head *head, rcu_callback_t func);
 void rcu_barrier_tasks(void);
 void rcu_barrier_tasks_rude(void);
 void synchronize_rcu(void);
+
+struct rcu_gp_oldstate;
 unsigned long get_completed_synchronize_rcu(void);
+void get_completed_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp);
+
+// Maximum number of unsigned long values corresponding to
+// not-yet-completed RCU grace periods.
+#define NUM_ACTIVE_RCU_POLL_OLDSTATE 2
+
+/**
+ * same_state_synchronize_rcu - Are two old-state values identical?
+ * @oldstate1: First old-state value.
+ * @oldstate2: Second old-state value.
+ *
+ * The two old-state values must have been obtained from either
+ * get_state_synchronize_rcu(), start_poll_synchronize_rcu(), or
+ * get_completed_synchronize_rcu().  Returns @true if the two values are
+ * identical and @false otherwise.  This allows structures whose lifetimes
+ * are tracked by old-state values to push these values to a list header,
+ * allowing those structures to be slightly smaller.
+ */
+static inline bool same_state_synchronize_rcu(unsigned long oldstate1, unsigned long oldstate2)
+{
+	return oldstate1 == oldstate2;
+}
 
 #ifdef CONFIG_PREEMPT_RCU
 
@@ -496,13 +520,21 @@ do {									      \
  * against NULL.  Although rcu_access_pointer() may also be used in cases
  * where update-side locks prevent the value of the pointer from changing,
  * you should instead use rcu_dereference_protected() for this use case.
+ * Within an RCU read-side critical section, there is little reason to
+ * use rcu_access_pointer().
+ *
+ * It is usually best to test the rcu_access_pointer() return value
+ * directly in order to avoid accidental dereferences being introduced
+ * by later inattentive changes.  In other words, assigning the
+ * rcu_access_pointer() return value to a local variable results in an
+ * accident waiting to happen.
  *
  * It is also permissible to use rcu_access_pointer() when read-side
- * access to the pointer was removed at least one grace period ago, as
- * is the case in the context of the RCU callback that is freeing up
- * the data, or after a synchronize_rcu() returns.  This can be useful
- * when tearing down multi-linked structures after a grace period
- * has elapsed.
+ * access to the pointer was removed at least one grace period ago, as is
+ * the case in the context of the RCU callback that is freeing up the data,
+ * or after a synchronize_rcu() returns.  This can be useful when tearing
+ * down multi-linked structures after a grace period has elapsed.  However,
+ * rcu_dereference_protected() is normally preferred for this use case.
  */
 #define rcu_access_pointer(p) __rcu_access_pointer((p), __UNIQUE_ID(rcu), __rcu)
 

include/linux/rcutiny.h

@@ -14,25 +14,75 @@
 
 #include <asm/param.h> /* for HZ */
 
+struct rcu_gp_oldstate {
+	unsigned long rgos_norm;
+};
+
+// Maximum number of rcu_gp_oldstate values corresponding to
+// not-yet-completed RCU grace periods.
+#define NUM_ACTIVE_RCU_POLL_FULL_OLDSTATE 2
+
+/*
+ * Are the two oldstate values the same?  See the Tree RCU version for
+ * docbook header.
+ */
+static inline bool same_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp1,
+						   struct rcu_gp_oldstate *rgosp2)
+{
+	return rgosp1->rgos_norm == rgosp2->rgos_norm;
+}
+
 unsigned long get_state_synchronize_rcu(void);
+
+static inline void get_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp)
+{
+	rgosp->rgos_norm = get_state_synchronize_rcu();
+}
+
 unsigned long start_poll_synchronize_rcu(void);
+
+static inline void start_poll_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp)
+{
+	rgosp->rgos_norm = start_poll_synchronize_rcu();
+}
+
 bool poll_state_synchronize_rcu(unsigned long oldstate);
 
+static inline bool poll_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp)
+{
+	return poll_state_synchronize_rcu(rgosp->rgos_norm);
+}
+
 static inline void cond_synchronize_rcu(unsigned long oldstate)
 {
 	might_sleep();
 }
 
+static inline void cond_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp)
+{
+	cond_synchronize_rcu(rgosp->rgos_norm);
+}
+
 static inline unsigned long start_poll_synchronize_rcu_expedited(void)
 {
 	return start_poll_synchronize_rcu();
 }
 
+static inline void start_poll_synchronize_rcu_expedited_full(struct rcu_gp_oldstate *rgosp)
+{
+	rgosp->rgos_norm = start_poll_synchronize_rcu_expedited();
+}
+
 static inline void cond_synchronize_rcu_expedited(unsigned long oldstate)
 {
 	cond_synchronize_rcu(oldstate);
 }
 
+static inline void cond_synchronize_rcu_expedited_full(struct rcu_gp_oldstate *rgosp)
+{
+	cond_synchronize_rcu_expedited(rgosp->rgos_norm);
+}
+
 extern void rcu_barrier(void);
 
 static inline void synchronize_rcu_expedited(void)

include/linux/rcutree.h

@@ -40,12 +40,52 @@ bool rcu_eqs_special_set(int cpu);
 void rcu_momentary_dyntick_idle(void);
 void kfree_rcu_scheduler_running(void);
 bool rcu_gp_might_be_stalled(void);
+
+struct rcu_gp_oldstate {
+	unsigned long rgos_norm;
+	unsigned long rgos_exp;
+};
+
+// Maximum number of rcu_gp_oldstate values corresponding to
+// not-yet-completed RCU grace periods.
+#define NUM_ACTIVE_RCU_POLL_FULL_OLDSTATE 4
+
+/**
+ * same_state_synchronize_rcu_full - Are two old-state values identical?
+ * @rgosp1: First old-state value.
+ * @rgosp2: Second old-state value.
+ *
+ * The two old-state values must have been obtained from either
+ * get_state_synchronize_rcu_full(), start_poll_synchronize_rcu_full(),
+ * or get_completed_synchronize_rcu_full().  Returns @true if the two
+ * values are identical and @false otherwise.  This allows structures
+ * whose lifetimes are tracked by old-state values to push these values
+ * to a list header, allowing those structures to be slightly smaller.
+ *
+ * Note that equality is judged on a bitwise basis, so that an
+ * @rcu_gp_oldstate structure with an already-completed state in one field
+ * will compare not-equal to a structure with an already-completed state
+ * in the other field.  After all, the @rcu_gp_oldstate structure is opaque
+ * so how did such a situation come to pass in the first place?
+ */
+static inline bool same_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp1,
+						   struct rcu_gp_oldstate *rgosp2)
+{
+	return rgosp1->rgos_norm == rgosp2->rgos_norm && rgosp1->rgos_exp == rgosp2->rgos_exp;
+}
+
 unsigned long start_poll_synchronize_rcu_expedited(void);
+void start_poll_synchronize_rcu_expedited_full(struct rcu_gp_oldstate *rgosp);
 void cond_synchronize_rcu_expedited(unsigned long oldstate);
+void cond_synchronize_rcu_expedited_full(struct rcu_gp_oldstate *rgosp);
 unsigned long get_state_synchronize_rcu(void);
+void get_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp);
 unsigned long start_poll_synchronize_rcu(void);
+void start_poll_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp);
 bool poll_state_synchronize_rcu(unsigned long oldstate);
+bool poll_state_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp);
 void cond_synchronize_rcu(unsigned long oldstate);
+void cond_synchronize_rcu_full(struct rcu_gp_oldstate *rgosp);
 
 bool rcu_is_idle_cpu(int cpu);
 

include/linux/srcutiny.h

@@ -15,10 +15,10 @@
 
 struct srcu_struct {
 	short srcu_lock_nesting[2];	/* srcu_read_lock() nesting depth. */
-	unsigned short srcu_idx;	/* Current reader array element in bit 0x2. */
-	unsigned short srcu_idx_max;	/* Furthest future srcu_idx request. */
 	u8 srcu_gp_running;		/* GP workqueue running? */
 	u8 srcu_gp_waiting;		/* GP waiting for readers? */
+	unsigned long srcu_idx;		/* Current reader array element in bit 0x2. */
+	unsigned long srcu_idx_max;	/* Furthest future srcu_idx request. */
 	struct swait_queue_head srcu_wq;
 					/* Last srcu_read_unlock() wakes GP. */
 	struct rcu_head *srcu_cb_head;	/* Pending callbacks: Head. */
@@ -82,10 +82,12 @@ static inline void srcu_torture_stats_print(struct srcu_struct *ssp,
 	int idx;
 
 	idx = ((data_race(READ_ONCE(ssp->srcu_idx)) + 1) & 0x2) >> 1;
-	pr_alert("%s%s Tiny SRCU per-CPU(idx=%d): (%hd,%hd)\n",
+	pr_alert("%s%s Tiny SRCU per-CPU(idx=%d): (%hd,%hd) gp: %lu->%lu\n",
 		 tt, tf, idx,
 		 data_race(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])),
+		 data_race(READ_ONCE(ssp->srcu_idx)),
+		 data_race(READ_ONCE(ssp->srcu_idx_max)));
 }
 
 #endif