Merge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull RCU updates from Ingo Molnar:
 "The changes in this cycle are:

   - RCU flavor consolidation cleanups and optmizations

   - Documentation updates

   - Miscellaneous fixes

   - SRCU updates

   - RCU-sync flavor consolidation

   - Torture-test updates

   - Linux-kernel memory-consistency-model updates, most notably the
     addition of plain C-language accesses"

* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (61 commits)
  tools/memory-model: Improve data-race detection
  tools/memory-model: Change definition of rcu-fence
  tools/memory-model: Expand definition of barrier
  tools/memory-model: Do not use "herd" to refer to "herd7"
  tools/memory-model: Fix comment in MP+poonceonces.litmus
  Documentation: atomic_t.txt: Explain ordering provided by smp_mb__{before,after}_atomic()
  rcu: Don't return a value from rcu_assign_pointer()
  rcu: Force inlining of rcu_read_lock()
  rcu: Fix irritating whitespace error in rcu_assign_pointer()
  rcu: Upgrade sync_exp_work_done() to smp_mb()
  rcutorture: Upper case solves the case of the vanishing NULL pointer
  torture: Suppress propagating trace_printk() warning
  rcutorture: Dump trace buffer for callback pipe drain failures
  torture: Add --trust-make to suppress "make clean"
  torture: Make --cpus override idleness calculations
  torture: Run kernel build in source directory
  torture: Add function graph-tracing cheat sheet
  torture: Capture qemu output
  rcutorture: Tweak kvm options
  rcutorture: Add trivial RCU implementation
  ...

Documentation/RCU/rcuref.txt

@@ -12,6 +12,7 @@ please read on.
 Reference counting on elements of lists which are protected by traditional
 reader/writer spinlocks or semaphores are straightforward:
 
+CODE LISTING A:
 1.				2.
 add()				search_and_reference()
 {				{
@@ -28,7 +29,8 @@ add()				search_and_reference()
 release_referenced()			delete()
 {					{
     ...					    write_lock(&list_lock);
-    atomic_dec(&el->rc, relfunc)	    ...
+    if(atomic_dec_and_test(&el->rc))	    ...
+	kfree(el);
     ...					    remove_element
 }					    write_unlock(&list_lock);
  					    ...
@@ -44,6 +46,7 @@ search_and_reference() could potentially hold reference to an element which
 has already been deleted from the list/array.  Use atomic_inc_not_zero()
 in this scenario as follows:
 
+CODE LISTING B:
 1.					2.
 add()					search_and_reference()
 {					{
@@ -79,6 +82,7 @@ search_and_reference() code path.  In such cases, the
 atomic_dec_and_test() may be moved from delete() to el_free()
 as follows:
 
+CODE LISTING C:
 1.					2.
 add()					search_and_reference()
 {					{
@@ -114,6 +118,17 @@ element can therefore safely be freed.  This in turn guarantees that if
 any reader finds the element, that reader may safely acquire a reference
 without checking the value of the reference counter.
 
+A clear advantage of the RCU-based pattern in listing C over the one
+in listing B is that any call to search_and_reference() that locates
+a given object will succeed in obtaining a reference to that object,
+even given a concurrent invocation of delete() for that same object.
+Similarly, a clear advantage of both listings B and C over listing A is
+that a call to delete() is not delayed even if there are an arbitrarily
+large number of calls to search_and_reference() searching for the same
+object that delete() was invoked on.  Instead, all that is delayed is
+the eventual invocation of kfree(), which is usually not a problem on
+modern computer systems, even the small ones.
+
 In cases where delete() can sleep, synchronize_rcu() can be called from
 delete(), so that el_free() can be subsumed into delete as follows:
 
@@ -130,3 +145,7 @@ delete()
     	kfree(el);
     ...
 }
+
+As additional examples in the kernel, the pattern in listing C is used by
+reference counting of struct pid, while the pattern in listing B is used by
+struct posix_acl.

Documentation/RCU/stallwarn.txt

@@ -153,7 +153,7 @@ rcupdate.rcu_task_stall_timeout
 	This boot/sysfs parameter controls the RCU-tasks stall warning
 	interval.  A value of zero or less suppresses RCU-tasks stall
 	warnings.  A positive value sets the stall-warning interval
-	in jiffies.  An RCU-tasks stall warning starts with the line:
+	in seconds.  An RCU-tasks stall warning starts with the line:
 
 		INFO: rcu_tasks detected stalls on tasks:
 

Documentation/RCU/whatisRCU.txt

@@ -212,7 +212,7 @@ synchronize_rcu()
 
 rcu_assign_pointer()
 
-	typeof(p) rcu_assign_pointer(p, typeof(p) v);
+	void rcu_assign_pointer(p, typeof(p) v);
 
 	Yes, rcu_assign_pointer() -is- implemented as a macro, though it
 	would be cool to be able to declare a function in this manner.
@@ -220,9 +220,9 @@ rcu_assign_pointer()
 
 	The updater uses this function to assign a new value to an
 	RCU-protected pointer, in order to safely communicate the change
-	in value from the updater to the reader.  This function returns
+	in value from the updater to the reader.  This macro does not
-	the new value, and also executes any memory-barrier instructions
+	evaluate to an rvalue, but it does execute any memory-barrier
-	required for a given CPU architecture.
+	instructions required for a given CPU architecture.
 
 	Perhaps just as important, it serves to document (1) which
 	pointers are protected by RCU and (2) the point at which a

Documentation/admin-guide/kernel-parameters.txt

@@ -3752,6 +3752,12 @@
 			the propagation of recent CPU-hotplug changes up
 			the rcu_node combining tree.
 
+	rcutree.use_softirq=	[KNL]
+			If set to zero, move all RCU_SOFTIRQ processing to
+			per-CPU rcuc kthreads.  Defaults to a non-zero
+			value, meaning that RCU_SOFTIRQ is used by default.
+			Specify rcutree.use_softirq=0 to use rcuc kthreads.
+
 	rcutree.rcu_fanout_exact= [KNL]
 			Disable autobalancing of the rcu_node combining
 			tree.  This is used by rcutorture, and might

Documentation/atomic_t.txt

@@ -187,8 +187,14 @@ The barriers:
 
   smp_mb__{before,after}_atomic()
 
-only apply to the RMW ops and can be used to augment/upgrade the ordering
+only apply to the RMW atomic ops and can be used to augment/upgrade the
-inherent to the used atomic op. These barriers provide a full smp_mb().
+ordering inherent to the op. These barriers act almost like a full smp_mb():
+smp_mb__before_atomic() orders all earlier accesses against the RMW op
+itself and all accesses following it, and smp_mb__after_atomic() orders all
+later accesses against the RMW op and all accesses preceding it. However,
+accesses between the smp_mb__{before,after}_atomic() and the RMW op are not
+ordered, so it is advisable to place the barrier right next to the RMW atomic
+op whenever possible.
 
 These helper barriers exist because architectures have varying implicit
 ordering on their SMP atomic primitives. For example our TSO architectures
@@ -212,7 +218,9 @@ Further, while something like:
   atomic_dec(&X);
 
 is a 'typical' RELEASE pattern, the barrier is strictly stronger than
-a RELEASE. Similarly for something like:
+a RELEASE because it orders preceding instructions against both the read
+and write parts of the atomic_dec(), and against all following instructions
+as well. Similarly, something like:
 
   atomic_inc(&X);
   smp_mb__after_atomic();
@@ -244,7 +252,8 @@ strictly stronger than ACQUIRE. As illustrated:
 
 This should not happen; but a hypothetical atomic_inc_acquire() --
 (void)atomic_fetch_inc_acquire() for instance -- would allow the outcome,
-since then:
+because it would not order the W part of the RMW against the following
+WRITE_ONCE.  Thus:
 
   P1			P2
 

Documentation/core-api/circular-buffers.rst

@@ -3,7 +3,7 @@ Circular Buffers
 ================
 
 :Author: David Howells <dhowells@redhat.com>
-:Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+:Author: Paul E. McKenney <paulmck@linux.ibm.com>
 
 
 Linux provides a number of features that can be used to implement circular

Documentation/memory-barriers.txt

@@ -3,7 +3,7 @@
 			 ============================
 
 By: David Howells <dhowells@redhat.com>
-    Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+    Paul E. McKenney <paulmck@linux.ibm.com>
     Will Deacon <will.deacon@arm.com>
     Peter Zijlstra <peterz@infradead.org>
 

Documentation/translations/ko_KR/memory-barriers.txt

@@ -24,7 +24,7 @@ Documentation/memory-barriers.txt
 			 =========================
 
 저자: David Howells <dhowells@redhat.com>
-      Paul E. McKenney <paulmck@linux.vnet.ibm.com>
+      Paul E. McKenney <paulmck@linux.ibm.com>
       Will Deacon <will.deacon@arm.com>
       Peter Zijlstra <peterz@infradead.org>
 

include/linux/lockdep.h

@@ -632,11 +632,18 @@ do {									\
 			  "IRQs not disabled as expected\n");		\
 	} while (0)
 
+#define lockdep_assert_in_irq() do {					\
+		WARN_ONCE(debug_locks && !current->lockdep_recursion &&	\
+			  !current->hardirq_context,			\
+			  "Not in hardirq as expected\n");		\
+	} while (0)
+
 #else
 # define might_lock(lock) do { } while (0)
 # define might_lock_read(lock) do { } while (0)
 # define lockdep_assert_irqs_enabled() do { } while (0)
 # define lockdep_assert_irqs_disabled() do { } while (0)
+# define lockdep_assert_in_irq() do { } while (0)
 #endif
 
 #ifdef CONFIG_LOCKDEP

include/linux/module.h

@@ -21,6 +21,7 @@
 #include <linux/rbtree_latch.h>
 #include <linux/error-injection.h>
 #include <linux/tracepoint-defs.h>
+#include <linux/srcu.h>
 
 #include <linux/percpu.h>
 #include <asm/module.h>
@@ -450,6 +451,10 @@ struct module {
 	unsigned int num_tracepoints;
 	tracepoint_ptr_t *tracepoints_ptrs;
 #endif
+#ifdef CONFIG_TREE_SRCU
+	unsigned int num_srcu_structs;
+	struct srcu_struct **srcu_struct_ptrs;
+#endif
 #ifdef CONFIG_BPF_EVENTS
 	unsigned int num_bpf_raw_events;
 	struct bpf_raw_event_map *bpf_raw_events;

include/linux/percpu-rwsem.h

@@ -17,14 +17,18 @@ struct percpu_rw_semaphore {
 	int			readers_block;
 };
 
-#define DEFINE_STATIC_PERCPU_RWSEM(name)				\
+#define __DEFINE_PERCPU_RWSEM(name, is_static)				\
 static DEFINE_PER_CPU(unsigned int, __percpu_rwsem_rc_##name);		\
-static struct percpu_rw_semaphore name = {				\
+is_static struct percpu_rw_semaphore name = {				\
-	.rss = __RCU_SYNC_INITIALIZER(name.rss, RCU_SCHED_SYNC),	\
+	.rss = __RCU_SYNC_INITIALIZER(name.rss),			\
 	.read_count = &__percpu_rwsem_rc_##name,			\
 	.rw_sem = __RWSEM_INITIALIZER(name.rw_sem),			\
 	.writer = __RCUWAIT_INITIALIZER(name.writer),			\
 }
+#define DEFINE_PERCPU_RWSEM(name)		\
+	__DEFINE_PERCPU_RWSEM(name, /* not static */)
+#define DEFINE_STATIC_PERCPU_RWSEM(name)	\
+	__DEFINE_PERCPU_RWSEM(name, static)
 
 extern int __percpu_down_read(struct percpu_rw_semaphore *, int);
 extern void __percpu_up_read(struct percpu_rw_semaphore *);

include/linux/rcu_sync.h

@@ -13,62 +13,44 @@
 #include <linux/wait.h>
 #include <linux/rcupdate.h>
 
-enum rcu_sync_type { RCU_SYNC, RCU_SCHED_SYNC, RCU_BH_SYNC };
-
 /* Structure to mediate between updaters and fastpath-using readers.  */
 struct rcu_sync {
 	int			gp_state;
 	int			gp_count;
 	wait_queue_head_t	gp_wait;
 
-	int			cb_state;
 	struct rcu_head		cb_head;
-
-	enum rcu_sync_type	gp_type;
 };
 
-extern void rcu_sync_lockdep_assert(struct rcu_sync *);
-
 /**
  * rcu_sync_is_idle() - Are readers permitted to use their fastpaths?
  * @rsp: Pointer to rcu_sync structure to use for synchronization
  *
- * Returns true if readers are permitted to use their fastpaths.
+ * Returns true if readers are permitted to use their fastpaths.  Must be
- * Must be invoked within an RCU read-side critical section whose
+ * invoked within some flavor of RCU read-side critical section.
- * flavor matches that of the rcu_sync struture.
  */
 static inline bool rcu_sync_is_idle(struct rcu_sync *rsp)
 {
-#ifdef CONFIG_PROVE_RCU
+	RCU_LOCKDEP_WARN(!rcu_read_lock_held() &&
-	rcu_sync_lockdep_assert(rsp);
+			 !rcu_read_lock_bh_held() &&
-#endif
+			 !rcu_read_lock_sched_held(),
-	return !rsp->gp_state; /* GP_IDLE */
+			 "suspicious rcu_sync_is_idle() usage");
+	return !READ_ONCE(rsp->gp_state); /* GP_IDLE */
 }
 
-extern void rcu_sync_init(struct rcu_sync *, enum rcu_sync_type);
+extern void rcu_sync_init(struct rcu_sync *);
 extern void rcu_sync_enter_start(struct rcu_sync *);
 extern void rcu_sync_enter(struct rcu_sync *);
 extern void rcu_sync_exit(struct rcu_sync *);
 extern void rcu_sync_dtor(struct rcu_sync *);
 
-#define __RCU_SYNC_INITIALIZER(name, type) {				\
+#define __RCU_SYNC_INITIALIZER(name) {					\
 		.gp_state = 0,						\
 		.gp_count = 0,						\
 		.gp_wait = __WAIT_QUEUE_HEAD_INITIALIZER(name.gp_wait),	\
-		.cb_state = 0,						\
-		.gp_type = type,					\
 	}
 
-#define	__DEFINE_RCU_SYNC(name, type)	\
+#define	DEFINE_RCU_SYNC(name)	\
-	struct rcu_sync_struct name = __RCU_SYNC_INITIALIZER(name, type)
+	struct rcu_sync name = __RCU_SYNC_INITIALIZER(name)
-
-#define DEFINE_RCU_SYNC(name)		\
-	__DEFINE_RCU_SYNC(name, RCU_SYNC)
-
-#define DEFINE_RCU_SCHED_SYNC(name)	\
-	__DEFINE_RCU_SYNC(name, RCU_SCHED_SYNC)
-
-#define DEFINE_RCU_BH_SYNC(name)	\
-	__DEFINE_RCU_SYNC(name, RCU_BH_SYNC)
 
 #endif /* _LINUX_RCU_SYNC_H_ */

include/linux/rcupdate.h

@@ -365,16 +365,15 @@ static inline void rcu_preempt_sleep_check(void) { }
  * other macros that it invokes.
  */
 #define rcu_assign_pointer(p, v)					      \
-({									      \
+do {									      \
 	uintptr_t _r_a_p__v = (uintptr_t)(v);				      \
-	rcu_check_sparse(p, __rcu);				      \
+	rcu_check_sparse(p, __rcu);					      \
 									      \
 	if (__builtin_constant_p(v) && (_r_a_p__v) == (uintptr_t)NULL)	      \
 		WRITE_ONCE((p), (typeof(p))(_r_a_p__v));		      \
 	else								      \
 		smp_store_release(&p, RCU_INITIALIZER((typeof(p))_r_a_p__v)); \
-	_r_a_p__v;							      \
+} while (0)
-})
 
 /**
  * rcu_swap_protected() - swap an RCU and a regular pointer
@@ -586,7 +585,7 @@ static inline void rcu_preempt_sleep_check(void) { }
  * read-side critical sections may be preempted and they may also block, but
  * only when acquiring spinlocks that are subject to priority inheritance.
  */
-static inline void rcu_read_lock(void)
+static __always_inline void rcu_read_lock(void)
 {
 	__rcu_read_lock();
 	__acquire(RCU);
@@ -803,7 +802,7 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
 /**
  * kfree_rcu() - kfree an object after a grace period.
  * @ptr:	pointer to kfree
- * @rcu_head:	the name of the struct rcu_head within the type of @ptr.
+ * @rhf:	the name of the struct rcu_head within the type of @ptr.
  *
  * Many rcu callbacks functions just call kfree() on the base structure.
  * These functions are trivial, but their size adds up, and furthermore
@@ -826,9 +825,13 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
  * The BUILD_BUG_ON check must not involve any function calls, hence the
  * checks are done in macros here.
  */
-#define kfree_rcu(ptr, rcu_head)					\
+#define kfree_rcu(ptr, rhf)						\
-	__kfree_rcu(&((ptr)->rcu_head), offsetof(typeof(*(ptr)), rcu_head))
+do {									\
-
+	typeof (ptr) ___p = (ptr);					\
+									\
+	if (___p)							\
+		__kfree_rcu(&((___p)->rhf), offsetof(typeof(*(ptr)), rhf)); \
+} while (0)
 
 /*
  * Place this after a lock-acquisition primitive to guarantee that

include/linux/sched.h

@@ -565,7 +565,7 @@ union rcu_special {
 		u8			blocked;
 		u8			need_qs;
 		u8			exp_hint; /* Hint for performance. */
-		u8			pad; /* No garbage from compiler! */
+		u8			deferred_qs;
 	} b; /* Bits. */
 	u32 s; /* Set of bits. */
 };

include/linux/srcutree.h

@@ -120,9 +120,17 @@ struct srcu_struct {
  *
  * See include/linux/percpu-defs.h for the rules on per-CPU variables.
  */
-#define __DEFINE_SRCU(name, is_static)					\
+#ifdef MODULE
-	static DEFINE_PER_CPU(struct srcu_data, name##_srcu_data);\
+# define __DEFINE_SRCU(name, is_static)					\
-	is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name, name##_srcu_data)
+	is_static struct srcu_struct name;				\
+	struct srcu_struct * const __srcu_struct_##name			\
+		__section("___srcu_struct_ptrs") = &name
+#else
+# define __DEFINE_SRCU(name, is_static)					\
+	static DEFINE_PER_CPU(struct srcu_data, name##_srcu_data);	\
+	is_static struct srcu_struct name =				\
+		__SRCU_STRUCT_INIT(name, name##_srcu_data)
+#endif
 #define DEFINE_SRCU(name)		__DEFINE_SRCU(name, /* not static */)
 #define DEFINE_STATIC_SRCU(name)	__DEFINE_SRCU(name, static)
 

include/linux/torture.h

@@ -66,7 +66,7 @@ int torture_shutdown_init(int ssecs, void (*cleanup)(void));
 
 /* Task stuttering, which forces load/no-load transitions. */
 bool stutter_wait(const char *title);
-int torture_stutter_init(int s);
+int torture_stutter_init(int s, int sgap);
 
 /* Initialization and cleanup. */
 bool torture_init_begin(char *ttype, int v);

kernel/cgroup/cgroup.c

@@ -101,7 +101,7 @@ static DEFINE_SPINLOCK(cgroup_idr_lock);
  */
 static DEFINE_SPINLOCK(cgroup_file_kn_lock);
 
-struct percpu_rw_semaphore cgroup_threadgroup_rwsem;
+DEFINE_PERCPU_RWSEM(cgroup_threadgroup_rwsem);
 
 #define cgroup_assert_mutex_or_rcu_locked()				\
 	RCU_LOCKDEP_WARN(!rcu_read_lock_held() &&			\
@@ -5666,7 +5666,6 @@ int __init cgroup_init(void)
 	int ssid;
 
 	BUILD_BUG_ON(CGROUP_SUBSYS_COUNT > 16);
-	BUG_ON(percpu_init_rwsem(&cgroup_threadgroup_rwsem));
 	BUG_ON(cgroup_init_cftypes(NULL, cgroup_base_files));
 	BUG_ON(cgroup_init_cftypes(NULL, cgroup1_base_files));
 

kernel/events/uprobes.c

@@ -46,7 +46,7 @@ static DEFINE_SPINLOCK(uprobes_treelock);	/* serialize rbtree access */
 static struct mutex uprobes_mmap_mutex[UPROBES_HASH_SZ];
 #define uprobes_mmap_hash(v)	(&uprobes_mmap_mutex[((unsigned long)(v)) % UPROBES_HASH_SZ])
 
-static struct percpu_rw_semaphore dup_mmap_sem;
+DEFINE_STATIC_PERCPU_RWSEM(dup_mmap_sem);
 
 /* Have a copy of original instruction */
 #define UPROBE_COPY_INSN	0
@@ -2302,7 +2302,5 @@ void __init uprobes_init(void)
 	for (i = 0; i < UPROBES_HASH_SZ; i++)
 		mutex_init(&uprobes_mmap_mutex[i]);
 
-	BUG_ON(percpu_init_rwsem(&dup_mmap_sem));
-
 	BUG_ON(register_die_notifier(&uprobe_exception_nb));
 }

kernel/locking/locktorture.c

@@ -975,7 +975,7 @@ static int __init lock_torture_init(void)
 			goto unwind;
 	}
 	if (stutter > 0) {
-		firsterr = torture_stutter_init(stutter);
+		firsterr = torture_stutter_init(stutter, stutter);
 		if (firsterr)
 			goto unwind;
 	}

kernel/locking/percpu-rwsem.c

@@ -18,7 +18,7 @@ int __percpu_init_rwsem(struct percpu_rw_semaphore *sem,
 		return -ENOMEM;
 
 	/* ->rw_sem represents the whole percpu_rw_semaphore for lockdep */
-	rcu_sync_init(&sem->rss, RCU_SCHED_SYNC);
+	rcu_sync_init(&sem->rss);
 	__init_rwsem(&sem->rw_sem, name, rwsem_key);
 	rcuwait_init(&sem->writer);
 	sem->readers_block = 0;

kernel/module.c

@@ -3083,6 +3083,11 @@ static int find_module_sections(struct module *mod, struct load_info *info)
 					     sizeof(*mod->tracepoints_ptrs),
 					     &mod->num_tracepoints);
 #endif
+#ifdef CONFIG_TREE_SRCU
+	mod->srcu_struct_ptrs = section_objs(info, "___srcu_struct_ptrs",
+					     sizeof(*mod->srcu_struct_ptrs),
+					     &mod->num_srcu_structs);
+#endif
 #ifdef CONFIG_BPF_EVENTS
 	mod->bpf_raw_events = section_objs(info, "__bpf_raw_tp_map",
 					   sizeof(*mod->bpf_raw_events),

kernel/rcu/rcu.h

@@ -446,6 +446,7 @@ void rcu_request_urgent_qs_task(struct task_struct *t);
 enum rcutorture_type {
 	RCU_FLAVOR,
 	RCU_TASKS_FLAVOR,
+	RCU_TRIVIAL_FLAVOR,
 	SRCU_FLAVOR,
 	INVALID_RCU_FLAVOR
 };
@@ -479,6 +480,10 @@ void do_trace_rcu_torture_read(const char *rcutorturename,
 #endif
 #endif
 
+#if IS_ENABLED(CONFIG_RCU_TORTURE_TEST) || IS_MODULE(CONFIG_RCU_TORTURE_TEST)
+long rcutorture_sched_setaffinity(pid_t pid, const struct cpumask *in_mask);
+#endif
+
 #ifdef CONFIG_TINY_SRCU
 
 static inline void srcutorture_get_gp_data(enum rcutorture_type test_type,

kernel/rcu/rcutorture.c

@@ -299,6 +299,7 @@ struct rcu_torture_ops {
 	int irq_capable;
 	int can_boost;
 	int extendables;
+	int slow_gps;
 	const char *name;
 };
 
@@ -667,9 +668,51 @@ static struct rcu_torture_ops tasks_ops = {
 	.fqs		= NULL,
 	.stats		= NULL,
 	.irq_capable	= 1,
+	.slow_gps	= 1,
 	.name		= "tasks"
 };
 
+/*
+ * Definitions for trivial CONFIG_PREEMPT=n-only torture testing.
+ * This implementation does not necessarily work well with CPU hotplug.
+ */
+
+static void synchronize_rcu_trivial(void)
+{
+	int cpu;
+
+	for_each_online_cpu(cpu) {
+		rcutorture_sched_setaffinity(current->pid, cpumask_of(cpu));
+		WARN_ON_ONCE(raw_smp_processor_id() != cpu);
+	}
+}
+
+static int rcu_torture_read_lock_trivial(void) __acquires(RCU)
+{
+	preempt_disable();
+	return 0;
+}
+
+static void rcu_torture_read_unlock_trivial(int idx) __releases(RCU)
+{
+	preempt_enable();
+}
+
+static struct rcu_torture_ops trivial_ops = {
+	.ttype		= RCU_TRIVIAL_FLAVOR,
+	.init		= rcu_sync_torture_init,
+	.readlock	= rcu_torture_read_lock_trivial,
+	.read_delay	= rcu_read_delay,  /* just reuse rcu's version. */
+	.readunlock	= rcu_torture_read_unlock_trivial,
+	.get_gp_seq	= rcu_no_completed,
+	.sync		= synchronize_rcu_trivial,
+	.exp_sync	= synchronize_rcu_trivial,
+	.fqs		= NULL,
+	.stats		= NULL,
+	.irq_capable	= 1,
+	.name		= "trivial"
+};
+
 static unsigned long rcutorture_seq_diff(unsigned long new, unsigned long old)
 {
 	if (!cur_ops->gp_diff)
@@ -1010,10 +1053,17 @@ rcu_torture_writer(void *arg)
 				       !rcu_gp_is_normal();
 		}
 		rcu_torture_writer_state = RTWS_STUTTER;
-		if (stutter_wait("rcu_torture_writer"))
+		if (stutter_wait("rcu_torture_writer") &&
+		    !READ_ONCE(rcu_fwd_cb_nodelay) &&
+		    !cur_ops->slow_gps &&
+		    !torture_must_stop())
 			for (i = 0; i < ARRAY_SIZE(rcu_tortures); i++)
-				if (list_empty(&rcu_tortures[i].rtort_free))
+				if (list_empty(&rcu_tortures[i].rtort_free) &&
-					WARN_ON_ONCE(1);
+				    rcu_access_pointer(rcu_torture_current) !=
+				    &rcu_tortures[i]) {
+					rcu_ftrace_dump(DUMP_ALL);
+					WARN(1, "%s: rtort_pipe_count: %d\n", __func__, rcu_tortures[i].rtort_pipe_count);
+				}
 	} while (!torture_must_stop());
 	/* Reset expediting back to unexpedited. */
 	if (expediting > 0)
@@ -1358,8 +1408,9 @@ rcu_torture_stats_print(void)
 	}
 
 	pr_alert("%s%s ", torture_type, TORTURE_FLAG);
-	pr_cont("rtc: %p ver: %lu tfle: %d rta: %d rtaf: %d rtf: %d ",
+	pr_cont("rtc: %p %s: %lu tfle: %d rta: %d rtaf: %d rtf: %d ",
 		rcu_torture_current,
+		rcu_torture_current ? "ver" : "VER",
 		rcu_torture_current_version,
 		list_empty(&rcu_torture_freelist),
 		atomic_read(&n_rcu_torture_alloc),
@@ -1661,6 +1712,17 @@ static void rcu_torture_fwd_cb_cr(struct rcu_head *rhp)
 	spin_unlock_irqrestore(&rcu_fwd_lock, flags);
 }
 
+// Give the scheduler a chance, even on nohz_full CPUs.
+static void rcu_torture_fwd_prog_cond_resched(void)
+{
+	if (IS_ENABLED(CONFIG_PREEMPT) && IS_ENABLED(CONFIG_NO_HZ_FULL)) {
+		if (need_resched())
+			schedule();
+	} else {
+		cond_resched();
+	}
+}
+
 /*
  * Free all callbacks on the rcu_fwd_cb_head list, either because the
  * test is over or because we hit an OOM event.
@@ -1674,16 +1736,18 @@ static unsigned long rcu_torture_fwd_prog_cbfree(void)
 	for (;;) {
 		spin_lock_irqsave(&rcu_fwd_lock, flags);
 		rfcp = rcu_fwd_cb_head;
-		if (!rfcp)
+		if (!rfcp) {
+			spin_unlock_irqrestore(&rcu_fwd_lock, flags);
 			break;
+		}
 		rcu_fwd_cb_head = rfcp->rfc_next;
 		if (!rcu_fwd_cb_head)
 			rcu_fwd_cb_tail = &rcu_fwd_cb_head;
 		spin_unlock_irqrestore(&rcu_fwd_lock, flags);
 		kfree(rfcp);
 		freed++;
+		rcu_torture_fwd_prog_cond_resched();
 	}
-	spin_unlock_irqrestore(&rcu_fwd_lock, flags);
 	return freed;
 }
 
@@ -1707,6 +1771,8 @@ static void rcu_torture_fwd_prog_nr(int *tested, int *tested_tries)
 	}
 
 	/* Tight loop containing cond_resched(). */
+	WRITE_ONCE(rcu_fwd_cb_nodelay, true);
+	cur_ops->sync(); /* Later readers see above write. */
 	if  (selfpropcb) {
 		WRITE_ONCE(fcs.stop, 0);
 		cur_ops->call(&fcs.rh, rcu_torture_fwd_prog_cb);
@@ -1724,7 +1790,7 @@ static void rcu_torture_fwd_prog_nr(int *tested, int *tested_tries)
 		udelay(10);
 		cur_ops->readunlock(idx);
 		if (!fwd_progress_need_resched || need_resched())
-			cond_resched();
+			rcu_torture_fwd_prog_cond_resched();
 	}
 	(*tested_tries)++;
 	if (!time_before(jiffies, stopat) &&
@@ -1745,6 +1811,8 @@ static void rcu_torture_fwd_prog_nr(int *tested, int *tested_tries)
 		WARN_ON(READ_ONCE(fcs.stop) != 2);
 		destroy_rcu_head_on_stack(&fcs.rh);
 	}
+	schedule_timeout_uninterruptible(HZ / 10); /* Let kthreads recover. */
+	WRITE_ONCE(rcu_fwd_cb_nodelay, false);
 }
 
 /* Carry out call_rcu() forward-progress testing. */
@@ -1765,6 +1833,8 @@ static void rcu_torture_fwd_prog_cr(void)
 
 	if (READ_ONCE(rcu_fwd_emergency_stop))
 		return; /* Get out of the way quickly, no GP wait! */
+	if (!cur_ops->call)
+		return; /* Can't do call_rcu() fwd prog without ->call. */
 
 	/* Loop continuously posting RCU callbacks. */
 	WRITE_ONCE(rcu_fwd_cb_nodelay, true);
@@ -1805,7 +1875,7 @@ static void rcu_torture_fwd_prog_cr(void)
 			rfcp->rfc_gps = 0;
 		}
 		cur_ops->call(&rfcp->rh, rcu_torture_fwd_cb_cr);
-		cond_resched();
+		rcu_torture_fwd_prog_cond_resched();
 	}
 	stoppedat = jiffies;
 	n_launders_cb_snap = READ_ONCE(n_launders_cb);
@@ -1814,7 +1884,6 @@ static void rcu_torture_fwd_prog_cr(void)
 	cur_ops->cb_barrier(); /* Wait for callbacks to be invoked. */
 	(void)rcu_torture_fwd_prog_cbfree();
 
-	WRITE_ONCE(rcu_fwd_cb_nodelay, false);
 	if (!torture_must_stop() && !READ_ONCE(rcu_fwd_emergency_stop)) {
 		WARN_ON(n_max_gps < MIN_FWD_CBS_LAUNDERED);
 		pr_alert("%s Duration %lu barrier: %lu pending %ld n_launders: %ld n_launders_sa: %ld n_max_gps: %ld n_max_cbs: %ld cver %ld gps %ld\n",
@@ -1825,6 +1894,8 @@ static void rcu_torture_fwd_prog_cr(void)
 			 n_max_gps, n_max_cbs, cver, gps);
 		rcu_torture_fwd_cb_hist();
 	}
+	schedule_timeout_uninterruptible(HZ); /* Let CBs drain. */
+	WRITE_ONCE(rcu_fwd_cb_nodelay, false);
 }
 
 
@@ -2240,7 +2311,7 @@ rcu_torture_init(void)
 	int firsterr = 0;
 	static struct rcu_torture_ops *torture_ops[] = {
 		&rcu_ops, &rcu_busted_ops, &srcu_ops, &srcud_ops,
-		&busted_srcud_ops, &tasks_ops,
+		&busted_srcud_ops, &tasks_ops, &trivial_ops,
 	};
 
 	if (!torture_init_begin(torture_type, verbose))
@@ -2363,7 +2434,10 @@ rcu_torture_init(void)
 	if (stutter < 0)
 		stutter = 0;
 	if (stutter) {
-		firsterr = torture_stutter_init(stutter * HZ);
+		int t;
+
+		t = cur_ops->stall_dur ? cur_ops->stall_dur() : stutter * HZ;
+		firsterr = torture_stutter_init(stutter * HZ, t);
 		if (firsterr)
 			goto unwind;
 	}

kernel/rcu/srcutree.c

@@ -831,8 +831,8 @@ static void srcu_leak_callback(struct rcu_head *rhp)
  * srcu_read_lock(), and srcu_read_unlock() that are all passed the same
  * srcu_struct structure.
  */
-void __call_srcu(struct srcu_struct *ssp, struct rcu_head *rhp,
+static void __call_srcu(struct srcu_struct *ssp, struct rcu_head *rhp,
-		 rcu_callback_t func, bool do_norm)
+			rcu_callback_t func, bool do_norm)
 {
 	unsigned long flags;
 	int idx;
@@ -1310,3 +1310,68 @@ void __init srcu_init(void)
 		queue_work(rcu_gp_wq, &ssp->work.work);
 	}
 }
+
+#ifdef CONFIG_MODULES
+
+/* Initialize any global-scope srcu_struct structures used by this module. */
+static int srcu_module_coming(struct module *mod)
+{
+	int i;
+	struct srcu_struct **sspp = mod->srcu_struct_ptrs;
+	int ret;
+
+	for (i = 0; i < mod->num_srcu_structs; i++) {
+		ret = init_srcu_struct(*(sspp++));
+		if (WARN_ON_ONCE(ret))
+			return ret;
+	}
+	return 0;
+}
+
+/* Clean up any global-scope srcu_struct structures used by this module. */
+static void srcu_module_going(struct module *mod)
+{
+	int i;
+	struct srcu_struct **sspp = mod->srcu_struct_ptrs;
+
+	for (i = 0; i < mod->num_srcu_structs; i++)
+		cleanup_srcu_struct(*(sspp++));
+}
+
+/* Handle one module, either coming or going. */
+static int srcu_module_notify(struct notifier_block *self,
+			      unsigned long val, void *data)
+{
+	struct module *mod = data;
+	int ret = 0;
+
+	switch (val) {
+	case MODULE_STATE_COMING:
+		ret = srcu_module_coming(mod);
+		break;
+	case MODULE_STATE_GOING:
+		srcu_module_going(mod);
+		break;
+	default:
+		break;
+	}
+	return ret;
+}
+
+static struct notifier_block srcu_module_nb = {
+	.notifier_call = srcu_module_notify,
+	.priority = 0,
+};
+
+static __init int init_srcu_module_notifier(void)
+{
+	int ret;
+
+	ret = register_module_notifier(&srcu_module_nb);
+	if (ret)
+		pr_warn("Failed to register srcu module notifier\n");
+	return ret;
+}
+late_initcall(init_srcu_module_notifier);
+
+#endif /* #ifdef CONFIG_MODULES */

kernel/rcu/sync.c

@@ -10,65 +10,18 @@
 #include <linux/rcu_sync.h>
 #include <linux/sched.h>
 
-#ifdef CONFIG_PROVE_RCU
+enum { GP_IDLE = 0, GP_ENTER, GP_PASSED, GP_EXIT, GP_REPLAY };
-#define __INIT_HELD(func)	.held = func,
-#else
-#define __INIT_HELD(func)
-#endif
-
-static const struct {
-	void (*sync)(void);
-	void (*call)(struct rcu_head *, void (*)(struct rcu_head *));
-	void (*wait)(void);
-#ifdef CONFIG_PROVE_RCU
-	int  (*held)(void);
-#endif
-} gp_ops[] = {
-	[RCU_SYNC] = {
-		.sync = synchronize_rcu,
-		.call = call_rcu,
-		.wait = rcu_barrier,
-		__INIT_HELD(rcu_read_lock_held)
-	},
-	[RCU_SCHED_SYNC] = {
-		.sync = synchronize_rcu,
-		.call = call_rcu,
-		.wait = rcu_barrier,
-		__INIT_HELD(rcu_read_lock_sched_held)
-	},
-	[RCU_BH_SYNC] = {
-		.sync = synchronize_rcu,
-		.call = call_rcu,
-		.wait = rcu_barrier,
-		__INIT_HELD(rcu_read_lock_bh_held)
-	},
-};
-
-enum { GP_IDLE = 0, GP_PENDING, GP_PASSED };
-enum { CB_IDLE = 0, CB_PENDING, CB_REPLAY };
 
 #define	rss_lock	gp_wait.lock
 
-#ifdef CONFIG_PROVE_RCU
-void rcu_sync_lockdep_assert(struct rcu_sync *rsp)
-{
-	RCU_LOCKDEP_WARN(!gp_ops[rsp->gp_type].held(),
-			 "suspicious rcu_sync_is_idle() usage");
-}
-
-EXPORT_SYMBOL_GPL(rcu_sync_lockdep_assert);
-#endif
-
 /**
  * rcu_sync_init() - Initialize an rcu_sync structure
  * @rsp: Pointer to rcu_sync structure to be initialized
- * @type: Flavor of RCU with which to synchronize rcu_sync structure
  */
-void rcu_sync_init(struct rcu_sync *rsp, enum rcu_sync_type type)
+void rcu_sync_init(struct rcu_sync *rsp)
 {
 	memset(rsp, 0, sizeof(*rsp));
 	init_waitqueue_head(&rsp->gp_wait);
-	rsp->gp_type = type;
 }
 
 /**
@@ -86,6 +39,70 @@ void rcu_sync_enter_start(struct rcu_sync *rsp)
 	rsp->gp_state = GP_PASSED;
 }
 
+
+static void rcu_sync_func(struct rcu_head *rhp);
+
+static void rcu_sync_call(struct rcu_sync *rsp)
+{
+	call_rcu(&rsp->cb_head, rcu_sync_func);
+}
+
+/**
+ * rcu_sync_func() - Callback function managing reader access to fastpath
+ * @rhp: Pointer to rcu_head in rcu_sync structure to use for synchronization
+ *
+ * This function is passed to call_rcu() function by rcu_sync_enter() and
+ * rcu_sync_exit(), so that it is invoked after a grace period following the
+ * that invocation of enter/exit.
+ *
+ * If it is called by rcu_sync_enter() it signals that all the readers were
+ * switched onto slow path.
+ *
+ * If it is called by rcu_sync_exit() it takes action based on events that
+ * have taken place in the meantime, so that closely spaced rcu_sync_enter()
+ * and rcu_sync_exit() pairs need not wait for a grace period.
+ *
+ * If another rcu_sync_enter() is invoked before the grace period
+ * ended, reset state to allow the next rcu_sync_exit() to let the
+ * readers back onto their fastpaths (after a grace period).  If both
+ * another rcu_sync_enter() and its matching rcu_sync_exit() are invoked
+ * before the grace period ended, re-invoke call_rcu() on behalf of that
+ * rcu_sync_exit().  Otherwise, set all state back to idle so that readers
+ * can again use their fastpaths.
+ */
+static void rcu_sync_func(struct rcu_head *rhp)
+{
+	struct rcu_sync *rsp = container_of(rhp, struct rcu_sync, cb_head);
+	unsigned long flags;
+
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_state) == GP_IDLE);
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_state) == GP_PASSED);
+
+	spin_lock_irqsave(&rsp->rss_lock, flags);
+	if (rsp->gp_count) {
+		/*
+		 * We're at least a GP after the GP_IDLE->GP_ENTER transition.
+		 */
+		WRITE_ONCE(rsp->gp_state, GP_PASSED);
+		wake_up_locked(&rsp->gp_wait);
+	} else if (rsp->gp_state == GP_REPLAY) {
+		/*
+		 * A new rcu_sync_exit() has happened; requeue the callback to
+		 * catch a later GP.
+		 */
+		WRITE_ONCE(rsp->gp_state, GP_EXIT);
+		rcu_sync_call(rsp);
+	} else {
+		/*
+		 * We're at least a GP after the last rcu_sync_exit(); eveybody
+		 * will now have observed the write side critical section.
+		 * Let 'em rip!.
+		 */
+		WRITE_ONCE(rsp->gp_state, GP_IDLE);
+	}
+	spin_unlock_irqrestore(&rsp->rss_lock, flags);
+}
+
 /**
  * rcu_sync_enter() - Force readers onto slowpath
  * @rsp: Pointer to rcu_sync structure to use for synchronization
@@ -103,84 +120,43 @@ void rcu_sync_enter_start(struct rcu_sync *rsp)
  */
 void rcu_sync_enter(struct rcu_sync *rsp)
 {
-	bool need_wait, need_sync;
+	int gp_state;
 
 	spin_lock_irq(&rsp->rss_lock);
-	need_wait = rsp->gp_count++;
+	gp_state = rsp->gp_state;
-	need_sync = rsp->gp_state == GP_IDLE;
+	if (gp_state == GP_IDLE) {
-	if (need_sync)
+		WRITE_ONCE(rsp->gp_state, GP_ENTER);
-		rsp->gp_state = GP_PENDING;
+		WARN_ON_ONCE(rsp->gp_count);
+		/*
+		 * Note that we could simply do rcu_sync_call(rsp) here and
+		 * avoid the "if (gp_state == GP_IDLE)" block below.
+		 *
+		 * However, synchronize_rcu() can be faster if rcu_expedited
+		 * or rcu_blocking_is_gp() is true.
+		 *
+		 * Another reason is that we can't wait for rcu callback if
+		 * we are called at early boot time but this shouldn't happen.
+		 */
+	}
+	rsp->gp_count++;
 	spin_unlock_irq(&rsp->rss_lock);
 
-	WARN_ON_ONCE(need_wait && need_sync);
+	if (gp_state == GP_IDLE) {
-	if (need_sync) {
-		gp_ops[rsp->gp_type].sync();
-		rsp->gp_state = GP_PASSED;
-		wake_up_all(&rsp->gp_wait);
-	} else if (need_wait) {
-		wait_event(rsp->gp_wait, rsp->gp_state == GP_PASSED);
-	} else {
 		/*
-		 * Possible when there's a pending CB from a rcu_sync_exit().
+		 * See the comment above, this simply does the "synchronous"
-		 * Nobody has yet been allowed the 'fast' path and thus we can
+		 * call_rcu(rcu_sync_func) which does GP_ENTER -> GP_PASSED.
-		 * avoid doing any sync(). The callback will get 'dropped'.
 		 */
-		WARN_ON_ONCE(rsp->gp_state != GP_PASSED);
+		synchronize_rcu();
+		rcu_sync_func(&rsp->cb_head);
+		/* Not really needed, wait_event() would see GP_PASSED. */
+		return;
 	}
+
+	wait_event(rsp->gp_wait, READ_ONCE(rsp->gp_state) >= GP_PASSED);
 }
 
 /**
- * rcu_sync_func() - Callback function managing reader access to fastpath
+ * rcu_sync_exit() - Allow readers back onto fast path after grace period
- * @rhp: Pointer to rcu_head in rcu_sync structure to use for synchronization
- *
- * This function is passed to one of the call_rcu() functions by
- * rcu_sync_exit(), so that it is invoked after a grace period following the
- * that invocation of rcu_sync_exit().  It takes action based on events that
- * have taken place in the meantime, so that closely spaced rcu_sync_enter()
- * and rcu_sync_exit() pairs need not wait for a grace period.
- *
- * If another rcu_sync_enter() is invoked before the grace period
- * ended, reset state to allow the next rcu_sync_exit() to let the
- * readers back onto their fastpaths (after a grace period).  If both
- * another rcu_sync_enter() and its matching rcu_sync_exit() are invoked
- * before the grace period ended, re-invoke call_rcu() on behalf of that
- * rcu_sync_exit().  Otherwise, set all state back to idle so that readers
- * can again use their fastpaths.
- */
-static void rcu_sync_func(struct rcu_head *rhp)
-{
-	struct rcu_sync *rsp = container_of(rhp, struct rcu_sync, cb_head);
-	unsigned long flags;
-
-	WARN_ON_ONCE(rsp->gp_state != GP_PASSED);
-	WARN_ON_ONCE(rsp->cb_state == CB_IDLE);
-
-	spin_lock_irqsave(&rsp->rss_lock, flags);
-	if (rsp->gp_count) {
-		/*
-		 * A new rcu_sync_begin() has happened; drop the callback.
-		 */
-		rsp->cb_state = CB_IDLE;
-	} else if (rsp->cb_state == CB_REPLAY) {
-		/*
-		 * A new rcu_sync_exit() has happened; requeue the callback
-		 * to catch a later GP.
-		 */
-		rsp->cb_state = CB_PENDING;
-		gp_ops[rsp->gp_type].call(&rsp->cb_head, rcu_sync_func);
-	} else {
-		/*
-		 * We're at least a GP after rcu_sync_exit(); eveybody will now
-		 * have observed the write side critical section. Let 'em rip!.
-		 */
-		rsp->cb_state = CB_IDLE;
-		rsp->gp_state = GP_IDLE;
-	}
-	spin_unlock_irqrestore(&rsp->rss_lock, flags);
-}
-
-/**
- * rcu_sync_exit() - Allow readers back onto fast patch after grace period
  * @rsp: Pointer to rcu_sync structure to use for synchronization
  *
  * This function is used by updaters who have completed, and can therefore
@@ -191,13 +167,16 @@ static void rcu_sync_func(struct rcu_head *rhp)
  */
 void rcu_sync_exit(struct rcu_sync *rsp)
 {
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_state) == GP_IDLE);
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_count) == 0);
+
 	spin_lock_irq(&rsp->rss_lock);
 	if (!--rsp->gp_count) {
-		if (rsp->cb_state == CB_IDLE) {
+		if (rsp->gp_state == GP_PASSED) {
-			rsp->cb_state = CB_PENDING;
+			WRITE_ONCE(rsp->gp_state, GP_EXIT);
-			gp_ops[rsp->gp_type].call(&rsp->cb_head, rcu_sync_func);
+			rcu_sync_call(rsp);
-		} else if (rsp->cb_state == CB_PENDING) {
+		} else if (rsp->gp_state == GP_EXIT) {
-			rsp->cb_state = CB_REPLAY;
+			WRITE_ONCE(rsp->gp_state, GP_REPLAY);
 		}
 	}
 	spin_unlock_irq(&rsp->rss_lock);
@@ -209,18 +188,19 @@ void rcu_sync_exit(struct rcu_sync *rsp)
  */
 void rcu_sync_dtor(struct rcu_sync *rsp)
 {
-	int cb_state;
+	int gp_state;
 
-	WARN_ON_ONCE(rsp->gp_count);
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_count));
+	WARN_ON_ONCE(READ_ONCE(rsp->gp_state) == GP_PASSED);
 
 	spin_lock_irq(&rsp->rss_lock);
-	if (rsp->cb_state == CB_REPLAY)
+	if (rsp->gp_state == GP_REPLAY)
-		rsp->cb_state = CB_PENDING;
+		WRITE_ONCE(rsp->gp_state, GP_EXIT);
-	cb_state = rsp->cb_state;
+	gp_state = rsp->gp_state;
 	spin_unlock_irq(&rsp->rss_lock);
 
-	if (cb_state != CB_IDLE) {
+	if (gp_state != GP_IDLE) {
-		gp_ops[rsp->gp_type].wait();
+		rcu_barrier();
-		WARN_ON_ONCE(rsp->cb_state != CB_IDLE);
+		WARN_ON_ONCE(rsp->gp_state != GP_IDLE);
 	}
 }

kernel/rcu/tree.c

@@ -51,6 +51,12 @@
 #include <linux/tick.h>
 #include <linux/sysrq.h>
 #include <linux/kprobes.h>
+#include <linux/gfp.h>
+#include <linux/oom.h>
+#include <linux/smpboot.h>
+#include <linux/jiffies.h>
+#include <linux/sched/isolation.h>
+#include "../time/tick-internal.h"
 
 #include "tree.h"
 #include "rcu.h"
@@ -92,6 +98,9 @@ struct rcu_state rcu_state = {
 /* Dump rcu_node combining tree at boot to verify correct setup. */
 static bool dump_tree;
 module_param(dump_tree, bool, 0444);
+/* By default, use RCU_SOFTIRQ instead of rcuc kthreads. */
+static bool use_softirq = 1;
+module_param(use_softirq, bool, 0444);
 /* Control rcu_node-tree auto-balancing at boot time. */
 static bool rcu_fanout_exact;
 module_param(rcu_fanout_exact, bool, 0444);
@@ -138,7 +147,6 @@ static void rcu_init_new_rnp(struct rcu_node *rnp_leaf);
 static void rcu_cleanup_dead_rnp(struct rcu_node *rnp_leaf);
 static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu);
 static void invoke_rcu_core(void);
-static void invoke_rcu_callbacks(struct rcu_data *rdp);
 static void rcu_report_exp_rdp(struct rcu_data *rdp);
 static void sync_sched_exp_online_cleanup(int cpu);
 
@@ -368,19 +376,33 @@ static void __maybe_unused rcu_momentary_dyntick_idle(void)
 }
 
 /**
- * rcu_is_cpu_rrupt_from_idle - see if idle or immediately interrupted from idle
+ * rcu_is_cpu_rrupt_from_idle - see if interrupted from idle
  *
- * If the current CPU is idle or running at a first-level (not nested)
+ * If the current CPU is idle and running at a first-level (not nested)
  * interrupt from idle, return true.  The caller must have at least
  * disabled preemption.
  */
 static int rcu_is_cpu_rrupt_from_idle(void)
 {
-	return __this_cpu_read(rcu_data.dynticks_nesting) <= 0 &&
+	/* Called only from within the scheduling-clock interrupt */
-	       __this_cpu_read(rcu_data.dynticks_nmi_nesting) <= 1;
+	lockdep_assert_in_irq();
+
+	/* Check for counter underflows */
+	RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nesting) < 0,
+			 "RCU dynticks_nesting counter underflow!");
+	RCU_LOCKDEP_WARN(__this_cpu_read(rcu_data.dynticks_nmi_nesting) <= 0,
+			 "RCU dynticks_nmi_nesting counter underflow/zero!");
+
+	/* Are we at first interrupt nesting level? */
+	if (__this_cpu_read(rcu_data.dynticks_nmi_nesting) != 1)
+		return false;
+
+	/* Does CPU appear to be idle from an RCU standpoint? */
+	return __this_cpu_read(rcu_data.dynticks_nesting) == 0;
 }
 
-#define DEFAULT_RCU_BLIMIT 10     /* Maximum callbacks per rcu_do_batch. */
+#define DEFAULT_RCU_BLIMIT 10     /* Maximum callbacks per rcu_do_batch ... */
+#define DEFAULT_MAX_RCU_BLIMIT 10000 /* ... even during callback flood. */
 static long blimit = DEFAULT_RCU_BLIMIT;
 #define DEFAULT_RCU_QHIMARK 10000 /* If this many pending, ignore blimit. */
 static long qhimark = DEFAULT_RCU_QHIMARK;
@@ -2113,7 +2135,7 @@ static void rcu_do_batch(struct rcu_data *rdp)
 
 	/* Reinstate batch limit if we have worked down the excess. */
 	count = rcu_segcblist_n_cbs(&rdp->cblist);
-	if (rdp->blimit == LONG_MAX && count <= qlowmark)
+	if (rdp->blimit >= DEFAULT_MAX_RCU_BLIMIT && count <= qlowmark)
 		rdp->blimit = blimit;
 
 	/* Reset ->qlen_last_fqs_check trigger if enough CBs have drained. */
@@ -2253,7 +2275,7 @@ void rcu_force_quiescent_state(void)
 EXPORT_SYMBOL_GPL(rcu_force_quiescent_state);
 
 /* Perform RCU core processing work for the current CPU.  */
-static __latent_entropy void rcu_core(struct softirq_action *unused)
+static __latent_entropy void rcu_core(void)
 {
 	unsigned long flags;
 	struct rcu_data *rdp = raw_cpu_ptr(&rcu_data);
@@ -2287,38 +2309,127 @@ static __latent_entropy void rcu_core(struct softirq_action *unused)
 	rcu_check_gp_start_stall(rnp, rdp, rcu_jiffies_till_stall_check());
 
 	/* If there are callbacks ready, invoke them. */
-	if (rcu_segcblist_ready_cbs(&rdp->cblist))
+	if (rcu_segcblist_ready_cbs(&rdp->cblist) &&
-		invoke_rcu_callbacks(rdp);
+	    likely(READ_ONCE(rcu_scheduler_fully_active)))
+		rcu_do_batch(rdp);
 
 	/* Do any needed deferred wakeups of rcuo kthreads. */
 	do_nocb_deferred_wakeup(rdp);
 	trace_rcu_utilization(TPS("End RCU core"));
 }
 
-/*
+static void rcu_core_si(struct softirq_action *h)
- * Schedule RCU callback invocation.  If the running implementation of RCU
- * does not support RCU priority boosting, just do a direct call, otherwise
- * wake up the per-CPU kernel kthread.  Note that because we are running
- * on the current CPU with softirqs disabled, the rcu_cpu_kthread_task
- * cannot disappear out from under us.
- */
-static void invoke_rcu_callbacks(struct rcu_data *rdp)
 {
-	if (unlikely(!READ_ONCE(rcu_scheduler_fully_active)))
+	rcu_core();
-		return;
-	if (likely(!rcu_state.boost)) {
-		rcu_do_batch(rdp);
-		return;
-	}
-	invoke_rcu_callbacks_kthread();
 }
 
+static void rcu_wake_cond(struct task_struct *t, int status)
+{
+	/*
+	 * If the thread is yielding, only wake it when this
+	 * is invoked from idle
+	 */
+	if (t && (status != RCU_KTHREAD_YIELDING || is_idle_task(current)))
+		wake_up_process(t);
+}
+
+static void invoke_rcu_core_kthread(void)
+{
+	struct task_struct *t;
+	unsigned long flags;
+
+	local_irq_save(flags);
+	__this_cpu_write(rcu_data.rcu_cpu_has_work, 1);
+	t = __this_cpu_read(rcu_data.rcu_cpu_kthread_task);
+	if (t != NULL && t != current)
+		rcu_wake_cond(t, __this_cpu_read(rcu_data.rcu_cpu_kthread_status));
+	local_irq_restore(flags);
+}
+
+/*
+ * Wake up this CPU's rcuc kthread to do RCU core processing.
+ */
 static void invoke_rcu_core(void)
 {
-	if (cpu_online(smp_processor_id()))
+	if (!cpu_online(smp_processor_id()))
+		return;
+	if (use_softirq)
 		raise_softirq(RCU_SOFTIRQ);
+	else
+		invoke_rcu_core_kthread();
 }
 
+static void rcu_cpu_kthread_park(unsigned int cpu)
+{
+	per_cpu(rcu_data.rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU;
+}
+
+static int rcu_cpu_kthread_should_run(unsigned int cpu)
+{
+	return __this_cpu_read(rcu_data.rcu_cpu_has_work);
+}
+
+/*
+ * Per-CPU kernel thread that invokes RCU callbacks.  This replaces
+ * the RCU softirq used in configurations of RCU that do not support RCU
+ * priority boosting.
+ */
+static void rcu_cpu_kthread(unsigned int cpu)
+{
+	unsigned int *statusp = this_cpu_ptr(&rcu_data.rcu_cpu_kthread_status);
+	char work, *workp = this_cpu_ptr(&rcu_data.rcu_cpu_has_work);
+	int spincnt;
+
+	for (spincnt = 0; spincnt < 10; spincnt++) {
+		trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait"));
+		local_bh_disable();
+		*statusp = RCU_KTHREAD_RUNNING;
+		local_irq_disable();
+		work = *workp;
+		*workp = 0;
+		local_irq_enable();
+		if (work)
+			rcu_core();
+		local_bh_enable();
+		if (*workp == 0) {
+			trace_rcu_utilization(TPS("End CPU kthread@rcu_wait"));
+			*statusp = RCU_KTHREAD_WAITING;
+			return;
+		}
+	}
+	*statusp = RCU_KTHREAD_YIELDING;
+	trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield"));
+	schedule_timeout_interruptible(2);
+	trace_rcu_utilization(TPS("End CPU kthread@rcu_yield"));
+	*statusp = RCU_KTHREAD_WAITING;
+}
+
+static struct smp_hotplug_thread rcu_cpu_thread_spec = {
+	.store			= &rcu_data.rcu_cpu_kthread_task,
+	.thread_should_run	= rcu_cpu_kthread_should_run,
+	.thread_fn		= rcu_cpu_kthread,
+	.thread_comm		= "rcuc/%u",
+	.setup			= rcu_cpu_kthread_setup,
+	.park			= rcu_cpu_kthread_park,
+};
+
+/*
+ * Spawn per-CPU RCU core processing kthreads.
+ */
+static int __init rcu_spawn_core_kthreads(void)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu)
+		per_cpu(rcu_data.rcu_cpu_has_work, cpu) = 0;
+	if (!IS_ENABLED(CONFIG_RCU_BOOST) && use_softirq)
+		return 0;
+	WARN_ONCE(smpboot_register_percpu_thread(&rcu_cpu_thread_spec),
+		  "%s: Could not start rcuc kthread, OOM is now expected behavior\n", __func__);
+	return 0;
+}
+early_initcall(rcu_spawn_core_kthreads);
+
 /*
  * Handle any core-RCU processing required by a call_rcu() invocation.
  */
@@ -2354,7 +2465,7 @@ static void __call_rcu_core(struct rcu_data *rdp, struct rcu_head *head,
 			rcu_accelerate_cbs_unlocked(rdp->mynode, rdp);
 		} else {
 			/* Give the grace period a kick. */
-			rdp->blimit = LONG_MAX;
+			rdp->blimit = DEFAULT_MAX_RCU_BLIMIT;
 			if (rcu_state.n_force_qs == rdp->n_force_qs_snap &&
 			    rcu_segcblist_first_pend_cb(&rdp->cblist) != head)
 				rcu_force_quiescent_state();
@@ -3355,7 +3466,8 @@ void __init rcu_init(void)
 	rcu_init_one();
 	if (dump_tree)
 		rcu_dump_rcu_node_tree();
-	open_softirq(RCU_SOFTIRQ, rcu_core);
+	if (use_softirq)
+		open_softirq(RCU_SOFTIRQ, rcu_core_si);
 
 	/*
 	 * We don't need protection against CPU-hotplug here because

kernel/rcu/tree.h

@@ -154,13 +154,15 @@ struct rcu_data {
 	bool		core_needs_qs;	/* Core waits for quiesc state. */
 	bool		beenonline;	/* CPU online at least once. */
 	bool		gpwrap;		/* Possible ->gp_seq wrap. */
-	bool		deferred_qs;	/* This CPU awaiting a deferred QS? */
+	bool		exp_deferred_qs; /* This CPU awaiting a deferred QS? */
 	struct rcu_node *mynode;	/* This CPU's leaf of hierarchy */
 	unsigned long grpmask;		/* Mask to apply to leaf qsmask. */
 	unsigned long	ticks_this_gp;	/* The number of scheduling-clock */
 					/*  ticks this CPU has handled */
 					/*  during and after the last grace */
 					/* period it is aware of. */
+	struct irq_work defer_qs_iw;	/* Obtain later scheduler attention. */
+	bool defer_qs_iw_pending;	/* Scheduler attention pending? */
 
 	/* 2) batch handling */
 	struct rcu_segcblist cblist;	/* Segmented callback list, with */
@@ -407,8 +409,8 @@ void call_rcu(struct rcu_head *head, rcu_callback_t func);
 static void dump_blkd_tasks(struct rcu_node *rnp, int ncheck);
 static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags);
 static void rcu_preempt_boost_start_gp(struct rcu_node *rnp);
-static void invoke_rcu_callbacks_kthread(void);
 static bool rcu_is_callbacks_kthread(void);
+static void rcu_cpu_kthread_setup(unsigned int cpu);
 static void __init rcu_spawn_boost_kthreads(void);
 static void rcu_prepare_kthreads(int cpu);
 static void rcu_cleanup_after_idle(void);

kernel/rcu/tree_exp.h

@@ -250,7 +250,7 @@ static void rcu_report_exp_cpu_mult(struct rcu_node *rnp,
  */
 static void rcu_report_exp_rdp(struct rcu_data *rdp)
 {
-	WRITE_ONCE(rdp->deferred_qs, false);
+	WRITE_ONCE(rdp->exp_deferred_qs, false);
 	rcu_report_exp_cpu_mult(rdp->mynode, rdp->grpmask, true);
 }
 
@@ -259,8 +259,7 @@ static bool sync_exp_work_done(unsigned long s)
 {
 	if (rcu_exp_gp_seq_done(s)) {
 		trace_rcu_exp_grace_period(rcu_state.name, s, TPS("done"));
-		/* Ensure test happens before caller kfree(). */
+		smp_mb(); /* Ensure test happens before caller kfree(). */
-		smp_mb__before_atomic(); /* ^^^ */
 		return true;
 	}
 	return false;
@@ -384,7 +383,12 @@ retry_ipi:
 			mask_ofl_test |= mask;
 			continue;
 		}
+		if (get_cpu() == cpu) {
+			put_cpu();
+			continue;
+		}
 		ret = smp_call_function_single(cpu, rcu_exp_handler, NULL, 0);
+		put_cpu();
 		if (!ret) {
 			mask_ofl_ipi &= ~mask;
 			continue;
@@ -611,7 +615,7 @@ static void rcu_exp_handler(void *unused)
 		    rcu_dynticks_curr_cpu_in_eqs()) {
 			rcu_report_exp_rdp(rdp);
 		} else {
-			rdp->deferred_qs = true;
+			rdp->exp_deferred_qs = true;
 			set_tsk_need_resched(t);
 			set_preempt_need_resched();
 		}
@@ -633,7 +637,7 @@ static void rcu_exp_handler(void *unused)
 	if (t->rcu_read_lock_nesting > 0) {
 		raw_spin_lock_irqsave_rcu_node(rnp, flags);
 		if (rnp->expmask & rdp->grpmask) {
-			rdp->deferred_qs = true;
+			rdp->exp_deferred_qs = true;
 			t->rcu_read_unlock_special.b.exp_hint = true;
 		}
 		raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
@@ -656,7 +660,7 @@ static void rcu_exp_handler(void *unused)
 	 *
 	 * Otherwise, force a context switch after the CPU enables everything.
 	 */
-	rdp->deferred_qs = true;
+	rdp->exp_deferred_qs = true;
 	if (!(preempt_count() & (PREEMPT_MASK | SOFTIRQ_MASK)) ||
 	    WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs())) {
 		rcu_preempt_deferred_qs(t);
@@ -694,6 +698,16 @@ static int rcu_print_task_exp_stall(struct rcu_node *rnp)
 
 #else /* #ifdef CONFIG_PREEMPT_RCU */
 
+/* Request an expedited quiescent state. */
+static void rcu_exp_need_qs(void)
+{
+	__this_cpu_write(rcu_data.cpu_no_qs.b.exp, true);
+	/* Store .exp before .rcu_urgent_qs. */
+	smp_store_release(this_cpu_ptr(&rcu_data.rcu_urgent_qs), true);
+	set_tsk_need_resched(current);
+	set_preempt_need_resched();
+}
+
 /* Invoked on each online non-idle CPU for expedited quiescent state. */
 static void rcu_exp_handler(void *unused)
 {
@@ -709,25 +723,38 @@ static void rcu_exp_handler(void *unused)
 		rcu_report_exp_rdp(this_cpu_ptr(&rcu_data));
 		return;
 	}
-	__this_cpu_write(rcu_data.cpu_no_qs.b.exp, true);
+	rcu_exp_need_qs();
-	/* Store .exp before .rcu_urgent_qs. */
-	smp_store_release(this_cpu_ptr(&rcu_data.rcu_urgent_qs), true);
-	set_tsk_need_resched(current);
-	set_preempt_need_resched();
 }
 
 /* Send IPI for expedited cleanup if needed at end of CPU-hotplug operation. */
 static void sync_sched_exp_online_cleanup(int cpu)
 {
+	unsigned long flags;
+	int my_cpu;
 	struct rcu_data *rdp;
 	int ret;
 	struct rcu_node *rnp;
 
 	rdp = per_cpu_ptr(&rcu_data, cpu);
 	rnp = rdp->mynode;
-	if (!(READ_ONCE(rnp->expmask) & rdp->grpmask))
+	my_cpu = get_cpu();
+	/* Quiescent state either not needed or already requested, leave. */
+	if (!(READ_ONCE(rnp->expmask) & rdp->grpmask) ||
+	    __this_cpu_read(rcu_data.cpu_no_qs.b.exp)) {
+		put_cpu();
 		return;
+	}
+	/* Quiescent state needed on current CPU, so set it up locally. */
+	if (my_cpu == cpu) {
+		local_irq_save(flags);
+		rcu_exp_need_qs();
+		local_irq_restore(flags);
+		put_cpu();
+		return;
+	}
+	/* Quiescent state needed on some other CPU, send IPI. */
 	ret = smp_call_function_single(cpu, rcu_exp_handler, NULL, 0);
+	put_cpu();
 	WARN_ON_ONCE(ret);
 }
 
@@ -765,7 +792,6 @@ static int rcu_print_task_exp_stall(struct rcu_node *rnp)
  */
 void synchronize_rcu_expedited(void)
 {
-	struct rcu_data *rdp;
 	struct rcu_exp_work rew;
 	struct rcu_node *rnp;
 	unsigned long s;
@@ -802,7 +828,6 @@ void synchronize_rcu_expedited(void)
 	}
 
 	/* Wait for expedited grace period to complete. */
-	rdp = per_cpu_ptr(&rcu_data, raw_smp_processor_id());
 	rnp = rcu_get_root();
 	wait_event(rnp->exp_wq[rcu_seq_ctr(s) & 0x3],
 		   sync_exp_work_done(s));

kernel/rcu/tree_plugin.h

@@ -11,29 +11,7 @@
  *	   Paul E. McKenney <paulmck@linux.ibm.com>
  */
 
-#include <linux/delay.h>
-#include <linux/gfp.h>
-#include <linux/oom.h>
-#include <linux/sched/debug.h>
-#include <linux/smpboot.h>
-#include <linux/sched/isolation.h>
-#include <uapi/linux/sched/types.h>
-#include "../time/tick-internal.h"
-
-#ifdef CONFIG_RCU_BOOST
 #include "../locking/rtmutex_common.h"
-#else /* #ifdef CONFIG_RCU_BOOST */
-
-/*
- * Some architectures do not define rt_mutexes, but if !CONFIG_RCU_BOOST,
- * all uses are in dead code.  Provide a definition to keep the compiler
- * happy, but add WARN_ON_ONCE() to complain if used in the wrong place.
- * This probably needs to be excluded from -rt builds.
- */
-#define rt_mutex_owner(a) ({ WARN_ON_ONCE(1); NULL; })
-#define rt_mutex_futex_unlock(x) WARN_ON_ONCE(1)
-
-#endif /* #else #ifdef CONFIG_RCU_BOOST */
 
 #ifdef CONFIG_RCU_NOCB_CPU
 static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */
@@ -94,6 +72,8 @@ static void __init rcu_bootup_announce_oddness(void)
 		pr_info("\tRCU debug GP init slowdown %d jiffies.\n", gp_init_delay);
 	if (gp_cleanup_delay)
 		pr_info("\tRCU debug GP init slowdown %d jiffies.\n", gp_cleanup_delay);
+	if (!use_softirq)
+		pr_info("\tRCU_SOFTIRQ processing moved to rcuc kthreads.\n");
 	if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG))
 		pr_info("\tRCU debug extended QS entry/exit.\n");
 	rcupdate_announce_bootup_oddness();
@@ -257,10 +237,10 @@ static void rcu_preempt_ctxt_queue(struct rcu_node *rnp, struct rcu_data *rdp)
 	 * no need to check for a subsequent expedited GP.  (Though we are
 	 * still in a quiescent state in any case.)
 	 */
-	if (blkd_state & RCU_EXP_BLKD && rdp->deferred_qs)
+	if (blkd_state & RCU_EXP_BLKD && rdp->exp_deferred_qs)
 		rcu_report_exp_rdp(rdp);
 	else
-		WARN_ON_ONCE(rdp->deferred_qs);
+		WARN_ON_ONCE(rdp->exp_deferred_qs);
 }
 
 /*
@@ -357,7 +337,7 @@ void rcu_note_context_switch(bool preempt)
 	 * means that we continue to block the current grace period.
 	 */
 	rcu_qs();
-	if (rdp->deferred_qs)
+	if (rdp->exp_deferred_qs)
 		rcu_report_exp_rdp(rdp);
 	trace_rcu_utilization(TPS("End context switch"));
 	barrier(); /* Avoid RCU read-side critical sections leaking up. */
@@ -471,14 +451,15 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags)
 	 */
 	special = t->rcu_read_unlock_special;
 	rdp = this_cpu_ptr(&rcu_data);
-	if (!special.s && !rdp->deferred_qs) {
+	if (!special.s && !rdp->exp_deferred_qs) {
 		local_irq_restore(flags);
 		return;
 	}
+	t->rcu_read_unlock_special.b.deferred_qs = false;
 	if (special.b.need_qs) {
 		rcu_qs();
 		t->rcu_read_unlock_special.b.need_qs = false;
-		if (!t->rcu_read_unlock_special.s && !rdp->deferred_qs) {
+		if (!t->rcu_read_unlock_special.s && !rdp->exp_deferred_qs) {
 			local_irq_restore(flags);
 			return;
 		}
@@ -490,7 +471,7 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags)
 	 * tasks are handled when removing the task from the
 	 * blocked-tasks list below.
 	 */
-	if (rdp->deferred_qs) {
+	if (rdp->exp_deferred_qs) {
 		rcu_report_exp_rdp(rdp);
 		if (!t->rcu_read_unlock_special.s) {
 			local_irq_restore(flags);
@@ -579,7 +560,7 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags)
  */
 static bool rcu_preempt_need_deferred_qs(struct task_struct *t)
 {
-	return (__this_cpu_read(rcu_data.deferred_qs) ||
+	return (__this_cpu_read(rcu_data.exp_deferred_qs) ||
 		READ_ONCE(t->rcu_read_unlock_special.s)) &&
 	       t->rcu_read_lock_nesting <= 0;
 }
@@ -606,6 +587,17 @@ static void rcu_preempt_deferred_qs(struct task_struct *t)
 		t->rcu_read_lock_nesting += RCU_NEST_BIAS;
 }
 
+/*
+ * Minimal handler to give the scheduler a chance to re-evaluate.
+ */
+static void rcu_preempt_deferred_qs_handler(struct irq_work *iwp)
+{
+	struct rcu_data *rdp;
+
+	rdp = container_of(iwp, struct rcu_data, defer_qs_iw);
+	rdp->defer_qs_iw_pending = false;
+}
+
 /*
  * Handle special cases during rcu_read_unlock(), such as needing to
  * notify RCU core processing or task having blocked during the RCU
@@ -625,16 +617,41 @@ static void rcu_read_unlock_special(struct task_struct *t)
 	local_irq_save(flags);
 	irqs_were_disabled = irqs_disabled_flags(flags);
 	if (preempt_bh_were_disabled || irqs_were_disabled) {
-		WRITE_ONCE(t->rcu_read_unlock_special.b.exp_hint, false);
+		bool exp;
-		/* Need to defer quiescent state until everything is enabled. */
+		struct rcu_data *rdp = this_cpu_ptr(&rcu_data);
-		if (irqs_were_disabled) {
+		struct rcu_node *rnp = rdp->mynode;
-			/* Enabling irqs does not reschedule, so... */
+
+		t->rcu_read_unlock_special.b.exp_hint = false;
+		exp = (t->rcu_blocked_node && t->rcu_blocked_node->exp_tasks) ||
+		      (rdp->grpmask & rnp->expmask) ||
+		      tick_nohz_full_cpu(rdp->cpu);
+		// Need to defer quiescent state until everything is enabled.
+		if ((exp || in_irq()) && irqs_were_disabled && use_softirq &&
+		    (in_irq() || !t->rcu_read_unlock_special.b.deferred_qs)) {
+			// Using softirq, safe to awaken, and we get
+			// no help from enabling irqs, unlike bh/preempt.
 			raise_softirq_irqoff(RCU_SOFTIRQ);
+		} else if (exp && irqs_were_disabled && !use_softirq &&
+			   !t->rcu_read_unlock_special.b.deferred_qs) {
+			// Safe to awaken and we get no help from enabling
+			// irqs, unlike bh/preempt.
+			invoke_rcu_core();
 		} else {
-			/* Enabling BH or preempt does reschedule, so... */
+			// Enabling BH or preempt does reschedule, so...
+			// Also if no expediting or NO_HZ_FULL, slow is OK.
 			set_tsk_need_resched(current);
 			set_preempt_need_resched();
+			if (IS_ENABLED(CONFIG_IRQ_WORK) &&
+			    !rdp->defer_qs_iw_pending && exp) {
+				// Get scheduler to re-evaluate and call hooks.
+				// If !IRQ_WORK, FQS scan will eventually IPI.
+				init_irq_work(&rdp->defer_qs_iw,
+					      rcu_preempt_deferred_qs_handler);
+				rdp->defer_qs_iw_pending = true;
+				irq_work_queue_on(&rdp->defer_qs_iw, rdp->cpu);
+			}
 		}
+		t->rcu_read_unlock_special.b.deferred_qs = true;
 		local_irq_restore(flags);
 		return;
 	}
@@ -760,7 +777,7 @@ dump_blkd_tasks(struct rcu_node *rnp, int ncheck)
 	i = 0;
 	list_for_each(lhp, &rnp->blkd_tasks) {
 		pr_cont(" %p", lhp);
-		if (++i >= 10)
+		if (++i >= ncheck)
 			break;
 	}
 	pr_cont("\n");
@@ -944,18 +961,21 @@ dump_blkd_tasks(struct rcu_node *rnp, int ncheck)
 
 #endif /* #else #ifdef CONFIG_PREEMPT_RCU */
 
-#ifdef CONFIG_RCU_BOOST
+/*
-
+ * If boosting, set rcuc kthreads to realtime priority.
-static void rcu_wake_cond(struct task_struct *t, int status)
+ */
+static void rcu_cpu_kthread_setup(unsigned int cpu)
 {
-	/*
+#ifdef CONFIG_RCU_BOOST
-	 * If the thread is yielding, only wake it when this
+	struct sched_param sp;
-	 * is invoked from idle
+
-	 */
+	sp.sched_priority = kthread_prio;
-	if (status != RCU_KTHREAD_YIELDING || is_idle_task(current))
+	sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);
-		wake_up_process(t);
+#endif /* #ifdef CONFIG_RCU_BOOST */
 }
 
+#ifdef CONFIG_RCU_BOOST
+
 /*
  * Carry out RCU priority boosting on the task indicated by ->exp_tasks
  * or ->boost_tasks, advancing the pointer to the next task in the
@@ -1090,23 +1110,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
 	}
 }
 
-/*
- * Wake up the per-CPU kthread to invoke RCU callbacks.
- */
-static void invoke_rcu_callbacks_kthread(void)
-{
-	unsigned long flags;
-
-	local_irq_save(flags);
-	__this_cpu_write(rcu_data.rcu_cpu_has_work, 1);
-	if (__this_cpu_read(rcu_data.rcu_cpu_kthread_task) != NULL &&
-	    current != __this_cpu_read(rcu_data.rcu_cpu_kthread_task)) {
-		rcu_wake_cond(__this_cpu_read(rcu_data.rcu_cpu_kthread_task),
-			      __this_cpu_read(rcu_data.rcu_cpu_kthread_status));
-	}
-	local_irq_restore(flags);
-}
-
 /*
  * Is the current CPU running the RCU-callbacks kthread?
  * Caller must have preemption disabled.
@@ -1160,59 +1163,6 @@ static int rcu_spawn_one_boost_kthread(struct rcu_node *rnp)
 	return 0;
 }
 
-static void rcu_cpu_kthread_setup(unsigned int cpu)
-{
-	struct sched_param sp;
-
-	sp.sched_priority = kthread_prio;
-	sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);
-}
-
-static void rcu_cpu_kthread_park(unsigned int cpu)
-{
-	per_cpu(rcu_data.rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU;
-}
-
-static int rcu_cpu_kthread_should_run(unsigned int cpu)
-{
-	return __this_cpu_read(rcu_data.rcu_cpu_has_work);
-}
-
-/*
- * Per-CPU kernel thread that invokes RCU callbacks.  This replaces
- * the RCU softirq used in configurations of RCU that do not support RCU
- * priority boosting.
- */
-static void rcu_cpu_kthread(unsigned int cpu)
-{
-	unsigned int *statusp = this_cpu_ptr(&rcu_data.rcu_cpu_kthread_status);
-	char work, *workp = this_cpu_ptr(&rcu_data.rcu_cpu_has_work);
-	int spincnt;
-
-	for (spincnt = 0; spincnt < 10; spincnt++) {
-		trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait"));
-		local_bh_disable();
-		*statusp = RCU_KTHREAD_RUNNING;
-		local_irq_disable();
-		work = *workp;
-		*workp = 0;
-		local_irq_enable();
-		if (work)
-			rcu_do_batch(this_cpu_ptr(&rcu_data));
-		local_bh_enable();
-		if (*workp == 0) {
-			trace_rcu_utilization(TPS("End CPU kthread@rcu_wait"));
-			*statusp = RCU_KTHREAD_WAITING;
-			return;
-		}
-	}
-	*statusp = RCU_KTHREAD_YIELDING;
-	trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield"));
-	schedule_timeout_interruptible(2);
-	trace_rcu_utilization(TPS("End CPU kthread@rcu_yield"));
-	*statusp = RCU_KTHREAD_WAITING;
-}
-
 /*
  * Set the per-rcu_node kthread's affinity to cover all CPUs that are
  * served by the rcu_node in question.  The CPU hotplug lock is still
@@ -1243,27 +1193,13 @@ static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu)
 	free_cpumask_var(cm);
 }
 
-static struct smp_hotplug_thread rcu_cpu_thread_spec = {
-	.store			= &rcu_data.rcu_cpu_kthread_task,
-	.thread_should_run	= rcu_cpu_kthread_should_run,
-	.thread_fn		= rcu_cpu_kthread,
-	.thread_comm		= "rcuc/%u",
-	.setup			= rcu_cpu_kthread_setup,
-	.park			= rcu_cpu_kthread_park,
-};
-
 /*
  * Spawn boost kthreads -- called as soon as the scheduler is running.
  */
 static void __init rcu_spawn_boost_kthreads(void)
 {
 	struct rcu_node *rnp;
-	int cpu;
 
-	for_each_possible_cpu(cpu)
-		per_cpu(rcu_data.rcu_cpu_has_work, cpu) = 0;
-	if (WARN_ONCE(smpboot_register_percpu_thread(&rcu_cpu_thread_spec), "%s: Could not start rcub kthread, OOM is now expected behavior\n", __func__))
-		return;
 	rcu_for_each_leaf_node(rnp)
 		(void)rcu_spawn_one_boost_kthread(rnp);
 }
@@ -1286,11 +1222,6 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
 	raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
 }
 
-static void invoke_rcu_callbacks_kthread(void)
-{
-	WARN_ON_ONCE(1);
-}
-
 static bool rcu_is_callbacks_kthread(void)
 {
 	return false;

kernel/rcu/tree_stall.h

@@ -630,7 +630,9 @@ static void rcu_check_gp_start_stall(struct rcu_node *rnp, struct rcu_data *rdp,
 	    time_before(j, rcu_state.gp_req_activity + gpssdelay) ||
 	    time_before(j, rcu_state.gp_activity + gpssdelay) ||
 	    atomic_xchg(&warned, 1)) {
-		raw_spin_unlock_rcu_node(rnp_root); /* irqs remain disabled. */
+		if (rnp_root != rnp)
+			/* irqs remain disabled. */
+			raw_spin_unlock_rcu_node(rnp_root);
 		raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
 		return;
 	}

kernel/rcu/update.c

@@ -423,6 +423,19 @@ EXPORT_SYMBOL_GPL(do_trace_rcu_torture_read);
 	do { } while (0)
 #endif
 
+#if IS_ENABLED(CONFIG_RCU_TORTURE_TEST) || IS_MODULE(CONFIG_RCU_TORTURE_TEST)
+/* Get rcutorture access to sched_setaffinity(). */
+long rcutorture_sched_setaffinity(pid_t pid, const struct cpumask *in_mask)
+{
+	int ret;
+
+	ret = sched_setaffinity(pid, in_mask);
+	WARN_ONCE(ret, "%s: sched_setaffinity() returned %d\n", __func__, ret);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(rcutorture_sched_setaffinity);
+#endif
+
 #ifdef CONFIG_RCU_STALL_COMMON
 int rcu_cpu_stall_suppress __read_mostly; /* 1 = suppress stall warnings. */
 EXPORT_SYMBOL_GPL(rcu_cpu_stall_suppress);

kernel/torture.c

@@ -570,6 +570,7 @@ static void torture_shutdown_cleanup(void)
 static struct task_struct *stutter_task;
 static int stutter_pause_test;
 static int stutter;
+static int stutter_gap;
 
 /*
  * Block until the stutter interval ends.  This must be called periodically
@@ -578,10 +579,12 @@ static int stutter;
 bool stutter_wait(const char *title)
 {
 	int spt;
+	bool ret = false;
 
 	cond_resched_tasks_rcu_qs();
 	spt = READ_ONCE(stutter_pause_test);
 	for (; spt; spt = READ_ONCE(stutter_pause_test)) {
+		ret = true;
 		if (spt == 1) {
 			schedule_timeout_interruptible(1);
 		} else if (spt == 2) {
@@ -592,7 +595,7 @@ bool stutter_wait(const char *title)
 		}
 		torture_shutdown_absorb(title);
 	}
-	return !!spt;
+	return ret;
 }
 EXPORT_SYMBOL_GPL(stutter_wait);
 
@@ -602,17 +605,24 @@ EXPORT_SYMBOL_GPL(stutter_wait);
  */
 static int torture_stutter(void *arg)
 {
+	int wtime;
+
 	VERBOSE_TOROUT_STRING("torture_stutter task started");
 	do {
 		if (!torture_must_stop() && stutter > 1) {
-			WRITE_ONCE(stutter_pause_test, 1);
+			wtime = stutter;
-			schedule_timeout_interruptible(stutter - 1);
+			if (stutter > HZ + 1) {
+				WRITE_ONCE(stutter_pause_test, 1);
+				wtime = stutter - HZ - 1;
+				schedule_timeout_interruptible(wtime);
+				wtime = HZ + 1;
+			}
 			WRITE_ONCE(stutter_pause_test, 2);
-			schedule_timeout_interruptible(1);
+			schedule_timeout_interruptible(wtime);
 		}
 		WRITE_ONCE(stutter_pause_test, 0);
 		if (!torture_must_stop())
-			schedule_timeout_interruptible(stutter);
+			schedule_timeout_interruptible(stutter_gap);
 		torture_shutdown_absorb("torture_stutter");
 	} while (!torture_must_stop());
 	torture_kthread_stopping("torture_stutter");
@@ -622,9 +632,10 @@ static int torture_stutter(void *arg)
 /*
  * Initialize and kick off the torture_stutter kthread.
  */
-int torture_stutter_init(const int s)
+int torture_stutter_init(const int s, const int sgap)
 {
 	stutter = s;
+	stutter_gap = sgap;
 	return torture_create_kthread(torture_stutter, NULL, stutter_task);
 }
 EXPORT_SYMBOL_GPL(torture_stutter_init);

tools/include/linux/rcu.h

@@ -19,7 +19,7 @@ static inline bool rcu_is_watching(void)
 	return false;
 }
 
-#define rcu_assign_pointer(p, v) ((p) = (v))
+#define rcu_assign_pointer(p, v)	do { (p) = (v); } while (0)
-#define RCU_INIT_POINTER(p, v) p=(v)
+#define RCU_INIT_POINTER(p, v)	do { (p) = (v); } while (0)
 
 #endif

tools/memory-model/linux-kernel.bell

@@ -24,6 +24,7 @@ instructions RMW[{'once,'acquire,'release}]
 enum Barriers = 'wmb (*smp_wmb*) ||
 		'rmb (*smp_rmb*) ||
 		'mb (*smp_mb*) ||
+		'barrier (*barrier*) ||
 		'rcu-lock (*rcu_read_lock*)  ||
 		'rcu-unlock (*rcu_read_unlock*) ||
 		'sync-rcu (*synchronize_rcu*) ||
@@ -76,3 +77,8 @@ flag ~empty rcu-rscs & (po ; [Sync-srcu] ; po) as invalid-sleep
 
 (* Validate SRCU dynamic match *)
 flag ~empty different-values(srcu-rscs) as srcu-bad-nesting
+
+(* Compute marked and plain memory accesses *)
+let Marked = (~M) | IW | Once | Release | Acquire | domain(rmw) | range(rmw) |
+		LKR | LKW | UL | LF | RL | RU
+let Plain = M \ Marked

tools/memory-model/linux-kernel.cat

@@ -24,8 +24,14 @@ include "lock.cat"
 (* Basic relations *)
 (*******************)
 
+(* Release Acquire *)
+let acq-po = [Acquire] ; po ; [M]
+let po-rel = [M] ; po ; [Release]
+let po-unlock-rf-lock-po = po ; [UL] ; rf ; [LKR] ; po
+
 (* Fences *)
-let rmb = [R \ Noreturn] ; fencerel(Rmb) ; [R \ Noreturn]
+let R4rmb = R \ Noreturn	(* Reads for which rmb works *)
+let rmb = [R4rmb] ; fencerel(Rmb) ; [R4rmb]
 let wmb = [W] ; fencerel(Wmb) ; [W]
 let mb = ([M] ; fencerel(Mb) ; [M]) |
 	([M] ; fencerel(Before-atomic) ; [RMW] ; po? ; [M]) |
@@ -34,13 +40,14 @@ let mb = ([M] ; fencerel(Mb) ; [M]) |
 	([M] ; po ; [UL] ; (co | po) ; [LKW] ;
 		fencerel(After-unlock-lock) ; [M])
 let gp = po ; [Sync-rcu | Sync-srcu] ; po?
-
 let strong-fence = mb | gp
 
-(* Release Acquire *)
+let nonrw-fence = strong-fence | po-rel | acq-po
-let acq-po = [Acquire] ; po ; [M]
+let fence = nonrw-fence | wmb | rmb
-let po-rel = [M] ; po ; [Release]
+let barrier = fencerel(Barrier | Rmb | Wmb | Mb | Sync-rcu | Sync-srcu |
-let po-unlock-rf-lock-po = po ; [UL] ; rf ; [LKR] ; po
+		Before-atomic | After-atomic | Acquire | Release |
+		Rcu-lock | Rcu-unlock | Srcu-lock | Srcu-unlock) |
+	(po ; [Release]) | ([Acquire] ; po)
 
 (**********************************)
 (* Fundamental coherence ordering *)
@@ -61,21 +68,22 @@ empty rmw & (fre ; coe) as atomic
 let dep = addr | data
 let rwdep = (dep | ctrl) ; [W]
 let overwrite = co | fr
-let to-w = rwdep | (overwrite & int)
+let to-w = rwdep | (overwrite & int) | (addr ; [Plain] ; wmb)
-let to-r = addr | (dep ; rfi)
+let to-r = addr | (dep ; [Marked] ; rfi)
-let fence = strong-fence | wmb | po-rel | rmb | acq-po
 let ppo = to-r | to-w | fence | (po-unlock-rf-lock-po & int)
 
 (* Propagation: Ordering from release operations and strong fences. *)
-let A-cumul(r) = rfe? ; r
+let A-cumul(r) = (rfe ; [Marked])? ; r
-let cumul-fence = A-cumul(strong-fence | po-rel) | wmb | po-unlock-rf-lock-po
+let cumul-fence = [Marked] ; (A-cumul(strong-fence | po-rel) | wmb |
-let prop = (overwrite & ext)? ; cumul-fence* ; rfe?
+	po-unlock-rf-lock-po) ; [Marked]
+let prop = [Marked] ; (overwrite & ext)? ; cumul-fence* ;
+	[Marked] ; rfe? ; [Marked]
 
 (*
  * Happens Before: Ordering from the passage of time.
  * No fences needed here for prop because relation confined to one process.
  *)
-let hb = ppo | rfe | ((prop \ id) & int)
+let hb = [Marked] ; (ppo | rfe | ((prop \ id) & int)) ; [Marked]
 acyclic hb as happens-before
 
 (****************************************)
@@ -83,7 +91,7 @@ acyclic hb as happens-before
 (****************************************)
 
 (* Propagation: Each non-rf link needs a strong fence. *)
-let pb = prop ; strong-fence ; hb*
+let pb = prop ; strong-fence ; hb* ; [Marked]
 acyclic pb as propagation
 
 (*******)
@@ -114,24 +122,28 @@ let rcu-link = po? ; hb* ; pb* ; prop ; po
 
 (*
  * Any sequence containing at least as many grace periods as RCU read-side
- * critical sections (joined by rcu-link) acts as a generalized strong fence.
+ * critical sections (joined by rcu-link) induces order like a generalized
+ * inter-CPU strong fence.
  * Likewise for SRCU grace periods and read-side critical sections, provided
  * the synchronize_srcu() and srcu_read_[un]lock() calls refer to the same
  * struct srcu_struct location.
  *)
-let rec rcu-fence = rcu-gp | srcu-gp |
+let rec rcu-order = rcu-gp | srcu-gp |
 	(rcu-gp ; rcu-link ; rcu-rscsi) |
 	((srcu-gp ; rcu-link ; srcu-rscsi) & loc) |
 	(rcu-rscsi ; rcu-link ; rcu-gp) |
 	((srcu-rscsi ; rcu-link ; srcu-gp) & loc) |
-	(rcu-gp ; rcu-link ; rcu-fence ; rcu-link ; rcu-rscsi) |
+	(rcu-gp ; rcu-link ; rcu-order ; rcu-link ; rcu-rscsi) |
-	((srcu-gp ; rcu-link ; rcu-fence ; rcu-link ; srcu-rscsi) & loc) |
+	((srcu-gp ; rcu-link ; rcu-order ; rcu-link ; srcu-rscsi) & loc) |
-	(rcu-rscsi ; rcu-link ; rcu-fence ; rcu-link ; rcu-gp) |
+	(rcu-rscsi ; rcu-link ; rcu-order ; rcu-link ; rcu-gp) |
-	((srcu-rscsi ; rcu-link ; rcu-fence ; rcu-link ; srcu-gp) & loc) |
+	((srcu-rscsi ; rcu-link ; rcu-order ; rcu-link ; srcu-gp) & loc) |
-	(rcu-fence ; rcu-link ; rcu-fence)
+	(rcu-order ; rcu-link ; rcu-order)
+let rcu-fence = po ; rcu-order ; po?
+let fence = fence | rcu-fence
+let strong-fence = strong-fence | rcu-fence
 
 (* rb orders instructions just as pb does *)
-let rb = prop ; po ; rcu-fence ; po? ; hb* ; pb*
+let rb = prop ; rcu-fence ; hb* ; pb* ; [Marked]
 
 irreflexive rb as rcu
 
@@ -143,3 +155,49 @@ irreflexive rb as rcu
  * let xb = hb | pb | rb
  * acyclic xb as executes-before
  *)
+
+(*********************************)
+(* Plain accesses and data races *)
+(*********************************)
+
+(* Warn about plain writes and marked accesses in the same region *)
+let mixed-accesses = ([Plain & W] ; (po-loc \ barrier) ; [Marked]) |
+	([Marked] ; (po-loc \ barrier) ; [Plain & W])
+flag ~empty mixed-accesses as mixed-accesses
+
+(* Executes-before and visibility *)
+let xbstar = (hb | pb | rb)*
+let vis = cumul-fence* ; rfe? ; [Marked] ;
+	((strong-fence ; [Marked] ; xbstar) | (xbstar & int))
+
+(* Boundaries for lifetimes of plain accesses *)
+let w-pre-bounded = [Marked] ; (addr | fence)?
+let r-pre-bounded = [Marked] ; (addr | nonrw-fence |
+	([R4rmb] ; fencerel(Rmb) ; [~Noreturn]))?
+let w-post-bounded = fence? ; [Marked]
+let r-post-bounded = (nonrw-fence | ([~Noreturn] ; fencerel(Rmb) ; [R4rmb]))? ;
+	[Marked]
+
+(* Visibility and executes-before for plain accesses *)
+let ww-vis = fence | (strong-fence ; xbstar ; w-pre-bounded) |
+	(w-post-bounded ; vis ; w-pre-bounded)
+let wr-vis = fence | (strong-fence ; xbstar ; r-pre-bounded) |
+	(w-post-bounded ; vis ; r-pre-bounded)
+let rw-xbstar = fence | (r-post-bounded ; xbstar ; w-pre-bounded)
+
+(* Potential races *)
+let pre-race = ext & ((Plain * M) | ((M \ IW) * Plain))
+
+(* Coherence requirements for plain accesses *)
+let wr-incoh = pre-race & rf & rw-xbstar^-1
+let rw-incoh = pre-race & fr & wr-vis^-1
+let ww-incoh = pre-race & co & ww-vis^-1
+empty (wr-incoh | rw-incoh | ww-incoh) as plain-coherence
+
+(* Actual races *)
+let ww-nonrace = ww-vis & ((Marked * W) | rw-xbstar) & ((W * Marked) | wr-vis)
+let ww-race = (pre-race & co) \ ww-nonrace
+let wr-race = (pre-race & (co? ; rf)) \ wr-vis
+let rw-race = (pre-race & fr) \ rw-xbstar
+
+flag ~empty (ww-race | wr-race | rw-race) as data-race

tools/memory-model/linux-kernel.def

@@ -24,6 +24,7 @@ smp_mb__before_atomic() { __fence{before-atomic}; }
 smp_mb__after_atomic() { __fence{after-atomic}; }
 smp_mb__after_spinlock() { __fence{after-spinlock}; }
 smp_mb__after_unlock_lock() { __fence{after-unlock-lock}; }
+barrier() { __fence{barrier}; }
 
 // Exchange
 xchg(X,V)  __xchg{mb}(X,V)

tools/memory-model/litmus-tests/MP+poonceonces.litmus

@@ -1,7 +1,7 @@
 C MP+poonceonces
 
 (*
- * Result: Maybe
+ * Result: Sometimes
  *
  * Can the counter-intuitive message-passing outcome be prevented with
  * no ordering at all?

tools/memory-model/litmus-tests/README

@@ -244,7 +244,7 @@ produce the name:
 Adding the ".litmus" suffix: SB+rfionceonce-poonceonces.litmus
 
 The descriptors that describe connections between consecutive accesses
-within the cycle through a given litmus test can be provided by the herd
+within the cycle through a given litmus test can be provided by the herd7
 tool (Rfi, Po, Fre, and so on) or by the linux-kernel.bell file (Once,
 Release, Acquire, and so on).
 

tools/memory-model/lock.cat

@@ -11,7 +11,7 @@
 include "cross.cat"
 
 (*
- * The lock-related events generated by herd are as follows:
+ * The lock-related events generated by herd7 are as follows:
  *
  * LKR		Lock-Read: the read part of a spin_lock() or successful
  *			spin_trylock() read-modify-write event pair

tools/memory-model/scripts/README

@@ -22,7 +22,7 @@ checklitmushist.sh
 
 	Run all litmus tests having .litmus.out files from previous
 	initlitmushist.sh or newlitmushist.sh runs, comparing the
-	herd output to that of the original runs.
+	herd7 output to that of the original runs.
 
 checklitmus.sh
 
@@ -43,7 +43,7 @@ initlitmushist.sh
 
 judgelitmus.sh
 
-	Given a .litmus file and its .litmus.out herd output, check the
+	Given a .litmus file and its .litmus.out herd7 output, check the
 	.litmus.out file against the .litmus file's "Result:" comment to
 	judge whether the test ran correctly.  Not normally run manually,
 	provided instead for use by other scripts.

tools/memory-model/scripts/checkalllitmus.sh

@@ -1,7 +1,7 @@
 #!/bin/sh
 # SPDX-License-Identifier: GPL-2.0+
 #
-# Run herd tests on all .litmus files in the litmus-tests directory
+# Run herd7 tests on all .litmus files in the litmus-tests directory
 # and check each file's result against a "Result:" comment within that
 # litmus test.  If the verification result does not match that specified
 # in the litmus test, this script prints an error message prefixed with

tools/memory-model/scripts/checklitmus.sh

@@ -1,7 +1,7 @@
 #!/bin/sh
 # SPDX-License-Identifier: GPL-2.0+
 #
-# Run a herd test and invokes judgelitmus.sh to check the result against
+# Run a herd7 test and invokes judgelitmus.sh to check the result against
 # a "Result:" comment within the litmus test.  It also outputs verification
 # results to a file whose name is that of the specified litmus test, but
 # with ".out" appended.

tools/memory-model/scripts/parseargs.sh

@@ -91,7 +91,7 @@ do
 		shift
 		;;
 	--herdopts|--herdopt)
-		checkarg --destdir "(herd options)" "$#" "$2" '.*' '^--'
+		checkarg --destdir "(herd7 options)" "$#" "$2" '.*' '^--'
 		LKMM_HERD_OPTIONS="$2"
 		shift
 		;;

tools/memory-model/scripts/runlitmushist.sh

@@ -79,7 +79,7 @@ then
 	echo ' ---' Summary: 1>&2
 	grep '!!!' $T/*.sh.out 1>&2
 	nfail="`grep '!!!' $T/*.sh.out | wc -l`"
-	echo 'Number of failed herd runs (e.g., timeout): ' $nfail 1>&2
+	echo 'Number of failed herd7 runs (e.g., timeout): ' $nfail 1>&2
 	exit 1
 else
 	echo All runs completed successfully. 1>&2

tools/testing/radix-tree/linux/rcupdate.h

@@ -7,6 +7,6 @@
 #define rcu_dereference_raw(p) rcu_dereference(p)
 #define rcu_dereference_protected(p, cond) rcu_dereference(p)
 #define rcu_dereference_check(p, cond) rcu_dereference(p)
-#define RCU_INIT_POINTER(p, v)	(p) = (v)
+#define RCU_INIT_POINTER(p, v)	do { (p) = (v); } while (0)
 
 #endif

tools/testing/selftests/rcutorture/Makefile

@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: GPL-2.0+
+all:
+	( cd ../../../..; tools/testing/selftests/rcutorture/bin/kvm.sh --duration 10 --configs TREE01 )

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

@@ -1,7 +1,7 @@
 #!/bin/bash
 # SPDX-License-Identifier: GPL-2.0+
 #
-# Usage: configinit.sh config-spec-file build-output-dir results-dir
+# Usage: configinit.sh config-spec-file results-dir
 #
 # Create a .config file from the spec file.  Run from the kernel source tree.
 # Exits with 0 if all went well, with 1 if all went well but the config
@@ -11,10 +11,6 @@
 # desired settings, for example, "CONFIG_NO_HZ=y".  For best results,
 # this should be a full pathname.
 #
-# The second argument is a optional path to a build output directory,
-# for example, "O=/tmp/foo".  If this argument is omitted, the .config
-# file will be generated directly in the current directory.
-#
 # Copyright (C) IBM Corporation, 2013
 #
 # Authors: Paul E. McKenney <paulmck@linux.ibm.com>
@@ -26,34 +22,23 @@ mkdir $T
 # Capture config spec file.
 
 c=$1
-buildloc=$2
+resdir=$2
-resdir=$3
-builddir=
-if echo $buildloc | grep -q '^O='
-then
-	builddir=`echo $buildloc | sed -e 's/^O=//'`
-	if test ! -d $builddir
-	then
-		mkdir $builddir
-	fi
-else
-	echo Bad build directory: \"$buildloc\"
-	exit 2
-fi
 
 sed -e 's/^\(CONFIG[0-9A-Z_]*\)=.*$/grep -v "^# \1" |/' < $c > $T/u.sh
 sed -e 's/^\(CONFIG[0-9A-Z_]*=\).*$/grep -v \1 |/' < $c >> $T/u.sh
 grep '^grep' < $T/u.sh > $T/upd.sh
 echo "cat - $c" >> $T/upd.sh
-make mrproper
+if test -z "$TORTURE_TRUST_MAKE"
-make $buildloc distclean > $resdir/Make.distclean 2>&1
+then
-make $buildloc $TORTURE_DEFCONFIG > $resdir/Make.defconfig.out 2>&1
+	make clean > $resdir/Make.clean 2>&1
-mv $builddir/.config $builddir/.config.sav
+fi
-sh $T/upd.sh < $builddir/.config.sav > $builddir/.config
+make $TORTURE_DEFCONFIG > $resdir/Make.defconfig.out 2>&1
-cp $builddir/.config $builddir/.config.new
+mv .config .config.sav
-yes '' | make $buildloc oldconfig > $resdir/Make.oldconfig.out 2> $resdir/Make.oldconfig.err
+sh $T/upd.sh < .config.sav > .config
+cp .config .config.new
+yes '' | make oldconfig > $resdir/Make.oldconfig.out 2> $resdir/Make.oldconfig.err
 
 # verify new config matches specification.
-configcheck.sh $builddir/.config $c
+configcheck.sh .config $c
 
 exit 0

tools/testing/selftests/rcutorture/bin/cpus2use.sh

@@ -9,6 +9,11 @@
 #
 # Authors: Paul E. McKenney <paulmck@linux.ibm.com>
 
+if test -n "$TORTURE_ALLOTED_CPUS"
+then
+	echo $TORTURE_ALLOTED_CPUS
+	exit 0
+fi
 ncpus=`grep '^processor' /proc/cpuinfo | wc -l`
 idlecpus=`mpstat | tail -1 | \
 	awk -v ncpus=$ncpus '{ print ncpus * ($7 + $NF) / 100 }'`

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

@@ -172,7 +172,7 @@ identify_qemu_append () {
 	local console=ttyS0
 	case "$1" in
 	qemu-system-x86_64|qemu-system-i386)
-		echo noapic selinux=0 initcall_debug debug
+		echo selinux=0 initcall_debug debug
 		;;
 	qemu-system-aarch64)
 		console=ttyAMA0
@@ -191,8 +191,19 @@ identify_qemu_append () {
 # Output arguments for qemu arguments based on the TORTURE_QEMU_MAC
 # and TORTURE_QEMU_INTERACTIVE environment variables.
 identify_qemu_args () {
+	local KVM_CPU=""
+	case "$1" in
+	qemu-system-x86_64)
+		KVM_CPU=kvm64
+		;;
+	qemu-system-i386)
+		KVM_CPU=kvm32
+		;;
+	esac
 	case "$1" in
 	qemu-system-x86_64|qemu-system-i386)
+		echo -machine q35,accel=kvm
+		echo -cpu ${KVM_CPU}
 		;;
 	qemu-system-aarch64)
 		echo -machine virt,gic-version=host -cpu host

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

@@ -34,10 +34,15 @@ do
 		exit 0;
 	fi
 
-	# Set affinity to randomly selected CPU
+	# Set affinity to randomly selected online CPU
-	cpus=`ls /sys/devices/system/cpu/*/online |
+	cpus=`grep 1 /sys/devices/system/cpu/*/online |
-		sed -e 's,/[^/]*$,,' -e 's/^[^0-9]*//' |
+		sed -e 's,/[^/]*$,,' -e 's/^[^0-9]*//'`
-		grep -v '^0*$'`
+
+	# Do not leave out poor old cpu0 which may not be hot-pluggable
+	if [ ! -f "/sys/devices/system/cpu/cpu0/online" ]; then
+		cpus="0 $cpus"
+	fi
+
 	cpumask=`awk -v cpus="$cpus" -v me=$me -v n=$n 'BEGIN {
 		srand(n + me + systime());
 		ncpus = split(cpus, ca);

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

@@ -3,7 +3,7 @@
 #
 # Build a kvm-ready Linux kernel from the tree in the current directory.
 #
-# Usage: kvm-build.sh config-template build-dir resdir
+# Usage: kvm-build.sh config-template resdir
 #
 # Copyright (C) IBM Corporation, 2011
 #
@@ -15,8 +15,7 @@ then
 	echo "kvm-build.sh :$config_template: Not a readable file"
 	exit 1
 fi
-builddir=${2}
+resdir=${2}
-resdir=${3}
 
 T=${TMPDIR-/tmp}/test-linux.sh.$$
 trap 'rm -rf $T' 0
@@ -29,14 +28,14 @@ CONFIG_VIRTIO_PCI=y
 CONFIG_VIRTIO_CONSOLE=y
 ___EOF___
 
-configinit.sh $T/config O=$builddir $resdir
+configinit.sh $T/config $resdir
 retval=$?
 if test $retval -gt 1
 then
 	exit 2
 fi
 ncpus=`cpus2use.sh`
-make O=$builddir -j$ncpus $TORTURE_KMAKE_ARG > $resdir/Make.out 2>&1
+make -j$ncpus $TORTURE_KMAKE_ARG > $resdir/Make.out 2>&1
 retval=$?
 if test $retval -ne 0 || grep "rcu[^/]*": < $resdir/Make.out | egrep -q "Stop|Error|error:|warning:" || egrep -q "Stop|Error|error:" < $resdir/Make.out
 then

tools/testing/selftests/rcutorture/bin/kvm-find-errors.sh

@@ -11,6 +11,7 @@
 #
 # The "directory" above should end with the date/time directory, for example,
 # "tools/testing/selftests/rcutorture/res/2018.02.25-14:27:27".
+# Returns error status reflecting the success (or not) of the specified run.
 #
 # Copyright (C) IBM Corporation, 2018
 #
@@ -56,6 +57,8 @@ done
 if test -n "$files"
 then
 	$editor $files
+	exit 1
 else
 	echo No errors in console logs.
+	exit 0
 fi

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

@@ -7,6 +7,8 @@
 #
 # Usage: kvm-recheck.sh resdir ...
 #
+# Returns status reflecting the success or not of the last run specified.
+#
 # Copyright (C) IBM Corporation, 2011
 #
 # Authors: Paul E. McKenney <paulmck@linux.ibm.com>
@@ -28,8 +30,16 @@ do
 		TORTURE_SUITE="`cat $i/../TORTURE_SUITE`"
 		rm -f $i/console.log.*.diags
 		kvm-recheck-${TORTURE_SUITE}.sh $i
-		if test -f "$i/console.log"
+		if test -f "$i/qemu-retval" && test "`cat $i/qemu-retval`" -ne 0 && test "`cat $i/qemu-retval`" -ne 137
 		then
+			echo QEMU error, output:
+			cat $i/qemu-output
+		elif test -f "$i/console.log"
+		then
+			if test -f "$i/qemu-retval" && test "`cat $i/qemu-retval`" -eq 137
+			then
+				echo QEMU killed
+			fi
 			configcheck.sh $i/.config $i/ConfigFragment
 			if test -r $i/Make.oldconfig.err
 			then
@@ -58,3 +68,4 @@ do
 		fi
 	done
 done
+EDITOR=echo kvm-find-errors.sh "${@: -1}" > /dev/null 2>&1

tools/testing/selftests/rcutorture/bin/kvm-test-1-run.sh

@@ -36,11 +36,6 @@ config_template=${1}
 config_dir=`echo $config_template | sed -e 's,/[^/]*$,,'`
 title=`echo $config_template | sed -e 's/^.*\///'`
 builddir=${2}
-if test -z "$builddir" -o ! -d "$builddir" -o ! -w "$builddir"
-then
-	echo "kvm-test-1-run.sh :$builddir: Not a writable directory, cannot build into it"
-	exit 1
-fi
 resdir=${3}
 if test -z "$resdir" -o ! -d "$resdir" -o ! -w "$resdir"
 then
@@ -85,18 +80,18 @@ then
 	ln -s $base_resdir/.config $resdir  # for kvm-recheck.sh
 	# Arch-independent indicator
 	touch $resdir/builtkernel
-elif kvm-build.sh $T/Kc2 $builddir $resdir
+elif kvm-build.sh $T/Kc2 $resdir
 then
 	# Had to build a kernel for this test.
-	QEMU="`identify_qemu $builddir/vmlinux`"
+	QEMU="`identify_qemu vmlinux`"
 	BOOT_IMAGE="`identify_boot_image $QEMU`"
-	cp $builddir/vmlinux $resdir
+	cp vmlinux $resdir
-	cp $builddir/.config $resdir
+	cp .config $resdir
-	cp $builddir/Module.symvers $resdir > /dev/null || :
+	cp Module.symvers $resdir > /dev/null || :
-	cp $builddir/System.map $resdir > /dev/null || :
+	cp System.map $resdir > /dev/null || :
 	if test -n "$BOOT_IMAGE"
 	then
-		cp $builddir/$BOOT_IMAGE $resdir
+		cp $BOOT_IMAGE $resdir
 		KERNEL=$resdir/${BOOT_IMAGE##*/}
 		# Arch-independent indicator
 		touch $resdir/builtkernel
@@ -107,7 +102,7 @@ then
 	parse-build.sh $resdir/Make.out $title
 else
 	# Build failed.
-	cp $builddir/.config $resdir || :
+	cp .config $resdir || :
 	echo Build failed, not running KVM, see $resdir.
 	if test -f $builddir.wait
 	then
@@ -165,7 +160,7 @@ then
 fi
 echo "NOTE: $QEMU either did not run or was interactive" > $resdir/console.log
 echo $QEMU $qemu_args -m $TORTURE_QEMU_MEM -kernel $KERNEL -append \"$qemu_append $boot_args\" > $resdir/qemu-cmd
-( $QEMU $qemu_args -m $TORTURE_QEMU_MEM -kernel $KERNEL -append "$qemu_append $boot_args"& echo $! > $resdir/qemu_pid; wait `cat  $resdir/qemu_pid`; echo $? > $resdir/qemu-retval ) &
+( $QEMU $qemu_args -m $TORTURE_QEMU_MEM -kernel $KERNEL -append "$qemu_append $boot_args" > $resdir/qemu-output 2>&1 & echo $! > $resdir/qemu_pid; wait `cat  $resdir/qemu_pid`; echo $? > $resdir/qemu-retval ) &
 commandcompleted=0
 sleep 10 # Give qemu's pid a chance to reach the file
 if test -s "$resdir/qemu_pid"

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

@@ -24,6 +24,7 @@ dur=$((30*60))
 dryrun=""
 KVM="`pwd`/tools/testing/selftests/rcutorture"; export KVM
 PATH=${KVM}/bin:$PATH; export PATH
+TORTURE_ALLOTED_CPUS=""
 TORTURE_DEFCONFIG=defconfig
 TORTURE_BOOT_IMAGE=""
 TORTURE_INITRD="$KVM/initrd"; export TORTURE_INITRD
@@ -32,6 +33,7 @@ TORTURE_KMAKE_ARG=""
 TORTURE_QEMU_MEM=512
 TORTURE_SHUTDOWN_GRACE=180
 TORTURE_SUITE=rcu
+TORTURE_TRUST_MAKE=""
 resdir=""
 configs=""
 cpus=0
@@ -62,6 +64,7 @@ usage () {
 	echo "       --qemu-cmd qemu-system-..."
 	echo "       --results absolute-pathname"
 	echo "       --torture rcu"
+	echo "       --trust-make"
 	exit 1
 }
 
@@ -89,6 +92,7 @@ do
 	--cpus)
 		checkarg --cpus "(number)" "$#" "$2" '^[0-9]*$' '^--'
 		cpus=$2
+		TORTURE_ALLOTED_CPUS="$2"
 		shift
 		;;
 	--datestamp)
@@ -173,6 +177,9 @@ do
 			jitter=0
 		fi
 		;;
+	--trust-make)
+		TORTURE_TRUST_MAKE="y"
+		;;
 	*)
 		echo Unknown argument $1
 		usage
@@ -285,6 +292,7 @@ cat << ___EOF___ > $T/script
 CONFIGFRAG="$CONFIGFRAG"; export CONFIGFRAG
 KVM="$KVM"; export KVM
 PATH="$PATH"; export PATH
+TORTURE_ALLOTED_CPUS="$TORTURE_ALLOTED_CPUS"; export TORTURE_ALLOTED_CPUS
 TORTURE_BOOT_IMAGE="$TORTURE_BOOT_IMAGE"; export TORTURE_BOOT_IMAGE
 TORTURE_BUILDONLY="$TORTURE_BUILDONLY"; export TORTURE_BUILDONLY
 TORTURE_DEFCONFIG="$TORTURE_DEFCONFIG"; export TORTURE_DEFCONFIG
@@ -297,6 +305,7 @@ TORTURE_QEMU_MAC="$TORTURE_QEMU_MAC"; export TORTURE_QEMU_MAC
 TORTURE_QEMU_MEM="$TORTURE_QEMU_MEM"; export TORTURE_QEMU_MEM
 TORTURE_SHUTDOWN_GRACE="$TORTURE_SHUTDOWN_GRACE"; export TORTURE_SHUTDOWN_GRACE
 TORTURE_SUITE="$TORTURE_SUITE"; export TORTURE_SUITE
+TORTURE_TRUST_MAKE="$TORTURE_TRUST_MAKE"; export TORTURE_TRUST_MAKE
 if ! test -e $resdir
 then
 	mkdir -p "$resdir" || :
@@ -342,7 +351,7 @@ function dump(first, pastlast, batchnum)
 	print "needqemurun="
 	jn=1
 	for (j = first; j < pastlast; j++) {
-		builddir=KVM "/b1"
+		builddir=KVM "/b" j - first + 1
 		cpusr[jn] = cpus[j];
 		if (cfrep[cf[j]] == "") {
 			cfr[jn] = cf[j];
@@ -358,7 +367,6 @@ function dump(first, pastlast, batchnum)
 		print "echo ", cfr[jn], cpusr[jn] ovf ": Starting build. `date` | tee -a " rd "log";
 		print "rm -f " builddir ".*";
 		print "touch " builddir ".wait";
-		print "mkdir " builddir " > /dev/null 2>&1 || :";
 		print "mkdir " rd cfr[jn] " || :";
 		print "kvm-test-1-run.sh " CONFIGDIR cf[j], builddir, rd cfr[jn], dur " \"" TORTURE_QEMU_ARG "\" \"" TORTURE_BOOTARGS "\" > " rd cfr[jn]  "/kvm-test-1-run.sh.out 2>&1 &"
 		print "echo ", cfr[jn], cpusr[jn] ovf ": Waiting for build to complete. `date` | tee -a " rd "log";
@@ -464,3 +472,5 @@ else
 fi
 
 # Tracing: trace_event=rcu:rcu_grace_period,rcu:rcu_future_grace_period,rcu:rcu_grace_period_init,rcu:rcu_nocb_wake,rcu:rcu_preempt_task,rcu:rcu_unlock_preempted_task,rcu:rcu_quiescent_state_report,rcu:rcu_fqs,rcu:rcu_callback,rcu:rcu_kfree_callback,rcu:rcu_batch_start,rcu:rcu_invoke_callback,rcu:rcu_invoke_kfree_callback,rcu:rcu_batch_end,rcu:rcu_torture_read,rcu:rcu_barrier
+# Function-graph tracing: ftrace=function_graph ftrace_graph_filter=sched_setaffinity,migration_cpu_stop
+# Also --kconfig "CONFIG_FUNCTION_TRACER=y CONFIG_FUNCTION_GRAPH_TRACER=y"

tools/testing/selftests/rcutorture/bin/parse-build.sh

@@ -21,7 +21,7 @@ mkdir $T
 
 . functions.sh
 
-if grep -q CC < $F
+if grep -q CC < $F || test -n "$TORTURE_TRUST_MAKE"
 then
 	:
 else

tools/testing/selftests/rcutorture/bin/parse-console.sh

@@ -106,6 +106,7 @@ fi | tee -a $file.diags
 
 egrep 'Badness|WARNING:|Warn|BUG|===========|Call Trace:|Oops:|detected stalls on CPUs/tasks:|self-detected stall on CPU|Stall ended before state dump start|\?\?\? Writer stall state|rcu_.*kthread starved for' < $file |
 grep -v 'ODEBUG: ' |
+grep -v 'This means that this is a DEBUG kernel and it is' |
 grep -v 'Warning: unable to open an initial console' > $T.diags
 if test -s $T.diags
 then

tools/testing/selftests/rcutorture/configs/rcu/CFcommon

@@ -1,2 +1,5 @@
 CONFIG_RCU_TORTURE_TEST=y
 CONFIG_PRINTK_TIME=y
+CONFIG_HYPERVISOR_GUEST=y
+CONFIG_PARAVIRT=y
+CONFIG_KVM_GUEST=y

tools/testing/selftests/rcutorture/configs/rcu/TREE01.boot

@@ -3,3 +3,4 @@ rcutree.gp_preinit_delay=3
 rcutree.gp_init_delay=3
 rcutree.gp_cleanup_delay=3
 rcu_nocbs=0
+rcutorture.fwd_progress=0

tools/testing/selftests/rcutorture/configs/rcu/TRIVIAL

@@ -0,0 +1,14 @@
+CONFIG_SMP=y
+CONFIG_NR_CPUS=8
+CONFIG_PREEMPT_NONE=y
+CONFIG_PREEMPT_VOLUNTARY=n
+CONFIG_PREEMPT=n
+CONFIG_HZ_PERIODIC=n
+CONFIG_NO_HZ_IDLE=y
+CONFIG_NO_HZ_FULL=n
+CONFIG_HOTPLUG_CPU=n
+CONFIG_SUSPEND=n
+CONFIG_HIBERNATION=n
+CONFIG_DEBUG_LOCK_ALLOC=n
+CONFIG_DEBUG_OBJECTS_RCU_HEAD=n
+CONFIG_RCU_EXPERT=y

tools/testing/selftests/rcutorture/configs/rcu/TRIVIAL.boot

@@ -0,0 +1,3 @@
+rcutorture.torture_type=trivial
+rcutorture.onoff_interval=0
+rcutorture.shuffle_interval=0