Merge branches 'barrier.2012.05.09a', 'fixes.2012.04.26a', 'inline.2012.05.02b' and 'srcu.2012.05.07b' into HEAD
barrier: Reduce the amount of disturbance by rcu_barrier() to the rest of
the system. This branch also includes improvements to
RCU_FAST_NO_HZ, which are included here due to conflicts.
fixes: Miscellaneous fixes.
inline: Remaining changes from an abortive attempt to inline
preemptible RCU's __rcu_read_lock(). These are (1) making
exit_rcu() avoid unnecessary work and (2) avoiding having
preemptible RCU record a blocked thread when the scheduler
declines to do a context switch.
srcu: Lai Jiangshan's algorithmic implementation of SRCU, including
call_srcu().
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index c1601e5..ab84a01 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -2330,18 +2330,100 @@
ramdisk_size= [RAM] Sizes of RAM disks in kilobytes
See Documentation/blockdev/ramdisk.txt.
- rcupdate.blimit= [KNL,BOOT]
+ rcutree.blimit= [KNL,BOOT]
Set maximum number of finished RCU callbacks to process
in one batch.
- rcupdate.qhimark= [KNL,BOOT]
+ rcutree.qhimark= [KNL,BOOT]
Set threshold of queued
RCU callbacks over which batch limiting is disabled.
- rcupdate.qlowmark= [KNL,BOOT]
+ rcutree.qlowmark= [KNL,BOOT]
Set threshold of queued RCU callbacks below which
batch limiting is re-enabled.
+ rcutree.rcu_cpu_stall_suppress= [KNL,BOOT]
+ Suppress RCU CPU stall warning messages.
+
+ rcutree.rcu_cpu_stall_timeout= [KNL,BOOT]
+ Set timeout for RCU CPU stall warning messages.
+
+ rcutorture.fqs_duration= [KNL,BOOT]
+ Set duration of force_quiescent_state bursts.
+
+ rcutorture.fqs_holdoff= [KNL,BOOT]
+ Set holdoff time within force_quiescent_state bursts.
+
+ rcutorture.fqs_stutter= [KNL,BOOT]
+ Set wait time between force_quiescent_state bursts.
+
+ rcutorture.irqreader= [KNL,BOOT]
+ Test RCU readers from irq handlers.
+
+ rcutorture.n_barrier_cbs= [KNL,BOOT]
+ Set callbacks/threads for rcu_barrier() testing.
+
+ rcutorture.nfakewriters= [KNL,BOOT]
+ Set number of concurrent RCU writers. These just
+ stress RCU, they don't participate in the actual
+ test, hence the "fake".
+
+ rcutorture.nreaders= [KNL,BOOT]
+ Set number of RCU readers.
+
+ rcutorture.onoff_holdoff= [KNL,BOOT]
+ Set time (s) after boot for CPU-hotplug testing.
+
+ rcutorture.onoff_interval= [KNL,BOOT]
+ Set time (s) between CPU-hotplug operations, or
+ zero to disable CPU-hotplug testing.
+
+ rcutorture.shuffle_interval= [KNL,BOOT]
+ Set task-shuffle interval (s). Shuffling tasks
+ allows some CPUs to go into dyntick-idle mode
+ during the rcutorture test.
+
+ rcutorture.shutdown_secs= [KNL,BOOT]
+ Set time (s) after boot system shutdown. This
+ is useful for hands-off automated testing.
+
+ rcutorture.stall_cpu= [KNL,BOOT]
+ Duration of CPU stall (s) to test RCU CPU stall
+ warnings, zero to disable.
+
+ rcutorture.stall_cpu_holdoff= [KNL,BOOT]
+ Time to wait (s) after boot before inducing stall.
+
+ rcutorture.stat_interval= [KNL,BOOT]
+ Time (s) between statistics printk()s.
+
+ rcutorture.stutter= [KNL,BOOT]
+ Time (s) to stutter testing, for example, specifying
+ five seconds causes the test to run for five seconds,
+ wait for five seconds, and so on. This tests RCU's
+ ability to transition abruptly to and from idle.
+
+ rcutorture.test_boost= [KNL,BOOT]
+ Test RCU priority boosting? 0=no, 1=maybe, 2=yes.
+ "Maybe" means test if the RCU implementation
+ under test support RCU priority boosting.
+
+ rcutorture.test_boost_duration= [KNL,BOOT]
+ Duration (s) of each individual boost test.
+
+ rcutorture.test_boost_interval= [KNL,BOOT]
+ Interval (s) between each boost test.
+
+ rcutorture.test_no_idle_hz= [KNL,BOOT]
+ Test RCU's dyntick-idle handling. See also the
+ rcutorture.shuffle_interval parameter.
+
+ rcutorture.torture_type= [KNL,BOOT]
+ Specify the RCU implementation to test.
+
+ rcutorture.verbose= [KNL,BOOT]
+ Enable additional printk() statements.
+
rdinit= [KNL]
Format: <full_path>
Run specified binary instead of /init from the ramdisk,
diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c
index 43b39d6..88e466b 100644
--- a/arch/um/drivers/mconsole_kern.c
+++ b/arch/um/drivers/mconsole_kern.c
@@ -705,6 +705,7 @@
struct task_struct *from = current, *to = arg;
to->thread.saved_task = from;
+ rcu_switch_from(from);
switch_to(from, to, from);
}
diff --git a/include/linux/rculist.h b/include/linux/rculist.h
index d079290..e0f0fab 100644
--- a/include/linux/rculist.h
+++ b/include/linux/rculist.h
@@ -30,6 +30,7 @@
* This is only for internal list manipulation where we know
* the prev/next entries already!
*/
+#ifndef CONFIG_DEBUG_LIST
static inline void __list_add_rcu(struct list_head *new,
struct list_head *prev, struct list_head *next)
{
@@ -38,6 +39,10 @@
rcu_assign_pointer(list_next_rcu(prev), new);
next->prev = new;
}
+#else
+extern void __list_add_rcu(struct list_head *new,
+ struct list_head *prev, struct list_head *next);
+#endif
/**
* list_add_rcu - add a new entry to rcu-protected list
@@ -108,7 +113,7 @@
*/
static inline void list_del_rcu(struct list_head *entry)
{
- __list_del(entry->prev, entry->next);
+ __list_del_entry(entry);
entry->prev = LIST_POISON2;
}
@@ -228,18 +233,43 @@
})
/**
- * list_first_entry_rcu - get the first element from a list
+ * Where are list_empty_rcu() and list_first_entry_rcu()?
+ *
+ * Implementing those functions following their counterparts list_empty() and
+ * list_first_entry() is not advisable because they lead to subtle race
+ * conditions as the following snippet shows:
+ *
+ * if (!list_empty_rcu(mylist)) {
+ * struct foo *bar = list_first_entry_rcu(mylist, struct foo, list_member);
+ * do_something(bar);
+ * }
+ *
+ * The list may not be empty when list_empty_rcu checks it, but it may be when
+ * list_first_entry_rcu rereads the ->next pointer.
+ *
+ * Rereading the ->next pointer is not a problem for list_empty() and
+ * list_first_entry() because they would be protected by a lock that blocks
+ * writers.
+ *
+ * See list_first_or_null_rcu for an alternative.
+ */
+
+/**
+ * list_first_or_null_rcu - get the first element from a list
* @ptr: the list head to take the element from.
* @type: the type of the struct this is embedded in.
* @member: the name of the list_struct within the struct.
*
- * Note, that list is expected to be not empty.
+ * Note that if the list is empty, it returns NULL.
*
* This primitive may safely run concurrently with the _rcu list-mutation
* primitives such as list_add_rcu() as long as it's guarded by rcu_read_lock().
*/
-#define list_first_entry_rcu(ptr, type, member) \
- list_entry_rcu((ptr)->next, type, member)
+#define list_first_or_null_rcu(ptr, type, member) \
+ ({struct list_head *__ptr = (ptr); \
+ struct list_head __rcu *__next = list_next_rcu(__ptr); \
+ likely(__ptr != __next) ? container_of(__next, type, member) : NULL; \
+ })
/**
* list_for_each_entry_rcu - iterate over rcu list of given type
diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
index 20fb776..26d1a47 100644
--- a/include/linux/rcupdate.h
+++ b/include/linux/rcupdate.h
@@ -184,12 +184,14 @@
/* Internal to kernel */
extern void rcu_sched_qs(int cpu);
extern void rcu_bh_qs(int cpu);
+extern void rcu_preempt_note_context_switch(void);
extern void rcu_check_callbacks(int cpu, int user);
struct notifier_block;
extern void rcu_idle_enter(void);
extern void rcu_idle_exit(void);
extern void rcu_irq_enter(void);
extern void rcu_irq_exit(void);
+extern void exit_rcu(void);
/**
* RCU_NONIDLE - Indicate idle-loop code that needs RCU readers
@@ -922,6 +924,21 @@
kfree_call_rcu(head, (rcu_callback)offset);
}
+/*
+ * Does the specified offset indicate that the corresponding rcu_head
+ * structure can be handled by kfree_rcu()?
+ */
+#define __is_kfree_rcu_offset(offset) ((offset) < 4096)
+
+/*
+ * Helper macro for kfree_rcu() to prevent argument-expansion eyestrain.
+ */
+#define __kfree_rcu(head, offset) \
+ do { \
+ BUILD_BUG_ON(!__is_kfree_rcu_offset(offset)); \
+ call_rcu(head, (void (*)(struct rcu_head *))(unsigned long)(offset)); \
+ } while (0)
+
/**
* kfree_rcu() - kfree an object after a grace period.
* @ptr: pointer to kfree
@@ -944,6 +961,9 @@
*
* Note that the allowable offset might decrease in the future, for example,
* to allow something like kmem_cache_free_rcu().
+ *
+ * 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) \
__kfree_rcu(&((ptr)->rcu_head), offsetof(typeof(*(ptr)), rcu_head))
diff --git a/include/linux/rcutiny.h b/include/linux/rcutiny.h
index e93df77..adb5e5a 100644
--- a/include/linux/rcutiny.h
+++ b/include/linux/rcutiny.h
@@ -87,14 +87,6 @@
#ifdef CONFIG_TINY_RCU
-static inline void rcu_preempt_note_context_switch(void)
-{
-}
-
-static inline void exit_rcu(void)
-{
-}
-
static inline int rcu_needs_cpu(int cpu)
{
return 0;
@@ -102,8 +94,6 @@
#else /* #ifdef CONFIG_TINY_RCU */
-void rcu_preempt_note_context_switch(void);
-extern void exit_rcu(void);
int rcu_preempt_needs_cpu(void);
static inline int rcu_needs_cpu(int cpu)
@@ -116,7 +106,6 @@
static inline void rcu_note_context_switch(int cpu)
{
rcu_sched_qs(cpu);
- rcu_preempt_note_context_switch();
}
/*
diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h
index e8ee5dd..3c6083c 100644
--- a/include/linux/rcutree.h
+++ b/include/linux/rcutree.h
@@ -45,18 +45,6 @@
rcu_note_context_switch(cpu);
}
-#ifdef CONFIG_TREE_PREEMPT_RCU
-
-extern void exit_rcu(void);
-
-#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */
-
-static inline void exit_rcu(void)
-{
-}
-
-#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
-
extern void synchronize_rcu_bh(void);
extern void synchronize_sched_expedited(void);
extern void synchronize_rcu_expedited(void);
@@ -98,13 +86,6 @@
extern void rcu_bh_force_quiescent_state(void);
extern void rcu_sched_force_quiescent_state(void);
-/* A context switch is a grace period for RCU-sched and RCU-bh. */
-static inline int rcu_blocking_is_gp(void)
-{
- might_sleep(); /* Check for RCU read-side critical section. */
- return num_online_cpus() == 1;
-}
-
extern void rcu_scheduler_starting(void);
extern int rcu_scheduler_active __read_mostly;
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 81a173c..8f3fd94 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1905,12 +1905,22 @@
INIT_LIST_HEAD(&p->rcu_node_entry);
}
+static inline void rcu_switch_from(struct task_struct *prev)
+{
+ if (prev->rcu_read_lock_nesting != 0)
+ rcu_preempt_note_context_switch();
+}
+
#else
static inline void rcu_copy_process(struct task_struct *p)
{
}
+static inline void rcu_switch_from(struct task_struct *prev)
+{
+}
+
#endif
#ifdef CONFIG_SMP
diff --git a/include/trace/events/rcu.h b/include/trace/events/rcu.h
index 3370997..1480900 100644
--- a/include/trace/events/rcu.h
+++ b/include/trace/events/rcu.h
@@ -292,6 +292,8 @@
* "More callbacks": Still more callbacks, try again to clear them out.
* "Callbacks drained": All callbacks processed, off to dyntick idle!
* "Timer": Timer fired to cause CPU to continue processing callbacks.
+ * "Demigrate": Timer fired on wrong CPU, woke up correct CPU.
+ * "Cleanup after idle": Idle exited, timer canceled.
*/
TRACE_EVENT(rcu_prep_idle,
diff --git a/init/Kconfig b/init/Kconfig
index 6cfd71d..6d18ef8 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -458,6 +458,33 @@
Select a specific number if testing RCU itself.
Take the default if unsure.
+config RCU_FANOUT_LEAF
+ int "Tree-based hierarchical RCU leaf-level fanout value"
+ range 2 RCU_FANOUT if 64BIT
+ range 2 RCU_FANOUT if !64BIT
+ depends on TREE_RCU || TREE_PREEMPT_RCU
+ default 16
+ help
+ This option controls the leaf-level fanout of hierarchical
+ implementations of RCU, and allows trading off cache misses
+ against lock contention. Systems that synchronize their
+ scheduling-clock interrupts for energy-efficiency reasons will
+ want the default because the smaller leaf-level fanout keeps
+ lock contention levels acceptably low. Very large systems
+ (hundreds or thousands of CPUs) will instead want to set this
+ value to the maximum value possible in order to reduce the
+ number of cache misses incurred during RCU's grace-period
+ initialization. These systems tend to run CPU-bound, and thus
+ are not helped by synchronized interrupts, and thus tend to
+ skew them, which reduces lock contention enough that large
+ leaf-level fanouts work well.
+
+ Select a specific number if testing RCU itself.
+
+ Select the maximum permissible value for large systems.
+
+ Take the default if unsure.
+
config RCU_FANOUT_EXACT
bool "Disable tree-based hierarchical RCU auto-balancing"
depends on TREE_RCU || TREE_PREEMPT_RCU
@@ -515,10 +542,25 @@
depends on RCU_BOOST
default 1
help
- This option specifies the real-time priority to which preempted
- RCU readers are to be boosted. If you are working with CPU-bound
- real-time applications, you should specify a priority higher then
- the highest-priority CPU-bound application.
+ This option specifies the real-time priority to which long-term
+ preempted RCU readers are to be boosted. If you are working
+ with a real-time application that has one or more CPU-bound
+ threads running at a real-time priority level, you should set
+ RCU_BOOST_PRIO to a priority higher then the highest-priority
+ real-time CPU-bound thread. The default RCU_BOOST_PRIO value
+ of 1 is appropriate in the common case, which is real-time
+ applications that do not have any CPU-bound threads.
+
+ Some real-time applications might not have a single real-time
+ thread that saturates a given CPU, but instead might have
+ multiple real-time threads that, taken together, fully utilize
+ that CPU. In this case, you should set RCU_BOOST_PRIO to
+ a priority higher than the lowest-priority thread that is
+ conspiring to prevent the CPU from running any non-real-time
+ tasks. For example, if one thread at priority 10 and another
+ thread at priority 5 are between themselves fully consuming
+ the CPU time on a given CPU, then RCU_BOOST_PRIO should be
+ set to priority 6 or higher.
Specify the real-time priority, or take the default if unsure.
diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
index a86f174..95cba41 100644
--- a/kernel/rcupdate.c
+++ b/kernel/rcupdate.c
@@ -51,6 +51,34 @@
#include "rcu.h"
+#ifdef CONFIG_PREEMPT_RCU
+
+/*
+ * Check for a task exiting while in a preemptible-RCU read-side
+ * critical section, clean up if so. No need to issue warnings,
+ * as debug_check_no_locks_held() already does this if lockdep
+ * is enabled.
+ */
+void exit_rcu(void)
+{
+ struct task_struct *t = current;
+
+ if (likely(list_empty(¤t->rcu_node_entry)))
+ return;
+ t->rcu_read_lock_nesting = 1;
+ barrier();
+ t->rcu_read_unlock_special = RCU_READ_UNLOCK_BLOCKED;
+ __rcu_read_unlock();
+}
+
+#else /* #ifdef CONFIG_PREEMPT_RCU */
+
+void exit_rcu(void)
+{
+}
+
+#endif /* #else #ifdef CONFIG_PREEMPT_RCU */
+
#ifdef CONFIG_DEBUG_LOCK_ALLOC
static struct lock_class_key rcu_lock_key;
struct lockdep_map rcu_lock_map =
diff --git a/kernel/rcutiny_plugin.h b/kernel/rcutiny_plugin.h
index 22ecea0..fc31a2d 100644
--- a/kernel/rcutiny_plugin.h
+++ b/kernel/rcutiny_plugin.h
@@ -851,22 +851,6 @@
return rcu_preempt_ctrlblk.rcb.rcucblist != NULL;
}
-/*
- * Check for a task exiting while in a preemptible -RCU read-side
- * critical section, clean up if so. No need to issue warnings,
- * as debug_check_no_locks_held() already does this if lockdep
- * is enabled.
- */
-void exit_rcu(void)
-{
- struct task_struct *t = current;
-
- if (t->rcu_read_lock_nesting == 0)
- return;
- t->rcu_read_lock_nesting = 1;
- __rcu_read_unlock();
-}
-
#else /* #ifdef CONFIG_TINY_PREEMPT_RCU */
#ifdef CONFIG_RCU_TRACE
diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index 1050d6d..b3ea3ac 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -75,6 +75,8 @@
.gpnum = -300, \
.completed = -300, \
.onofflock = __RAW_SPIN_LOCK_UNLOCKED(&structname##_state.onofflock), \
+ .orphan_nxttail = &structname##_state.orphan_nxtlist, \
+ .orphan_donetail = &structname##_state.orphan_donelist, \
.fqslock = __RAW_SPIN_LOCK_UNLOCKED(&structname##_state.fqslock), \
.n_force_qs = 0, \
.n_force_qs_ngp = 0, \
@@ -145,6 +147,13 @@
unsigned long rcutorture_testseq;
unsigned long rcutorture_vernum;
+/* State information for rcu_barrier() and friends. */
+
+static DEFINE_PER_CPU(struct rcu_head, rcu_barrier_head) = {NULL};
+static atomic_t rcu_barrier_cpu_count;
+static DEFINE_MUTEX(rcu_barrier_mutex);
+static struct completion rcu_barrier_completion;
+
/*
* Return true if an RCU grace period is in progress. The ACCESS_ONCE()s
* permit this function to be invoked without holding the root rcu_node
@@ -192,7 +201,6 @@
{
trace_rcu_utilization("Start context switch");
rcu_sched_qs(cpu);
- rcu_preempt_note_context_switch(cpu);
trace_rcu_utilization("End context switch");
}
EXPORT_SYMBOL_GPL(rcu_note_context_switch);
@@ -1311,95 +1319,133 @@
#ifdef CONFIG_HOTPLUG_CPU
/*
- * Move a dying CPU's RCU callbacks to online CPU's callback list.
- * Also record a quiescent state for this CPU for the current grace period.
- * Synchronization and interrupt disabling are not required because
- * this function executes in stop_machine() context. Therefore, cleanup
- * operations that might block must be done later from the CPU_DEAD
- * notifier.
- *
- * Note that the outgoing CPU's bit has already been cleared in the
- * cpu_online_mask. This allows us to randomly pick a callback
- * destination from the bits set in that mask.
+ * Send the specified CPU's RCU callbacks to the orphanage. The
+ * specified CPU must be offline, and the caller must hold the
+ * ->onofflock.
*/
-static void rcu_cleanup_dying_cpu(struct rcu_state *rsp)
+static void
+rcu_send_cbs_to_orphanage(int cpu, struct rcu_state *rsp,
+ struct rcu_node *rnp, struct rcu_data *rdp)
{
int i;
- unsigned long mask;
- int receive_cpu = cpumask_any(cpu_online_mask);
- struct rcu_data *rdp = this_cpu_ptr(rsp->rda);
- struct rcu_data *receive_rdp = per_cpu_ptr(rsp->rda, receive_cpu);
- RCU_TRACE(struct rcu_node *rnp = rdp->mynode); /* For dying CPU. */
- /* First, adjust the counts. */
+ /*
+ * Orphan the callbacks. First adjust the counts. This is safe
+ * because ->onofflock excludes _rcu_barrier()'s adoption of
+ * the callbacks, thus no memory barrier is required.
+ */
if (rdp->nxtlist != NULL) {
- receive_rdp->qlen_lazy += rdp->qlen_lazy;
- receive_rdp->qlen += rdp->qlen;
+ rsp->qlen_lazy += rdp->qlen_lazy;
+ rsp->qlen += rdp->qlen;
+ rdp->n_cbs_orphaned += rdp->qlen;
rdp->qlen_lazy = 0;
rdp->qlen = 0;
}
/*
- * Next, move ready-to-invoke callbacks to be invoked on some
- * other CPU. These will not be required to pass through another
- * grace period: They are done, regardless of CPU.
+ * Next, move those callbacks still needing a grace period to
+ * the orphanage, where some other CPU will pick them up.
+ * Some of the callbacks might have gone partway through a grace
+ * period, but that is too bad. They get to start over because we
+ * cannot assume that grace periods are synchronized across CPUs.
+ * We don't bother updating the ->nxttail[] array yet, instead
+ * we just reset the whole thing later on.
*/
- if (rdp->nxtlist != NULL &&
- rdp->nxttail[RCU_DONE_TAIL] != &rdp->nxtlist) {
- struct rcu_head *oldhead;
- struct rcu_head **oldtail;
- struct rcu_head **newtail;
-
- oldhead = rdp->nxtlist;
- oldtail = receive_rdp->nxttail[RCU_DONE_TAIL];
- rdp->nxtlist = *rdp->nxttail[RCU_DONE_TAIL];
- *rdp->nxttail[RCU_DONE_TAIL] = *oldtail;
- *receive_rdp->nxttail[RCU_DONE_TAIL] = oldhead;
- newtail = rdp->nxttail[RCU_DONE_TAIL];
- for (i = RCU_DONE_TAIL; i < RCU_NEXT_SIZE; i++) {
- if (receive_rdp->nxttail[i] == oldtail)
- receive_rdp->nxttail[i] = newtail;
- if (rdp->nxttail[i] == newtail)
- rdp->nxttail[i] = &rdp->nxtlist;
- }
+ if (*rdp->nxttail[RCU_DONE_TAIL] != NULL) {
+ *rsp->orphan_nxttail = *rdp->nxttail[RCU_DONE_TAIL];
+ rsp->orphan_nxttail = rdp->nxttail[RCU_NEXT_TAIL];
+ *rdp->nxttail[RCU_DONE_TAIL] = NULL;
}
/*
- * Finally, put the rest of the callbacks at the end of the list.
- * The ones that made it partway through get to start over: We
- * cannot assume that grace periods are synchronized across CPUs.
- * (We could splice RCU_WAIT_TAIL into RCU_NEXT_READY_TAIL, but
- * this does not seem compelling. Not yet, anyway.)
+ * Then move the ready-to-invoke callbacks to the orphanage,
+ * where some other CPU will pick them up. These will not be
+ * required to pass though another grace period: They are done.
*/
if (rdp->nxtlist != NULL) {
- *receive_rdp->nxttail[RCU_NEXT_TAIL] = rdp->nxtlist;
- receive_rdp->nxttail[RCU_NEXT_TAIL] =
- rdp->nxttail[RCU_NEXT_TAIL];
- receive_rdp->n_cbs_adopted += rdp->qlen;
- rdp->n_cbs_orphaned += rdp->qlen;
-
- rdp->nxtlist = NULL;
- for (i = 0; i < RCU_NEXT_SIZE; i++)
- rdp->nxttail[i] = &rdp->nxtlist;
+ *rsp->orphan_donetail = rdp->nxtlist;
+ rsp->orphan_donetail = rdp->nxttail[RCU_DONE_TAIL];
}
+ /* Finally, initialize the rcu_data structure's list to empty. */
+ rdp->nxtlist = NULL;
+ for (i = 0; i < RCU_NEXT_SIZE; i++)
+ rdp->nxttail[i] = &rdp->nxtlist;
+}
+
+/*
+ * Adopt the RCU callbacks from the specified rcu_state structure's
+ * orphanage. The caller must hold the ->onofflock.
+ */
+static void rcu_adopt_orphan_cbs(struct rcu_state *rsp)
+{
+ int i;
+ struct rcu_data *rdp = __this_cpu_ptr(rsp->rda);
+
/*
- * Record a quiescent state for the dying CPU. This is safe
- * only because we have already cleared out the callbacks.
- * (Otherwise, the RCU core might try to schedule the invocation
- * of callbacks on this now-offline CPU, which would be bad.)
+ * If there is an rcu_barrier() operation in progress, then
+ * only the task doing that operation is permitted to adopt
+ * callbacks. To do otherwise breaks rcu_barrier() and friends
+ * by causing them to fail to wait for the callbacks in the
+ * orphanage.
*/
- mask = rdp->grpmask; /* rnp->grplo is constant. */
+ if (rsp->rcu_barrier_in_progress &&
+ rsp->rcu_barrier_in_progress != current)
+ return;
+
+ /* Do the accounting first. */
+ rdp->qlen_lazy += rsp->qlen_lazy;
+ rdp->qlen += rsp->qlen;
+ rdp->n_cbs_adopted += rsp->qlen;
+ rsp->qlen_lazy = 0;
+ rsp->qlen = 0;
+
+ /*
+ * We do not need a memory barrier here because the only way we
+ * can get here if there is an rcu_barrier() in flight is if
+ * we are the task doing the rcu_barrier().
+ */
+
+ /* First adopt the ready-to-invoke callbacks. */
+ if (rsp->orphan_donelist != NULL) {
+ *rsp->orphan_donetail = *rdp->nxttail[RCU_DONE_TAIL];
+ *rdp->nxttail[RCU_DONE_TAIL] = rsp->orphan_donelist;
+ for (i = RCU_NEXT_SIZE - 1; i >= RCU_DONE_TAIL; i--)
+ if (rdp->nxttail[i] == rdp->nxttail[RCU_DONE_TAIL])
+ rdp->nxttail[i] = rsp->orphan_donetail;
+ rsp->orphan_donelist = NULL;
+ rsp->orphan_donetail = &rsp->orphan_donelist;
+ }
+
+ /* And then adopt the callbacks that still need a grace period. */
+ if (rsp->orphan_nxtlist != NULL) {
+ *rdp->nxttail[RCU_NEXT_TAIL] = rsp->orphan_nxtlist;
+ rdp->nxttail[RCU_NEXT_TAIL] = rsp->orphan_nxttail;
+ rsp->orphan_nxtlist = NULL;
+ rsp->orphan_nxttail = &rsp->orphan_nxtlist;
+ }
+}
+
+/*
+ * Trace the fact that this CPU is going offline.
+ */
+static void rcu_cleanup_dying_cpu(struct rcu_state *rsp)
+{
+ RCU_TRACE(unsigned long mask);
+ RCU_TRACE(struct rcu_data *rdp = this_cpu_ptr(rsp->rda));
+ RCU_TRACE(struct rcu_node *rnp = rdp->mynode);
+
+ RCU_TRACE(mask = rdp->grpmask);
trace_rcu_grace_period(rsp->name,
rnp->gpnum + 1 - !!(rnp->qsmask & mask),
"cpuofl");
- rcu_report_qs_rdp(smp_processor_id(), rsp, rdp, rsp->gpnum);
- /* Note that rcu_report_qs_rdp() might call trace_rcu_grace_period(). */
}
/*
* The CPU has been completely removed, and some other CPU is reporting
- * this fact from process context. Do the remainder of the cleanup.
+ * this fact from process context. Do the remainder of the cleanup,
+ * including orphaning the outgoing CPU's RCU callbacks, and also
+ * adopting them, if there is no _rcu_barrier() instance running.
* There can only be one CPU hotplug operation at a time, so no other
* CPU can be attempting to update rcu_cpu_kthread_task.
*/
@@ -1409,17 +1455,21 @@
unsigned long mask;
int need_report = 0;
struct rcu_data *rdp = per_cpu_ptr(rsp->rda, cpu);
- struct rcu_node *rnp = rdp->mynode; /* Outgoing CPU's rnp. */
+ struct rcu_node *rnp = rdp->mynode; /* Outgoing CPU's rdp & rnp. */
/* Adjust any no-longer-needed kthreads. */
rcu_stop_cpu_kthread(cpu);
rcu_node_kthread_setaffinity(rnp, -1);
- /* Remove the dying CPU from the bitmasks in the rcu_node hierarchy. */
+ /* Remove the dead CPU from the bitmasks in the rcu_node hierarchy. */
/* Exclude any attempts to start a new grace period. */
raw_spin_lock_irqsave(&rsp->onofflock, flags);
+ /* Orphan the dead CPU's callbacks, and adopt them if appropriate. */
+ rcu_send_cbs_to_orphanage(cpu, rsp, rnp, rdp);
+ rcu_adopt_orphan_cbs(rsp);
+
/* Remove the outgoing CPU from the masks in the rcu_node hierarchy. */
mask = rdp->grpmask; /* rnp->grplo is constant. */
do {
@@ -1456,6 +1506,10 @@
#else /* #ifdef CONFIG_HOTPLUG_CPU */
+static void rcu_adopt_orphan_cbs(struct rcu_state *rsp)
+{
+}
+
static void rcu_cleanup_dying_cpu(struct rcu_state *rsp)
{
}
@@ -1524,9 +1578,6 @@
rcu_is_callbacks_kthread());
/* Update count, and requeue any remaining callbacks. */
- rdp->qlen_lazy -= count_lazy;
- rdp->qlen -= count;
- rdp->n_cbs_invoked += count;
if (list != NULL) {
*tail = rdp->nxtlist;
rdp->nxtlist = list;
@@ -1536,6 +1587,10 @@
else
break;
}
+ smp_mb(); /* List handling before counting for rcu_barrier(). */
+ rdp->qlen_lazy -= count_lazy;
+ rdp->qlen -= count;
+ rdp->n_cbs_invoked += count;
/* Reinstate batch limit if we have worked down the excess. */
if (rdp->blimit == LONG_MAX && rdp->qlen <= qlowmark)
@@ -1824,11 +1879,14 @@
rdp = this_cpu_ptr(rsp->rda);
/* Add the callback to our list. */
- *rdp->nxttail[RCU_NEXT_TAIL] = head;
- rdp->nxttail[RCU_NEXT_TAIL] = &head->next;
rdp->qlen++;
if (lazy)
rdp->qlen_lazy++;
+ else
+ rcu_idle_count_callbacks_posted();
+ smp_mb(); /* Count before adding callback for rcu_barrier(). */
+ *rdp->nxttail[RCU_NEXT_TAIL] = head;
+ rdp->nxttail[RCU_NEXT_TAIL] = &head->next;
if (__is_kfree_rcu_offset((unsigned long)func))
trace_rcu_kfree_callback(rsp->name, head, (unsigned long)func,
@@ -1894,6 +1952,38 @@
}
EXPORT_SYMBOL_GPL(call_rcu_bh);
+/*
+ * Because a context switch is a grace period for RCU-sched and RCU-bh,
+ * any blocking grace-period wait automatically implies a grace period
+ * if there is only one CPU online at any point time during execution
+ * of either synchronize_sched() or synchronize_rcu_bh(). It is OK to
+ * occasionally incorrectly indicate that there are multiple CPUs online
+ * when there was in fact only one the whole time, as this just adds
+ * some overhead: RCU still operates correctly.
+ *
+ * Of course, sampling num_online_cpus() with preemption enabled can
+ * give erroneous results if there are concurrent CPU-hotplug operations.
+ * For example, given a demonic sequence of preemptions in num_online_cpus()
+ * and CPU-hotplug operations, there could be two or more CPUs online at
+ * all times, but num_online_cpus() might well return one (or even zero).
+ *
+ * However, all such demonic sequences require at least one CPU-offline
+ * operation. Furthermore, rcu_blocking_is_gp() giving the wrong answer
+ * is only a problem if there is an RCU read-side critical section executing
+ * throughout. But RCU-sched and RCU-bh read-side critical sections
+ * disable either preemption or bh, which prevents a CPU from going offline.
+ * Therefore, the only way that rcu_blocking_is_gp() can incorrectly return
+ * that there is only one CPU when in fact there was more than one throughout
+ * is when there were no RCU readers in the system. If there are no
+ * RCU readers, the grace period by definition can be of zero length,
+ * regardless of the number of online CPUs.
+ */
+static inline int rcu_blocking_is_gp(void)
+{
+ might_sleep(); /* Check for RCU read-side critical section. */
+ return num_online_cpus() <= 1;
+}
+
/**
* synchronize_sched - wait until an rcu-sched grace period has elapsed.
*
@@ -2167,11 +2257,10 @@
rcu_preempt_cpu_has_callbacks(cpu);
}
-static DEFINE_PER_CPU(struct rcu_head, rcu_barrier_head) = {NULL};
-static atomic_t rcu_barrier_cpu_count;
-static DEFINE_MUTEX(rcu_barrier_mutex);
-static struct completion rcu_barrier_completion;
-
+/*
+ * RCU callback function for _rcu_barrier(). If we are last, wake
+ * up the task executing _rcu_barrier().
+ */
static void rcu_barrier_callback(struct rcu_head *notused)
{
if (atomic_dec_and_test(&rcu_barrier_cpu_count))
@@ -2201,27 +2290,94 @@
void (*call_rcu_func)(struct rcu_head *head,
void (*func)(struct rcu_head *head)))
{
- BUG_ON(in_interrupt());
+ int cpu;
+ unsigned long flags;
+ struct rcu_data *rdp;
+ struct rcu_head rh;
+
+ init_rcu_head_on_stack(&rh);
+
/* Take mutex to serialize concurrent rcu_barrier() requests. */
mutex_lock(&rcu_barrier_mutex);
- init_completion(&rcu_barrier_completion);
+
+ smp_mb(); /* Prevent any prior operations from leaking in. */
+
/*
- * Initialize rcu_barrier_cpu_count to 1, then invoke
- * rcu_barrier_func() on each CPU, so that each CPU also has
- * incremented rcu_barrier_cpu_count. Only then is it safe to
- * decrement rcu_barrier_cpu_count -- otherwise the first CPU
- * might complete its grace period before all of the other CPUs
- * did their increment, causing this function to return too
- * early. Note that on_each_cpu() disables irqs, which prevents
- * any CPUs from coming online or going offline until each online
- * CPU has queued its RCU-barrier callback.
+ * Initialize the count to one rather than to zero in order to
+ * avoid a too-soon return to zero in case of a short grace period
+ * (or preemption of this task). Also flag this task as doing
+ * an rcu_barrier(). This will prevent anyone else from adopting
+ * orphaned callbacks, which could cause otherwise failure if a
+ * CPU went offline and quickly came back online. To see this,
+ * consider the following sequence of events:
+ *
+ * 1. We cause CPU 0 to post an rcu_barrier_callback() callback.
+ * 2. CPU 1 goes offline, orphaning its callbacks.
+ * 3. CPU 0 adopts CPU 1's orphaned callbacks.
+ * 4. CPU 1 comes back online.
+ * 5. We cause CPU 1 to post an rcu_barrier_callback() callback.
+ * 6. Both rcu_barrier_callback() callbacks are invoked, awakening
+ * us -- but before CPU 1's orphaned callbacks are invoked!!!
*/
+ init_completion(&rcu_barrier_completion);
atomic_set(&rcu_barrier_cpu_count, 1);
- on_each_cpu(rcu_barrier_func, (void *)call_rcu_func, 1);
+ raw_spin_lock_irqsave(&rsp->onofflock, flags);
+ rsp->rcu_barrier_in_progress = current;
+ raw_spin_unlock_irqrestore(&rsp->onofflock, flags);
+
+ /*
+ * Force every CPU with callbacks to register a new callback
+ * that will tell us when all the preceding callbacks have
+ * been invoked. If an offline CPU has callbacks, wait for
+ * it to either come back online or to finish orphaning those
+ * callbacks.
+ */
+ for_each_possible_cpu(cpu) {
+ preempt_disable();
+ rdp = per_cpu_ptr(rsp->rda, cpu);
+ if (cpu_is_offline(cpu)) {
+ preempt_enable();
+ while (cpu_is_offline(cpu) && ACCESS_ONCE(rdp->qlen))
+ schedule_timeout_interruptible(1);
+ } else if (ACCESS_ONCE(rdp->qlen)) {
+ smp_call_function_single(cpu, rcu_barrier_func,
+ (void *)call_rcu_func, 1);
+ preempt_enable();
+ } else {
+ preempt_enable();
+ }
+ }
+
+ /*
+ * Now that all online CPUs have rcu_barrier_callback() callbacks
+ * posted, we can adopt all of the orphaned callbacks and place
+ * an rcu_barrier_callback() callback after them. When that is done,
+ * we are guaranteed to have an rcu_barrier_callback() callback
+ * following every callback that could possibly have been
+ * registered before _rcu_barrier() was called.
+ */
+ raw_spin_lock_irqsave(&rsp->onofflock, flags);
+ rcu_adopt_orphan_cbs(rsp);
+ rsp->rcu_barrier_in_progress = NULL;
+ raw_spin_unlock_irqrestore(&rsp->onofflock, flags);
+ atomic_inc(&rcu_barrier_cpu_count);
+ smp_mb__after_atomic_inc(); /* Ensure atomic_inc() before callback. */
+ call_rcu_func(&rh, rcu_barrier_callback);
+
+ /*
+ * Now that we have an rcu_barrier_callback() callback on each
+ * CPU, and thus each counted, remove the initial count.
+ */
if (atomic_dec_and_test(&rcu_barrier_cpu_count))
complete(&rcu_barrier_completion);
+
+ /* Wait for all rcu_barrier_callback() callbacks to be invoked. */
wait_for_completion(&rcu_barrier_completion);
+
+ /* Other rcu_barrier() invocations can now safely proceed. */
mutex_unlock(&rcu_barrier_mutex);
+
+ destroy_rcu_head_on_stack(&rh);
}
/**
@@ -2418,7 +2574,7 @@
for (i = NUM_RCU_LVLS - 1; i > 0; i--)
rsp->levelspread[i] = CONFIG_RCU_FANOUT;
- rsp->levelspread[0] = RCU_FANOUT_LEAF;
+ rsp->levelspread[0] = CONFIG_RCU_FANOUT_LEAF;
}
#else /* #ifdef CONFIG_RCU_FANOUT_EXACT */
static void __init rcu_init_levelspread(struct rcu_state *rsp)
diff --git a/kernel/rcutree.h b/kernel/rcutree.h
index cdd1be0..7f5d138 100644
--- a/kernel/rcutree.h
+++ b/kernel/rcutree.h
@@ -29,18 +29,14 @@
#include <linux/seqlock.h>
/*
- * Define shape of hierarchy based on NR_CPUS and CONFIG_RCU_FANOUT.
+ * Define shape of hierarchy based on NR_CPUS, CONFIG_RCU_FANOUT, and
+ * CONFIG_RCU_FANOUT_LEAF.
* In theory, it should be possible to add more levels straightforwardly.
* In practice, this did work well going from three levels to four.
* Of course, your mileage may vary.
*/
#define MAX_RCU_LVLS 4
-#if CONFIG_RCU_FANOUT > 16
-#define RCU_FANOUT_LEAF 16
-#else /* #if CONFIG_RCU_FANOUT > 16 */
-#define RCU_FANOUT_LEAF (CONFIG_RCU_FANOUT)
-#endif /* #else #if CONFIG_RCU_FANOUT > 16 */
-#define RCU_FANOUT_1 (RCU_FANOUT_LEAF)
+#define RCU_FANOUT_1 (CONFIG_RCU_FANOUT_LEAF)
#define RCU_FANOUT_2 (RCU_FANOUT_1 * CONFIG_RCU_FANOUT)
#define RCU_FANOUT_3 (RCU_FANOUT_2 * CONFIG_RCU_FANOUT)
#define RCU_FANOUT_4 (RCU_FANOUT_3 * CONFIG_RCU_FANOUT)
@@ -371,6 +367,17 @@
raw_spinlock_t onofflock; /* exclude on/offline and */
/* starting new GP. */
+ struct rcu_head *orphan_nxtlist; /* Orphaned callbacks that */
+ /* need a grace period. */
+ struct rcu_head **orphan_nxttail; /* Tail of above. */
+ struct rcu_head *orphan_donelist; /* Orphaned callbacks that */
+ /* are ready to invoke. */
+ struct rcu_head **orphan_donetail; /* Tail of above. */
+ long qlen_lazy; /* Number of lazy callbacks. */
+ long qlen; /* Total number of callbacks. */
+ struct task_struct *rcu_barrier_in_progress;
+ /* Task doing rcu_barrier(), */
+ /* or NULL if no barrier. */
raw_spinlock_t fqslock; /* Only one task forcing */
/* quiescent states. */
unsigned long jiffies_force_qs; /* Time at which to invoke */
@@ -423,7 +430,6 @@
/* Forward declarations for rcutree_plugin.h */
static void rcu_bootup_announce(void);
long rcu_batches_completed(void);
-static void rcu_preempt_note_context_switch(int cpu);
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp);
#ifdef CONFIG_HOTPLUG_CPU
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp,
@@ -471,6 +477,7 @@
static void rcu_prepare_for_idle_init(int cpu);
static void rcu_cleanup_after_idle(int cpu);
static void rcu_prepare_for_idle(int cpu);
+static void rcu_idle_count_callbacks_posted(void);
static void print_cpu_stall_info_begin(void);
static void print_cpu_stall_info(struct rcu_state *rsp, int cpu);
static void print_cpu_stall_info_end(void);
diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h
index c023464..2411000 100644
--- a/kernel/rcutree_plugin.h
+++ b/kernel/rcutree_plugin.h
@@ -153,7 +153,7 @@
*
* Caller must disable preemption.
*/
-static void rcu_preempt_note_context_switch(int cpu)
+void rcu_preempt_note_context_switch(void)
{
struct task_struct *t = current;
unsigned long flags;
@@ -164,7 +164,7 @@
(t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) {
/* Possibly blocking in an RCU read-side critical section. */
- rdp = per_cpu_ptr(rcu_preempt_state.rda, cpu);
+ rdp = __this_cpu_ptr(rcu_preempt_state.rda);
rnp = rdp->mynode;
raw_spin_lock_irqsave(&rnp->lock, flags);
t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED;
@@ -228,7 +228,7 @@
* means that we continue to block the current grace period.
*/
local_irq_save(flags);
- rcu_preempt_qs(cpu);
+ rcu_preempt_qs(smp_processor_id());
local_irq_restore(flags);
}
@@ -969,22 +969,6 @@
rcu_init_one(&rcu_preempt_state, &rcu_preempt_data);
}
-/*
- * Check for a task exiting while in a preemptible-RCU read-side
- * critical section, clean up if so. No need to issue warnings,
- * as debug_check_no_locks_held() already does this if lockdep
- * is enabled.
- */
-void exit_rcu(void)
-{
- struct task_struct *t = current;
-
- if (t->rcu_read_lock_nesting == 0)
- return;
- t->rcu_read_lock_nesting = 1;
- __rcu_read_unlock();
-}
-
#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */
static struct rcu_state *rcu_state = &rcu_sched_state;
@@ -1018,14 +1002,6 @@
EXPORT_SYMBOL_GPL(rcu_force_quiescent_state);
/*
- * Because preemptible RCU does not exist, we never have to check for
- * CPUs being in quiescent states.
- */
-static void rcu_preempt_note_context_switch(int cpu)
-{
-}
-
-/*
* Because preemptible RCU does not exist, there are never any preempted
* RCU readers.
*/
@@ -1938,6 +1914,14 @@
{
}
+/*
+ * Don't bother keeping a running count of the number of RCU callbacks
+ * posted because CONFIG_RCU_FAST_NO_HZ=n.
+ */
+static void rcu_idle_count_callbacks_posted(void)
+{
+}
+
#else /* #if !defined(CONFIG_RCU_FAST_NO_HZ) */
/*
@@ -1978,11 +1962,20 @@
#define RCU_IDLE_GP_DELAY 6 /* Roughly one grace period. */
#define RCU_IDLE_LAZY_GP_DELAY (6 * HZ) /* Roughly six seconds. */
+/* Loop counter for rcu_prepare_for_idle(). */
static DEFINE_PER_CPU(int, rcu_dyntick_drain);
+/* If rcu_dyntick_holdoff==jiffies, don't try to enter dyntick-idle mode. */
static DEFINE_PER_CPU(unsigned long, rcu_dyntick_holdoff);
-static DEFINE_PER_CPU(struct hrtimer, rcu_idle_gp_timer);
-static ktime_t rcu_idle_gp_wait; /* If some non-lazy callbacks. */
-static ktime_t rcu_idle_lazy_gp_wait; /* If only lazy callbacks. */
+/* Timer to awaken the CPU if it enters dyntick-idle mode with callbacks. */
+static DEFINE_PER_CPU(struct timer_list, rcu_idle_gp_timer);
+/* Scheduled expiry time for rcu_idle_gp_timer to allow reposting. */
+static DEFINE_PER_CPU(unsigned long, rcu_idle_gp_timer_expires);
+/* Enable special processing on first attempt to enter dyntick-idle mode. */
+static DEFINE_PER_CPU(bool, rcu_idle_first_pass);
+/* Running count of non-lazy callbacks posted, never decremented. */
+static DEFINE_PER_CPU(unsigned long, rcu_nonlazy_posted);
+/* Snapshot of rcu_nonlazy_posted to detect meaningful exits from idle. */
+static DEFINE_PER_CPU(unsigned long, rcu_nonlazy_posted_snap);
/*
* Allow the CPU to enter dyntick-idle mode if either: (1) There are no
@@ -1995,6 +1988,8 @@
*/
int rcu_needs_cpu(int cpu)
{
+ /* Flag a new idle sojourn to the idle-entry state machine. */
+ per_cpu(rcu_idle_first_pass, cpu) = 1;
/* If no callbacks, RCU doesn't need the CPU. */
if (!rcu_cpu_has_callbacks(cpu))
return 0;
@@ -2045,16 +2040,34 @@
}
/*
+ * Handler for smp_call_function_single(). The only point of this
+ * handler is to wake the CPU up, so the handler does only tracing.
+ */
+void rcu_idle_demigrate(void *unused)
+{
+ trace_rcu_prep_idle("Demigrate");
+}
+
+/*
* Timer handler used to force CPU to start pushing its remaining RCU
* callbacks in the case where it entered dyntick-idle mode with callbacks
* pending. The hander doesn't really need to do anything because the
* real work is done upon re-entry to idle, or by the next scheduling-clock
* interrupt should idle not be re-entered.
+ *
+ * One special case: the timer gets migrated without awakening the CPU
+ * on which the timer was scheduled on. In this case, we must wake up
+ * that CPU. We do so with smp_call_function_single().
*/
-static enum hrtimer_restart rcu_idle_gp_timer_func(struct hrtimer *hrtp)
+static void rcu_idle_gp_timer_func(unsigned long cpu_in)
{
+ int cpu = (int)cpu_in;
+
trace_rcu_prep_idle("Timer");
- return HRTIMER_NORESTART;
+ if (cpu != smp_processor_id())
+ smp_call_function_single(cpu, rcu_idle_demigrate, NULL, 0);
+ else
+ WARN_ON_ONCE(1); /* Getting here can hang the system... */
}
/*
@@ -2062,19 +2075,11 @@
*/
static void rcu_prepare_for_idle_init(int cpu)
{
- static int firsttime = 1;
- struct hrtimer *hrtp = &per_cpu(rcu_idle_gp_timer, cpu);
-
- hrtimer_init(hrtp, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
- hrtp->function = rcu_idle_gp_timer_func;
- if (firsttime) {
- unsigned int upj = jiffies_to_usecs(RCU_IDLE_GP_DELAY);
-
- rcu_idle_gp_wait = ns_to_ktime(upj * (u64)1000);
- upj = jiffies_to_usecs(RCU_IDLE_LAZY_GP_DELAY);
- rcu_idle_lazy_gp_wait = ns_to_ktime(upj * (u64)1000);
- firsttime = 0;
- }
+ per_cpu(rcu_dyntick_holdoff, cpu) = jiffies - 1;
+ setup_timer(&per_cpu(rcu_idle_gp_timer, cpu),
+ rcu_idle_gp_timer_func, cpu);
+ per_cpu(rcu_idle_gp_timer_expires, cpu) = jiffies - 1;
+ per_cpu(rcu_idle_first_pass, cpu) = 1;
}
/*
@@ -2084,7 +2089,8 @@
*/
static void rcu_cleanup_after_idle(int cpu)
{
- hrtimer_cancel(&per_cpu(rcu_idle_gp_timer, cpu));
+ del_timer(&per_cpu(rcu_idle_gp_timer, cpu));
+ trace_rcu_prep_idle("Cleanup after idle");
}
/*
@@ -2108,6 +2114,29 @@
*/
static void rcu_prepare_for_idle(int cpu)
{
+ struct timer_list *tp;
+
+ /*
+ * If this is an idle re-entry, for example, due to use of
+ * RCU_NONIDLE() or the new idle-loop tracing API within the idle
+ * loop, then don't take any state-machine actions, unless the
+ * momentary exit from idle queued additional non-lazy callbacks.
+ * Instead, repost the rcu_idle_gp_timer if this CPU has callbacks
+ * pending.
+ */
+ if (!per_cpu(rcu_idle_first_pass, cpu) &&
+ (per_cpu(rcu_nonlazy_posted, cpu) ==
+ per_cpu(rcu_nonlazy_posted_snap, cpu))) {
+ if (rcu_cpu_has_callbacks(cpu)) {
+ tp = &per_cpu(rcu_idle_gp_timer, cpu);
+ mod_timer_pinned(tp, per_cpu(rcu_idle_gp_timer_expires, cpu));
+ }
+ return;
+ }
+ per_cpu(rcu_idle_first_pass, cpu) = 0;
+ per_cpu(rcu_nonlazy_posted_snap, cpu) =
+ per_cpu(rcu_nonlazy_posted, cpu) - 1;
+
/*
* If there are no callbacks on this CPU, enter dyntick-idle mode.
* Also reset state to avoid prejudicing later attempts.
@@ -2140,11 +2169,15 @@
per_cpu(rcu_dyntick_drain, cpu) = 0;
per_cpu(rcu_dyntick_holdoff, cpu) = jiffies;
if (rcu_cpu_has_nonlazy_callbacks(cpu))
- hrtimer_start(&per_cpu(rcu_idle_gp_timer, cpu),
- rcu_idle_gp_wait, HRTIMER_MODE_REL);
+ per_cpu(rcu_idle_gp_timer_expires, cpu) =
+ jiffies + RCU_IDLE_GP_DELAY;
else
- hrtimer_start(&per_cpu(rcu_idle_gp_timer, cpu),
- rcu_idle_lazy_gp_wait, HRTIMER_MODE_REL);
+ per_cpu(rcu_idle_gp_timer_expires, cpu) =
+ jiffies + RCU_IDLE_LAZY_GP_DELAY;
+ tp = &per_cpu(rcu_idle_gp_timer, cpu);
+ mod_timer_pinned(tp, per_cpu(rcu_idle_gp_timer_expires, cpu));
+ per_cpu(rcu_nonlazy_posted_snap, cpu) =
+ per_cpu(rcu_nonlazy_posted, cpu);
return; /* Nothing more to do immediately. */
} else if (--per_cpu(rcu_dyntick_drain, cpu) <= 0) {
/* We have hit the limit, so time to give up. */
@@ -2184,6 +2217,19 @@
trace_rcu_prep_idle("Callbacks drained");
}
+/*
+ * Keep a running count of the number of non-lazy callbacks posted
+ * on this CPU. This running counter (which is never decremented) allows
+ * rcu_prepare_for_idle() to detect when something out of the idle loop
+ * posts a callback, even if an equal number of callbacks are invoked.
+ * Of course, callbacks should only be posted from within a trace event
+ * designed to be called from idle or from within RCU_NONIDLE().
+ */
+static void rcu_idle_count_callbacks_posted(void)
+{
+ __this_cpu_add(rcu_nonlazy_posted, 1);
+}
+
#endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */
#ifdef CONFIG_RCU_CPU_STALL_INFO
@@ -2192,14 +2238,12 @@
static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
{
- struct hrtimer *hrtp = &per_cpu(rcu_idle_gp_timer, cpu);
+ struct timer_list *tltp = &per_cpu(rcu_idle_gp_timer, cpu);
- sprintf(cp, "drain=%d %c timer=%lld",
+ sprintf(cp, "drain=%d %c timer=%lu",
per_cpu(rcu_dyntick_drain, cpu),
per_cpu(rcu_dyntick_holdoff, cpu) == jiffies ? 'H' : '.',
- hrtimer_active(hrtp)
- ? ktime_to_us(hrtimer_get_remaining(hrtp))
- : -1);
+ timer_pending(tltp) ? tltp->expires - jiffies : -1);
}
#else /* #ifdef CONFIG_RCU_FAST_NO_HZ */
diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c
index ed459ed..d4bc16d 100644
--- a/kernel/rcutree_trace.c
+++ b/kernel/rcutree_trace.c
@@ -271,13 +271,13 @@
gpnum = rsp->gpnum;
seq_printf(m, "c=%lu g=%lu s=%d jfq=%ld j=%x "
- "nfqs=%lu/nfqsng=%lu(%lu) fqlh=%lu\n",
+ "nfqs=%lu/nfqsng=%lu(%lu) fqlh=%lu oqlen=%ld/%ld\n",
rsp->completed, gpnum, rsp->fqs_state,
(long)(rsp->jiffies_force_qs - jiffies),
(int)(jiffies & 0xffff),
rsp->n_force_qs, rsp->n_force_qs_ngp,
rsp->n_force_qs - rsp->n_force_qs_ngp,
- rsp->n_force_qs_lh);
+ rsp->n_force_qs_lh, rsp->qlen_lazy, rsp->qlen);
for (rnp = &rsp->node[0]; rnp - &rsp->node[0] < NUM_RCU_NODES; rnp++) {
if (rnp->level != level) {
seq_puts(m, "\n");
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 4603b9d..5d89eb9 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2083,6 +2083,7 @@
#endif
/* Here we just switch the register state and the stack. */
+ rcu_switch_from(prev);
switch_to(prev, next, prev);
barrier();
diff --git a/kernel/timer.c b/kernel/timer.c
index a297ffc..837c552f 100644
--- a/kernel/timer.c
+++ b/kernel/timer.c
@@ -861,7 +861,13 @@
*
* mod_timer_pinned() is a way to update the expire field of an
* active timer (if the timer is inactive it will be activated)
- * and not allow the timer to be migrated to a different CPU.
+ * and to ensure that the timer is scheduled on the current CPU.
+ *
+ * Note that this does not prevent the timer from being migrated
+ * when the current CPU goes offline. If this is a problem for
+ * you, use CPU-hotplug notifiers to handle it correctly, for
+ * example, cancelling the timer when the corresponding CPU goes
+ * offline.
*
* mod_timer_pinned(timer, expires) is equivalent to:
*
diff --git a/lib/list_debug.c b/lib/list_debug.c
index 982b850..3810b48 100644
--- a/lib/list_debug.c
+++ b/lib/list_debug.c
@@ -10,6 +10,7 @@
#include <linux/list.h>
#include <linux/bug.h>
#include <linux/kernel.h>
+#include <linux/rculist.h>
/*
* Insert a new entry between two known consecutive entries.
@@ -75,3 +76,24 @@
entry->prev = LIST_POISON2;
}
EXPORT_SYMBOL(list_del);
+
+/*
+ * RCU variants.
+ */
+void __list_add_rcu(struct list_head *new,
+ struct list_head *prev, struct list_head *next)
+{
+ WARN(next->prev != prev,
+ "list_add_rcu corruption. next->prev should be "
+ "prev (%p), but was %p. (next=%p).\n",
+ prev, next->prev, next);
+ WARN(prev->next != next,
+ "list_add_rcu corruption. prev->next should be "
+ "next (%p), but was %p. (prev=%p).\n",
+ next, prev->next, prev);
+ new->next = next;
+ new->prev = prev;
+ rcu_assign_pointer(list_next_rcu(prev), new);
+ next->prev = new;
+}
+EXPORT_SYMBOL(__list_add_rcu);