#include <linux/rcupdate.h>
#include <linux/interrupt.h>
#include <linux/sched.h>
+#include <linux/nmi.h>
#include <asm/atomic.h>
#include <linux/bitops.h>
#include <linux/module.h>
struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state);
DEFINE_PER_CPU(struct rcu_data, rcu_bh_data);
+extern long rcu_batches_completed_sched(void);
+static struct rcu_node *rcu_get_root(struct rcu_state *rsp);
+static void cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp,
+ struct rcu_node *rnp, unsigned long flags);
+static void cpu_quiet_msk_finish(struct rcu_state *rsp, unsigned long flags);
+#ifdef CONFIG_HOTPLUG_CPU
+static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp);
+#endif /* #ifdef CONFIG_HOTPLUG_CPU */
+static void __rcu_process_callbacks(struct rcu_state *rsp,
+ struct rcu_data *rdp);
+static void __call_rcu(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu),
+ struct rcu_state *rsp);
+static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp);
+static void __cpuinit rcu_init_percpu_data(int cpu, struct rcu_state *rsp,
+ int preemptable);
+
+#include "rcutree_plugin.h"
+
/*
* Note a quiescent state. Because we do not need to know
* how many quiescent states passed, just if there was at least
*/
void rcu_sched_qs(int cpu)
{
- struct rcu_data *rdp = &per_cpu(rcu_sched_data, cpu);
+ unsigned long flags;
+ struct rcu_data *rdp;
+
+ local_irq_save(flags);
+ rdp = &per_cpu(rcu_sched_data, cpu);
rdp->passed_quiesc = 1;
rdp->passed_quiesc_completed = rdp->completed;
+ rcu_preempt_qs(cpu);
+ local_irq_restore(flags);
}
void rcu_bh_qs(int cpu)
{
- struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu);
+ unsigned long flags;
+ struct rcu_data *rdp;
+
+ local_irq_save(flags);
+ rdp = &per_cpu(rcu_bh_data, cpu);
rdp->passed_quiesc = 1;
rdp->passed_quiesc_completed = rdp->completed;
+ local_irq_restore(flags);
}
#ifdef CONFIG_NO_HZ
static int qlowmark = 100; /* Once only this many pending, use blimit. */
static void force_quiescent_state(struct rcu_state *rsp, int relaxed);
+static int rcu_pending(int cpu);
/*
* Return the number of RCU-sched batches processed thus far for debug & stats.
}
EXPORT_SYMBOL_GPL(rcu_batches_completed_sched);
-/*
- * Return the number of RCU batches processed thus far for debug & stats.
- * @@@ placeholder, maps to rcu_batches_completed_sched().
- */
-long rcu_batches_completed(void)
-{
- return rcu_batches_completed_sched();
-}
-EXPORT_SYMBOL_GPL(rcu_batches_completed);
-
/*
* Return the number of RCU BH batches processed thus far for debug & stats.
*/
return 1;
}
+ /* If preemptable RCU, no point in sending reschedule IPI. */
+ if (rdp->preemptable)
+ return 0;
+
/* The CPU is online, so send it a reschedule IPI. */
if (rdp->cpu != smp_processor_id())
smp_send_reschedule(rdp->cpu);
#endif /* #ifdef CONFIG_SMP */
#ifdef CONFIG_NO_HZ
-static DEFINE_RATELIMIT_STATE(rcu_rs, 10 * HZ, 5);
/**
* rcu_enter_nohz - inform RCU that current CPU is entering nohz
rdtp = &__get_cpu_var(rcu_dynticks);
rdtp->dynticks++;
rdtp->dynticks_nesting--;
- WARN_ON_RATELIMIT(rdtp->dynticks & 0x1, &rcu_rs);
+ WARN_ON_ONCE(rdtp->dynticks & 0x1);
local_irq_restore(flags);
}
rdtp = &__get_cpu_var(rcu_dynticks);
rdtp->dynticks++;
rdtp->dynticks_nesting++;
- WARN_ON_RATELIMIT(!(rdtp->dynticks & 0x1), &rcu_rs);
+ WARN_ON_ONCE(!(rdtp->dynticks & 0x1));
local_irq_restore(flags);
smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */
}
if (rdtp->dynticks & 0x1)
return;
rdtp->dynticks_nmi++;
- WARN_ON_RATELIMIT(!(rdtp->dynticks_nmi & 0x1), &rcu_rs);
+ WARN_ON_ONCE(!(rdtp->dynticks_nmi & 0x1));
smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */
}
return;
smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */
rdtp->dynticks_nmi++;
- WARN_ON_RATELIMIT(rdtp->dynticks_nmi & 0x1, &rcu_rs);
+ WARN_ON_ONCE(rdtp->dynticks_nmi & 0x1);
}
/**
if (rdtp->dynticks_nesting++)
return;
rdtp->dynticks++;
- WARN_ON_RATELIMIT(!(rdtp->dynticks & 0x1), &rcu_rs);
+ WARN_ON_ONCE(!(rdtp->dynticks & 0x1));
smp_mb(); /* CPUs seeing ++ must see later RCU read-side crit sects */
}
return;
smp_mb(); /* CPUs seeing ++ must see prior RCU read-side crit sects */
rdtp->dynticks++;
- WARN_ON_RATELIMIT(rdtp->dynticks & 0x1, &rcu_rs);
+ WARN_ON_ONCE(rdtp->dynticks & 0x1);
/* If the interrupt queued a callback, get out of dyntick mode. */
if (__get_cpu_var(rcu_sched_data).nxtlist ||
printk(KERN_ERR "INFO: RCU detected CPU stalls:");
for (; rnp_cur < rnp_end; rnp_cur++) {
+ rcu_print_task_stall(rnp);
if (rnp_cur->qsmask == 0)
continue;
for (cpu = 0; cpu <= rnp_cur->grphi - rnp_cur->grplo; cpu++)
}
printk(" (detected by %d, t=%ld jiffies)\n",
smp_processor_id(), (long)(jiffies - rsp->gp_start));
+ trigger_all_cpu_backtrace();
+
force_quiescent_state(rsp, 0); /* Kick them all. */
}
printk(KERN_ERR "INFO: RCU detected CPU %d stall (t=%lu jiffies)\n",
smp_processor_id(), jiffies - rsp->gp_start);
- dump_stack();
+ trigger_all_cpu_backtrace();
+
spin_lock_irqsave(&rnp->lock, flags);
if ((long)(jiffies - rsp->jiffies_stall) >= 0)
rsp->jiffies_stall =
jiffies + RCU_SECONDS_TILL_STALL_RECHECK;
spin_unlock_irqrestore(&rnp->lock, flags);
+
set_need_resched(); /* kick ourselves to get things going. */
}
local_irq_restore(flags);
}
+/*
+ * Clean up after the prior grace period and let rcu_start_gp() start up
+ * the next grace period if one is needed. Note that the caller must
+ * hold rnp->lock, as required by rcu_start_gp(), which will release it.
+ */
+static void cpu_quiet_msk_finish(struct rcu_state *rsp, unsigned long flags)
+ __releases(rnp->lock)
+{
+ rsp->completed = rsp->gpnum;
+ rcu_process_gp_end(rsp, rsp->rda[smp_processor_id()]);
+ rcu_start_gp(rsp, flags); /* releases root node's rnp->lock. */
+}
+
/*
* Similar to cpu_quiet(), for which it is a helper function. Allows
* a group of CPUs to be quieted at one go, though all the CPUs in the
return;
}
rnp->qsmask &= ~mask;
- if (rnp->qsmask != 0) {
+ if (rnp->qsmask != 0 || rcu_preempted_readers(rnp)) {
/* Other bits still set at this level, so done. */
spin_unlock_irqrestore(&rnp->lock, flags);
/*
* Get here if we are the last CPU to pass through a quiescent
- * state for this grace period. Clean up and let rcu_start_gp()
- * start up the next grace period if one is needed. Note that
- * we still hold rnp->lock, as required by rcu_start_gp(), which
- * will release it.
+ * state for this grace period. Invoke cpu_quiet_msk_finish()
+ * to clean up and start the next grace period if one is needed.
*/
- rsp->completed = rsp->gpnum;
- rcu_process_gp_end(rsp, rsp->rda[smp_processor_id()]);
- rcu_start_gp(rsp, flags); /* releases rnp->lock. */
+ cpu_quiet_msk_finish(rsp, flags); /* releases rnp->lock. */
}
/*
spin_lock(&rnp->lock); /* irqs already disabled. */
rnp->qsmaskinit &= ~mask;
if (rnp->qsmaskinit != 0) {
- spin_unlock(&rnp->lock); /* irqs already disabled. */
+ spin_unlock(&rnp->lock); /* irqs remain disabled. */
break;
}
+ rcu_preempt_offline_tasks(rsp, rnp);
mask = rnp->grpmask;
- spin_unlock(&rnp->lock); /* irqs already disabled. */
+ spin_unlock(&rnp->lock); /* irqs remain disabled. */
rnp = rnp->parent;
} while (rnp != NULL);
lastcomp = rsp->completed;
{
__rcu_offline_cpu(cpu, &rcu_sched_state);
__rcu_offline_cpu(cpu, &rcu_bh_state);
+ rcu_preempt_offline_cpu(cpu);
}
#else /* #ifdef CONFIG_HOTPLUG_CPU */
*/
void rcu_check_callbacks(int cpu, int user)
{
+ if (!rcu_pending(cpu))
+ return; /* if nothing for RCU to do. */
if (user ||
(idle_cpu(cpu) && rcu_scheduler_active &&
!in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) {
rcu_bh_qs(cpu);
}
+ rcu_preempt_check_callbacks(cpu);
raise_softirq(RCU_SOFTIRQ);
}
__rcu_process_callbacks(&rcu_sched_state,
&__get_cpu_var(rcu_sched_data));
__rcu_process_callbacks(&rcu_bh_state, &__get_cpu_var(rcu_bh_data));
+ rcu_preempt_process_callbacks();
/*
* Memory references from any later RCU read-side critical sections
}
EXPORT_SYMBOL_GPL(call_rcu_sched);
-/*
- * @@@ Queue an RCU callback for invocation after a grace period.
- * @@@ Placeholder pending rcutree_plugin.h.
- */
-void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
-{
- call_rcu_sched(head, func);
-}
-EXPORT_SYMBOL_GPL(call_rcu);
-
-
/*
* Queue an RCU for invocation after a quicker grace period.
*/
* by the current CPU, returning 1 if so. This function is part of the
* RCU implementation; it is -not- an exported member of the RCU API.
*/
-int rcu_pending(int cpu)
+static int rcu_pending(int cpu)
{
return __rcu_pending(&rcu_sched_state, &per_cpu(rcu_sched_data, cpu)) ||
- __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu));
+ __rcu_pending(&rcu_bh_state, &per_cpu(rcu_bh_data, cpu)) ||
+ rcu_preempt_pending(cpu);
}
/*
{
/* RCU callbacks either ready or pending? */
return per_cpu(rcu_sched_data, cpu).nxtlist ||
- per_cpu(rcu_bh_data, cpu).nxtlist;
+ per_cpu(rcu_bh_data, cpu).nxtlist ||
+ rcu_preempt_needs_cpu(cpu);
}
/*
* that this CPU cannot possibly have any RCU callbacks in flight yet.
*/
static void __cpuinit
-rcu_init_percpu_data(int cpu, struct rcu_state *rsp)
+rcu_init_percpu_data(int cpu, struct rcu_state *rsp, int preemptable)
{
unsigned long flags;
long lastcomp;
rdp->passed_quiesc = 0; /* We could be racing with new GP, */
rdp->qs_pending = 1; /* so set up to respond to current GP. */
rdp->beenonline = 1; /* We have now been online. */
+ rdp->preemptable = preemptable;
rdp->passed_quiesc_completed = lastcomp - 1;
rdp->blimit = blimit;
spin_unlock(&rnp->lock); /* irqs remain disabled. */
static void __cpuinit rcu_online_cpu(int cpu)
{
- rcu_init_percpu_data(cpu, &rcu_sched_state);
- rcu_init_percpu_data(cpu, &rcu_bh_state);
+ rcu_init_percpu_data(cpu, &rcu_sched_state, 0);
+ rcu_init_percpu_data(cpu, &rcu_bh_state, 0);
+ rcu_preempt_init_percpu_data(cpu);
}
/*
- * Handle CPU online/offline notifcation events.
+ * Handle CPU online/offline notification events.
*/
int __cpuinit rcu_cpu_notify(struct notifier_block *self,
unsigned long action, void *hcpu)
rnp = rsp->level[i];
for (j = 0; j < rsp->levelcnt[i]; j++, rnp++) {
spin_lock_init(&rnp->lock);
+ rnp->gpnum = 0;
rnp->qsmask = 0;
rnp->qsmaskinit = 0;
rnp->grplo = j * cpustride;
j / rsp->levelspread[i - 1];
}
rnp->level = i;
+ INIT_LIST_HEAD(&rnp->blocked_tasks[0]);
+ INIT_LIST_HEAD(&rnp->blocked_tasks[1]);
}
}
}
/*
- * Helper macro for __rcu_init(). To be used nowhere else!
- * Assigns leaf node pointers into each CPU's rcu_data structure.
+ * Helper macro for __rcu_init() and __rcu_init_preempt(). To be used
+ * nowhere else! Assigns leaf node pointers into each CPU's rcu_data
+ * structure.
*/
#define RCU_INIT_FLAVOR(rsp, rcu_data) \
do { \
} \
} while (0)
+#ifdef CONFIG_TREE_PREEMPT_RCU
+
+void __init __rcu_init_preempt(void)
+{
+ int i; /* All used by RCU_INIT_FLAVOR(). */
+ int j;
+ struct rcu_node *rnp;
+
+ RCU_INIT_FLAVOR(&rcu_preempt_state, rcu_preempt_data);
+}
+
+#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */
+
+void __init __rcu_init_preempt(void)
+{
+}
+
+#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
+
void __init __rcu_init(void)
{
- int i; /* All used by RCU_DATA_PTR_INIT(). */
+ int i; /* All used by RCU_INIT_FLAVOR(). */
int j;
struct rcu_node *rnp;
- printk(KERN_INFO "Hierarchical RCU implementation.\n");
+ rcu_bootup_announce();
#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
printk(KERN_INFO "RCU-based detection of stalled CPUs is enabled.\n");
#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
RCU_INIT_FLAVOR(&rcu_sched_state, rcu_sched_data);
RCU_INIT_FLAVOR(&rcu_bh_state, rcu_bh_data);
+ __rcu_init_preempt();
open_softirq(RCU_SOFTIRQ, rcu_process_callbacks);
}