diff options
Diffstat (limited to 'kernel/rcutree.c')
| -rw-r--r-- | kernel/rcutree.c | 1010 |
1 files changed, 680 insertions, 330 deletions
diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 7717b95c202..53ae9598f79 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -25,7 +25,7 @@ * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. * * For detailed explanation of Read-Copy Update mechanism see - - * Documentation/RCU + * Documentation/RCU */ #include <linux/types.h> #include <linux/kernel.h> @@ -35,6 +35,7 @@ #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> @@ -45,57 +46,78 @@ #include <linux/cpu.h> #include <linux/mutex.h> #include <linux/time.h> +#include <linux/kernel_stat.h> -#ifdef CONFIG_DEBUG_LOCK_ALLOC -static struct lock_class_key rcu_lock_key; -struct lockdep_map rcu_lock_map = - STATIC_LOCKDEP_MAP_INIT("rcu_read_lock", &rcu_lock_key); -EXPORT_SYMBOL_GPL(rcu_lock_map); -#endif +#include "rcutree.h" /* Data structures. */ +static struct lock_class_key rcu_node_class[NUM_RCU_LVLS]; + #define RCU_STATE_INITIALIZER(name) { \ .level = { &name.node[0] }, \ .levelcnt = { \ NUM_RCU_LVL_0, /* root of hierarchy. */ \ NUM_RCU_LVL_1, \ NUM_RCU_LVL_2, \ - NUM_RCU_LVL_3, /* == MAX_RCU_LVLS */ \ + NUM_RCU_LVL_3, \ + NUM_RCU_LVL_4, /* == MAX_RCU_LVLS */ \ }, \ - .signaled = RCU_SIGNAL_INIT, \ + .signaled = RCU_GP_IDLE, \ .gpnum = -300, \ .completed = -300, \ .onofflock = __SPIN_LOCK_UNLOCKED(&name.onofflock), \ + .orphan_cbs_list = NULL, \ + .orphan_cbs_tail = &name.orphan_cbs_list, \ + .orphan_qlen = 0, \ .fqslock = __SPIN_LOCK_UNLOCKED(&name.fqslock), \ .n_force_qs = 0, \ .n_force_qs_ngp = 0, \ } -struct rcu_state rcu_state = RCU_STATE_INITIALIZER(rcu_state); -DEFINE_PER_CPU(struct rcu_data, rcu_data); +struct rcu_state rcu_sched_state = RCU_STATE_INITIALIZER(rcu_sched_state); +DEFINE_PER_CPU(struct rcu_data, rcu_sched_data); struct rcu_state rcu_bh_state = RCU_STATE_INITIALIZER(rcu_bh_state); DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); +static int rcu_scheduler_active __read_mostly; + + +/* + * 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 + * structure's ->lock, but of course results can be subject to change. + */ +static int rcu_gp_in_progress(struct rcu_state *rsp) +{ + return ACCESS_ONCE(rsp->completed) != ACCESS_ONCE(rsp->gpnum); +} + /* - * Increment the quiescent state counter. - * The counter is a bit degenerated: We do not need to know + * Note a quiescent state. Because we do not need to know * how many quiescent states passed, just if there was at least - * one since the start of the grace period. Thus just a flag. + * one since the start of the grace period, this just sets a flag. */ -void rcu_qsctr_inc(int cpu) +void rcu_sched_qs(int cpu) { - struct rcu_data *rdp = &per_cpu(rcu_data, cpu); + struct rcu_data *rdp; + + rdp = &per_cpu(rcu_sched_data, cpu); + rdp->passed_quiesc_completed = rdp->gpnum - 1; + barrier(); rdp->passed_quiesc = 1; - rdp->passed_quiesc_completed = rdp->completed; + rcu_preempt_note_context_switch(cpu); } -void rcu_bh_qsctr_inc(int cpu) +void rcu_bh_qs(int cpu) { - struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu); + struct rcu_data *rdp; + + rdp = &per_cpu(rcu_bh_data, cpu); + rdp->passed_quiesc_completed = rdp->gpnum - 1; + barrier(); rdp->passed_quiesc = 1; - rdp->passed_quiesc_completed = rdp->completed; } #ifdef CONFIG_NO_HZ @@ -109,16 +131,21 @@ static int blimit = 10; /* Maximum callbacks per softirq. */ static int qhimark = 10000; /* If this many pending, ignore blimit. */ static int qlowmark = 100; /* Once only this many pending, use blimit. */ +module_param(blimit, int, 0); +module_param(qhimark, int, 0); +module_param(qlowmark, int, 0); + static void force_quiescent_state(struct rcu_state *rsp, int relaxed); +static int rcu_pending(int cpu); /* - * Return the number of RCU batches processed thus far for debug & stats. + * Return the number of RCU-sched batches processed thus far for debug & stats. */ -long rcu_batches_completed(void) +long rcu_batches_completed_sched(void) { - return rcu_state.completed; + return rcu_sched_state.completed; } -EXPORT_SYMBOL_GPL(rcu_batches_completed); +EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); /* * Return the number of RCU BH batches processed thus far for debug & stats. @@ -144,9 +171,7 @@ cpu_has_callbacks_ready_to_invoke(struct rcu_data *rdp) static int cpu_needs_another_gp(struct rcu_state *rsp, struct rcu_data *rdp) { - /* ACCESS_ONCE() because we are accessing outside of lock. */ - return *rdp->nxttail[RCU_DONE_TAIL] && - ACCESS_ONCE(rsp->completed) == ACCESS_ONCE(rsp->gpnum); + return *rdp->nxttail[RCU_DONE_TAIL] && !rcu_gp_in_progress(rsp); } /* @@ -181,6 +206,10 @@ static int rcu_implicit_offline_qs(struct rcu_data *rdp) 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); @@ -193,7 +222,6 @@ static int rcu_implicit_offline_qs(struct rcu_data *rdp) #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 @@ -213,7 +241,7 @@ void rcu_enter_nohz(void) 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); } @@ -232,7 +260,7 @@ void rcu_exit_nohz(void) 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 */ } @@ -251,7 +279,7 @@ void rcu_nmi_enter(void) 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 */ } @@ -270,7 +298,7 @@ void rcu_nmi_exit(void) 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); } /** @@ -286,7 +314,7 @@ void rcu_irq_enter(void) 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 */ } @@ -305,39 +333,20 @@ void rcu_irq_exit(void) 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_data).nxtlist || + if (__get_cpu_var(rcu_sched_data).nxtlist || __get_cpu_var(rcu_bh_data).nxtlist) set_need_resched(); } -/* - * Record the specified "completed" value, which is later used to validate - * dynticks counter manipulations. Specify "rsp->completed - 1" to - * unconditionally invalidate any future dynticks manipulations (which is - * useful at the beginning of a grace period). - */ -static void dyntick_record_completed(struct rcu_state *rsp, long comp) -{ - rsp->dynticks_completed = comp; -} - #ifdef CONFIG_SMP /* - * Recall the previously recorded value of the completion for dynticks. - */ -static long dyntick_recall_completed(struct rcu_state *rsp) -{ - return rsp->dynticks_completed; -} - -/* * Snapshot the specified CPU's dynticks counter so that we can later * credit them with an implicit quiescent state. Return 1 if this CPU - * is already in a quiescent state courtesy of dynticks idle mode. + * is in dynticks idle mode, which is an extended quiescent state. */ static int dyntick_save_progress_counter(struct rcu_data *rdp) { @@ -397,24 +406,8 @@ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp) #else /* #ifdef CONFIG_NO_HZ */ -static void dyntick_record_completed(struct rcu_state *rsp, long comp) -{ -} - #ifdef CONFIG_SMP -/* - * If there are no dynticks, then the only way that a CPU can passively - * be in a quiescent state is to be offline. Unlike dynticks idle, which - * is a point in time during the prior (already finished) grace period, - * an offline CPU is always in a quiescent state, and thus can be - * unconditionally applied. So just return the current value of completed. - */ -static long dyntick_recall_completed(struct rcu_state *rsp) -{ - return rsp->completed; -} - static int dyntick_save_progress_counter(struct rcu_data *rdp) { return 0; @@ -443,32 +436,39 @@ static void print_other_cpu_stall(struct rcu_state *rsp) long delta; unsigned long flags; struct rcu_node *rnp = rcu_get_root(rsp); - struct rcu_node *rnp_cur = rsp->level[NUM_RCU_LVLS - 1]; - struct rcu_node *rnp_end = &rsp->node[NUM_RCU_NODES]; /* Only let one CPU complain about others per time interval. */ spin_lock_irqsave(&rnp->lock, flags); delta = jiffies - rsp->jiffies_stall; - if (delta < RCU_STALL_RAT_DELAY || rsp->gpnum == rsp->completed) { + if (delta < RCU_STALL_RAT_DELAY || !rcu_gp_in_progress(rsp)) { spin_unlock_irqrestore(&rnp->lock, flags); return; } rsp->jiffies_stall = jiffies + RCU_SECONDS_TILL_STALL_RECHECK; + + /* + * Now rat on any tasks that got kicked up to the root rcu_node + * due to CPU offlining. + */ + rcu_print_task_stall(rnp); spin_unlock_irqrestore(&rnp->lock, flags); /* OK, time to rat on our buddy... */ printk(KERN_ERR "INFO: RCU detected CPU stalls:"); - for (; rnp_cur < rnp_end; rnp_cur++) { - if (rnp_cur->qsmask == 0) + rcu_for_each_leaf_node(rsp, rnp) { + rcu_print_task_stall(rnp); + if (rnp->qsmask == 0) continue; - for (cpu = 0; cpu <= rnp_cur->grphi - rnp_cur->grplo; cpu++) - if (rnp_cur->qsmask & (1UL << cpu)) - printk(" %d", rnp_cur->grplo + cpu); + for (cpu = 0; cpu <= rnp->grphi - rnp->grplo; cpu++) + if (rnp->qsmask & (1UL << cpu)) + printk(" %d", rnp->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. */ } @@ -479,12 +479,14 @@ static void print_cpu_stall(struct rcu_state *rsp) 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. */ } @@ -500,8 +502,7 @@ static void check_cpu_stall(struct rcu_state *rsp, struct rcu_data *rdp) /* We haven't checked in, so go dump stack. */ print_cpu_stall(rsp); - } else if (rsp->gpnum != rsp->completed && - delta >= RCU_STALL_RAT_DELAY) { + } else if (rcu_gp_in_progress(rsp) && delta >= RCU_STALL_RAT_DELAY) { /* They had two time units to dump stack, so complain. */ print_other_cpu_stall(rsp); @@ -523,13 +524,33 @@ static void check_cpu_stall(struct rcu_state *rsp, struct rcu_data *rdp) /* * Update CPU-local rcu_data state to record the newly noticed grace period. * This is used both when we started the grace period and when we notice - * that someone else started the grace period. + * that someone else started the grace period. The caller must hold the + * ->lock of the leaf rcu_node structure corresponding to the current CPU, + * and must have irqs disabled. */ +static void __note_new_gpnum(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) +{ + if (rdp->gpnum != rnp->gpnum) { + rdp->qs_pending = 1; + rdp->passed_quiesc = 0; + rdp->gpnum = rnp->gpnum; + } +} + static void note_new_gpnum(struct rcu_state *rsp, struct rcu_data *rdp) { - rdp->qs_pending = 1; - rdp->passed_quiesc = 0; - rdp->gpnum = rsp->gpnum; + unsigned long flags; + struct rcu_node *rnp; + + local_irq_save(flags); + rnp = rdp->mynode; + if (rdp->gpnum == ACCESS_ONCE(rnp->gpnum) || /* outside lock. */ + !spin_trylock(&rnp->lock)) { /* irqs already off, retry later. */ + local_irq_restore(flags); + return; + } + __note_new_gpnum(rsp, rnp, rdp); + spin_unlock_irqrestore(&rnp->lock, flags); } /* @@ -553,6 +574,79 @@ check_for_new_grace_period(struct rcu_state *rsp, struct rcu_data *rdp) } /* + * Advance this CPU's callbacks, but only if the current grace period + * has ended. This may be called only from the CPU to whom the rdp + * belongs. In addition, the corresponding leaf rcu_node structure's + * ->lock must be held by the caller, with irqs disabled. + */ +static void +__rcu_process_gp_end(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) +{ + /* Did another grace period end? */ + if (rdp->completed != rnp->completed) { + + /* Advance callbacks. No harm if list empty. */ + rdp->nxttail[RCU_DONE_TAIL] = rdp->nxttail[RCU_WAIT_TAIL]; + rdp->nxttail[RCU_WAIT_TAIL] = rdp->nxttail[RCU_NEXT_READY_TAIL]; + rdp->nxttail[RCU_NEXT_READY_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; + + /* Remember that we saw this grace-period completion. */ + rdp->completed = rnp->completed; + } +} + +/* + * Advance this CPU's callbacks, but only if the current grace period + * has ended. This may be called only from the CPU to whom the rdp + * belongs. + */ +static void +rcu_process_gp_end(struct rcu_state *rsp, struct rcu_data *rdp) +{ + unsigned long flags; + struct rcu_node *rnp; + + local_irq_save(flags); + rnp = rdp->mynode; + if (rdp->completed == ACCESS_ONCE(rnp->completed) || /* outside lock. */ + !spin_trylock(&rnp->lock)) { /* irqs already off, retry later. */ + local_irq_restore(flags); + return; + } + __rcu_process_gp_end(rsp, rnp, rdp); + spin_unlock_irqrestore(&rnp->lock, flags); +} + +/* + * Do per-CPU grace-period initialization for running CPU. The caller + * must hold the lock of the leaf rcu_node structure corresponding to + * this CPU. + */ +static void +rcu_start_gp_per_cpu(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) +{ + /* Prior grace period ended, so advance callbacks for current CPU. */ + __rcu_process_gp_end(rsp, rnp, rdp); + + /* + * Because this CPU just now started the new grace period, we know + * that all of its callbacks will be covered by this upcoming grace + * period, even the ones that were registered arbitrarily recently. + * Therefore, advance all outstanding callbacks to RCU_WAIT_TAIL. + * + * Other CPUs cannot be sure exactly when the grace period started. + * Therefore, their recently registered callbacks must pass through + * an additional RCU_NEXT_READY stage, so that they will be handled + * by the next RCU grace period. + */ + rdp->nxttail[RCU_NEXT_READY_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; + rdp->nxttail[RCU_WAIT_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; + + /* Set state so that this CPU will detect the next quiescent state. */ + __note_new_gpnum(rsp, rnp, rdp); +} + +/* * Start a new RCU grace period if warranted, re-initializing the hierarchy * in preparation for detecting the next grace period. The caller must hold * the root node's ->lock, which is released before return. Hard irqs must @@ -564,34 +658,43 @@ rcu_start_gp(struct rcu_state *rsp, unsigned long flags) { struct rcu_data *rdp = rsp->rda[smp_processor_id()]; struct rcu_node *rnp = rcu_get_root(rsp); - struct rcu_node *rnp_cur; - struct rcu_node *rnp_end; if (!cpu_needs_another_gp(rsp, rdp)) { - spin_unlock_irqrestore(&rnp->lock, flags); + if (rnp->completed == rsp->completed) { + spin_unlock_irqrestore(&rnp->lock, flags); + return; + } + spin_unlock(&rnp->lock); /* irqs remain disabled. */ + + /* + * Propagate new ->completed value to rcu_node structures + * so that other CPUs don't have to wait until the start + * of the next grace period to process their callbacks. + */ + rcu_for_each_node_breadth_first(rsp, rnp) { + spin_lock(&rnp->lock); /* irqs already disabled. */ + rnp->completed = rsp->completed; + spin_unlock(&rnp->lock); /* irqs remain disabled. */ + } + local_irq_restore(flags); return; } /* Advance to a new grace period and initialize state. */ rsp->gpnum++; + WARN_ON_ONCE(rsp->signaled == RCU_GP_INIT); rsp->signaled = RCU_GP_INIT; /* Hold off force_quiescent_state. */ rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS; record_gp_stall_check_time(rsp); - dyntick_record_completed(rsp, rsp->completed - 1); - note_new_gpnum(rsp, rdp); - - /* - * Because we are first, we know that all our callbacks will - * be covered by this upcoming grace period, even the ones - * that were registered arbitrarily recently. - */ - rdp->nxttail[RCU_NEXT_READY_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; - rdp->nxttail[RCU_WAIT_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; /* Special-case the common single-level case. */ if (NUM_RCU_NODES == 1) { + rcu_preempt_check_blocked_tasks(rnp); rnp->qsmask = rnp->qsmaskinit; + rnp->gpnum = rsp->gpnum; + rnp->completed = rsp->completed; rsp->signaled = RCU_SIGNAL_INIT; /* force_quiescent_state OK. */ + rcu_start_gp_per_cpu(rsp, rnp, rdp); spin_unlock_irqrestore(&rnp->lock, flags); return; } @@ -603,88 +706,71 @@ rcu_start_gp(struct rcu_state *rsp, unsigned long flags) spin_lock(&rsp->onofflock); /* irqs already disabled. */ /* - * Set the quiescent-state-needed bits in all the non-leaf RCU - * nodes for all currently online CPUs. This operation relies - * on the layout of the hierarchy within the rsp->node[] array. - * Note that other CPUs will access only the leaves of the - * hierarchy, which still indicate that no grace period is in - * progress. In addition, we have excluded CPU-hotplug operations. - * - * We therefore do not need to hold any locks. Any required - * memory barriers will be supplied by the locks guarding the - * leaf rcu_nodes in the hierarchy. - */ - - rnp_end = rsp->level[NUM_RCU_LVLS - 1]; - for (rnp_cur = &rsp->node[0]; rnp_cur < rnp_end; rnp_cur++) - rnp_cur->qsmask = rnp_cur->qsmaskinit; - - /* - * Now set up the leaf nodes. Here we must be careful. First, - * we need to hold the lock in order to exclude other CPUs, which - * might be contending for the leaf nodes' locks. Second, as - * soon as we initialize a given leaf node, its CPUs might run - * up the rest of the hierarchy. We must therefore acquire locks - * for each node that we touch during this stage. (But we still - * are excluding CPU-hotplug operations.) + * Set the quiescent-state-needed bits in all the rcu_node + * structures for all currently online CPUs in breadth-first + * order, starting from the root rcu_node structure. This + * operation relies on the layout of the hierarchy within the + * rsp->node[] array. Note that other CPUs will access only + * the leaves of the hierarchy, which still indicate that no + * grace period is in progress, at least until the corresponding + * leaf node has been initialized. In addition, we have excluded + * CPU-hotplug operations. * * Note that the grace period cannot complete until we finish * the initialization process, as there will be at least one * qsmask bit set in the root node until that time, namely the - * one corresponding to this CPU. + * one corresponding to this CPU, due to the fact that we have + * irqs disabled. */ - rnp_end = &rsp->node[NUM_RCU_NODES]; - rnp_cur = rsp->level[NUM_RCU_LVLS - 1]; - for (; rnp_cur < rnp_end; rnp_cur++) { - spin_lock(&rnp_cur->lock); /* irqs already disabled. */ - rnp_cur->qsmask = rnp_cur->qsmaskinit; - spin_unlock(&rnp_cur->lock); /* irqs already disabled. */ + rcu_for_each_node_breadth_first(rsp, rnp) { + spin_lock(&rnp->lock); /* irqs already disabled. */ + rcu_preempt_check_blocked_tasks(rnp); + rnp->qsmask = rnp->qsmaskinit; + rnp->gpnum = rsp->gpnum; + rnp->completed = rsp->completed; + if (rnp == rdp->mynode) + rcu_start_gp_per_cpu(rsp, rnp, rdp); + spin_unlock(&rnp->lock); /* irqs remain disabled. */ } + rnp = rcu_get_root(rsp); + spin_lock(&rnp->lock); /* irqs already disabled. */ rsp->signaled = RCU_SIGNAL_INIT; /* force_quiescent_state now OK. */ + spin_unlock(&rnp->lock); /* irqs remain disabled. */ spin_unlock_irqrestore(&rsp->onofflock, flags); } /* - * Advance this CPU's callbacks, but only if the current grace period - * has ended. This may be called only from the CPU to whom the rdp - * belongs. + * Report a full set of quiescent states to the specified rcu_state + * data structure. This involves cleaning up after the prior grace + * period and letting 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 -rcu_process_gp_end(struct rcu_state *rsp, struct rcu_data *rdp) +static void rcu_report_qs_rsp(struct rcu_state *rsp, unsigned long flags) + __releases(rcu_get_root(rsp)->lock) { - long completed_snap; - unsigned long flags; - - local_irq_save(flags); - completed_snap = ACCESS_ONCE(rsp->completed); /* outside of lock. */ - - /* Did another grace period end? */ - if (rdp->completed != completed_snap) { - - /* Advance callbacks. No harm if list empty. */ - rdp->nxttail[RCU_DONE_TAIL] = rdp->nxttail[RCU_WAIT_TAIL]; - rdp->nxttail[RCU_WAIT_TAIL] = rdp->nxttail[RCU_NEXT_READY_TAIL]; - rdp->nxttail[RCU_NEXT_READY_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; - - /* Remember that we saw this grace-period completion. */ - rdp->completed = completed_snap; - } - local_irq_restore(flags); + WARN_ON_ONCE(!rcu_gp_in_progress(rsp)); + rsp->completed = rsp->gpnum; + rsp->signaled = RCU_GP_IDLE; + 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 - * group must be represented by the same leaf rcu_node structure. - * That structure's lock must be held upon entry, and it is released - * before return. + * Similar to rcu_report_qs_rdp(), for which it is a helper function. + * Allows quiescent states for a group of CPUs to be reported at one go + * to the specified rcu_node structure, though all the CPUs in the group + * must be represented by the same rcu_node structure (which need not be + * a leaf rcu_node structure, though it often will be). That structure's + * lock must be held upon entry, and it is released before return. */ static void -cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, - unsigned long flags) +rcu_report_qs_rnp(unsigned long mask, struct rcu_state *rsp, + struct rcu_node *rnp, unsigned long flags) __releases(rnp->lock) { + struct rcu_node *rnp_c; + /* Walk up the rcu_node hierarchy. */ for (;;) { if (!(rnp->qsmask & mask)) { @@ -694,7 +780,7 @@ cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, 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); @@ -708,31 +794,31 @@ cpu_quiet_msk(unsigned long mask, struct rcu_state *rsp, struct rcu_node *rnp, break; } spin_unlock_irqrestore(&rnp->lock, flags); + rnp_c = rnp; rnp = rnp->parent; spin_lock_irqsave(&rnp->lock, flags); + WARN_ON_ONCE(rnp_c->qsmask); } /* * 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 rcu_report_qs_rsp() + * 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. */ + rcu_report_qs_rsp(rsp, flags); /* releases rnp->lock. */ } /* - * Record a quiescent state for the specified CPU, which must either be - * the current CPU or an offline CPU. The lastcomp argument is used to - * make sure we are still in the grace period of interest. We don't want - * to end the current grace period based on quiescent states detected in - * an earlier grace period! + * Record a quiescent state for the specified CPU to that CPU's rcu_data + * structure. This must be either called from the specified CPU, or + * called when the specified CPU is known to be offline (and when it is + * also known that no other CPU is concurrently trying to help the offline + * CPU). The lastcomp argument is used to make sure we are still in the + * grace period of interest. We don't want to end the current grace period + * based on quiescent states detected in an earlier grace period! */ static void -cpu_quiet(int cpu, struct rcu_state *rsp, struct rcu_data *rdp, long lastcomp) +rcu_report_qs_rdp(int cpu, struct rcu_state *rsp, struct rcu_data *rdp, long lastcomp) { unsigned long flags; unsigned long mask; @@ -740,15 +826,15 @@ cpu_quiet(int cpu, struct rcu_state *rsp, struct rcu_data *rdp, long lastcomp) rnp = rdp->mynode; spin_lock_irqsave(&rnp->lock, flags); - if (lastcomp != ACCESS_ONCE(rsp->completed)) { + if (lastcomp != rnp->completed) { /* * Someone beat us to it for this grace period, so leave. * The race with GP start is resolved by the fact that we * hold the leaf rcu_node lock, so that the per-CPU bits * cannot yet be initialized -- so we would simply find our - * CPU's bit already cleared in cpu_quiet_msk() if this race - * occurred. + * CPU's bit already cleared in rcu_report_qs_rnp() if this + * race occurred. */ rdp->passed_quiesc = 0; /* try again later! */ spin_unlock_irqrestore(&rnp->lock, flags); @@ -764,10 +850,9 @@ cpu_quiet(int cpu, struct rcu_state *rsp, struct rcu_data *rdp, long lastcomp) * This GP can't end until cpu checks in, so all of our * callbacks can be processed during the next GP. */ - rdp = rsp->rda[smp_processor_id()]; rdp->nxttail[RCU_NEXT_READY_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; - cpu_quiet_msk(mask, rsp, rnp, flags); /* releases rnp->lock */ + rcu_report_qs_rnp(mask, rsp, rnp, flags); /* rlses rnp->lock */ } } @@ -798,74 +883,113 @@ rcu_check_quiescent_state(struct rcu_state *rsp, struct rcu_data *rdp) if (!rdp->passed_quiesc) return; - /* Tell RCU we are done (but cpu_quiet() will be the judge of that). */ - cpu_quiet(rdp->cpu, rsp, rdp, rdp->passed_quiesc_completed); + /* + * Tell RCU we are done (but rcu_report_qs_rdp() will be the + * judge of that). + */ + rcu_report_qs_rdp(rdp->cpu, rsp, rdp, rdp->passed_quiesc_completed); } #ifdef CONFIG_HOTPLUG_CPU /* + * Move a dying CPU's RCU callbacks to the ->orphan_cbs_list for the + * specified flavor of RCU. The callbacks will be adopted by the next + * _rcu_barrier() invocation or by the CPU_DEAD notifier, whichever + * comes first. Because this is invoked from the CPU_DYING notifier, + * irqs are already disabled. + */ +static void rcu_send_cbs_to_orphanage(struct rcu_state *rsp) +{ + int i; + struct rcu_data *rdp = rsp->rda[smp_processor_id()]; + + if (rdp->nxtlist == NULL) + return; /* irqs disabled, so comparison is stable. */ + spin_lock(&rsp->onofflock); /* irqs already disabled. */ + *rsp->orphan_cbs_tail = rdp->nxtlist; + rsp->orphan_cbs_tail = rdp->nxttail[RCU_NEXT_TAIL]; + rdp->nxtlist = NULL; + for (i = 0; i < RCU_NEXT_SIZE; i++) + rdp->nxttail[i] = &rdp->nxtlist; + rsp->orphan_qlen += rdp->qlen; + rdp->qlen = 0; + spin_unlock(&rsp->onofflock); /* irqs remain disabled. */ +} + +/* + * Adopt previously orphaned RCU callbacks. + */ +static void rcu_adopt_orphan_cbs(struct rcu_state *rsp) +{ + unsigned long flags; + struct rcu_data *rdp; + + spin_lock_irqsave(&rsp->onofflock, flags); + rdp = rsp->rda[smp_processor_id()]; + if (rsp->orphan_cbs_list == NULL) { + spin_unlock_irqrestore(&rsp->onofflock, flags); + return; + } + *rdp->nxttail[RCU_NEXT_TAIL] = rsp->orphan_cbs_list; + rdp->nxttail[RCU_NEXT_TAIL] = rsp->orphan_cbs_tail; + rdp->qlen += rsp->orphan_qlen; + rsp->orphan_cbs_list = NULL; + rsp->orphan_cbs_tail = &rsp->orphan_cbs_list; + rsp->orphan_qlen = 0; + spin_unlock_irqrestore(&rsp->onofflock, flags); +} + +/* * Remove the outgoing CPU from the bitmasks in the rcu_node hierarchy * and move all callbacks from the outgoing CPU to the current one. */ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) { - int i; unsigned long flags; - long lastcomp; unsigned long mask; + int need_report = 0; struct rcu_data *rdp = rsp->rda[cpu]; - struct rcu_data *rdp_me; struct rcu_node *rnp; /* Exclude any attempts to start a new grace period. */ spin_lock_irqsave(&rsp->onofflock, flags); /* Remove the outgoing CPU from the masks in the rcu_node hierarchy. */ - rnp = rdp->mynode; + rnp = rdp->mynode; /* this is the outgoing CPU's rnp. */ mask = rdp->grpmask; /* rnp->grplo is constant. */ do { spin_lock(&rnp->lock); /* irqs already disabled. */ rnp->qsmaskinit &= ~mask; if (rnp->qsmaskinit != 0) { - spin_unlock(&rnp->lock); /* irqs already disabled. */ + if (rnp != rdp->mynode) + spin_unlock(&rnp->lock); /* irqs remain disabled. */ break; } + if (rnp == rdp->mynode) + need_report = rcu_preempt_offline_tasks(rsp, rnp, rdp); + else + spin_unlock(&rnp->lock); /* irqs remain disabled. */ mask = rnp->grpmask; - spin_unlock(&rnp->lock); /* irqs already disabled. */ rnp = rnp->parent; } while (rnp != NULL); - lastcomp = rsp->completed; - - spin_unlock(&rsp->onofflock); /* irqs remain disabled. */ - - /* Being offline is a quiescent state, so go record it. */ - cpu_quiet(cpu, rsp, rdp, lastcomp); /* - * Move callbacks from the outgoing CPU to the running CPU. - * Note that the outgoing CPU is now quiscent, so it is now - * (uncharacteristically) safe to access it rcu_data structure. - * Note also that we must carefully retain the order of the - * outgoing CPU's callbacks in order for rcu_barrier() to work - * correctly. Finally, note that we start all the callbacks - * afresh, even those that have passed through a grace period - * and are therefore ready to invoke. The theory is that hotplug - * events are rare, and that if they are frequent enough to - * indefinitely delay callbacks, you have far worse things to - * be worrying about. + * We still hold the leaf rcu_node structure lock here, and + * irqs are still disabled. The reason for this subterfuge is + * because invoking rcu_report_unblock_qs_rnp() with ->onofflock + * held leads to deadlock. */ - rdp_me = rsp->rda[smp_processor_id()]; - if (rdp->nxtlist != NULL) { - *rdp_me->nxttail[RCU_NEXT_TAIL] = rdp->nxtlist; - rdp_me->nxttail[RCU_NEXT_TAIL] = rdp->nxttail[RCU_NEXT_TAIL]; - rdp->nxtlist = NULL; - for (i = 0; i < RCU_NEXT_SIZE; i++) - rdp->nxttail[i] = &rdp->nxtlist; - rdp_me->qlen += rdp->qlen; - rdp->qlen = 0; - } - local_irq_restore(flags); + spin_unlock(&rsp->onofflock); /* irqs remain disabled. */ + rnp = rdp->mynode; + if (need_report & RCU_OFL_TASKS_NORM_GP) + rcu_report_unblock_qs_rnp(rnp, flags); + else + spin_unlock_irqrestore(&rnp->lock, flags); + if (need_report & RCU_OFL_TASKS_EXP_GP) + rcu_report_exp_rnp(rsp, rnp); + + rcu_adopt_orphan_cbs(rsp); } /* @@ -876,12 +1000,21 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) */ static void rcu_offline_cpu(int cpu) { - __rcu_offline_cpu(cpu, &rcu_state); + __rcu_offline_cpu(cpu, &rcu_sched_state); __rcu_offline_cpu(cpu, &rcu_bh_state); + rcu_preempt_offline_cpu(cpu); } #else /* #ifdef CONFIG_HOTPLUG_CPU */ +static void rcu_send_cbs_to_orphanage(struct rcu_state *rsp) +{ +} + +static void rcu_adopt_orphan_cbs(struct rcu_state *rsp) +{ +} + static void rcu_offline_cpu(int cpu) { } @@ -892,7 +1025,7 @@ static void rcu_offline_cpu(int cpu) * Invoke any RCU callbacks that have made it to the end of their grace * period. Thottle as specified by rdp->blimit. */ -static void rcu_do_batch(struct rcu_data *rdp) +static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp) { unsigned long flags; struct rcu_head *next, *list, **tail; @@ -945,6 +1078,13 @@ static void rcu_do_batch(struct rcu_data *rdp) if (rdp->blimit == LONG_MAX && rdp->qlen <= qlowmark) rdp->blimit = blimit; + /* Reset ->qlen_last_fqs_check trigger if enough CBs have drained. */ + if (rdp->qlen == 0 && rdp->qlen_last_fqs_check != 0) { + rdp->qlen_last_fqs_check = 0; + rdp->n_force_qs_snap = rsp->n_force_qs; + } else if (rdp->qlen < rdp->qlen_last_fqs_check - qhimark) + rdp->qlen_last_fqs_check = rdp->qlen; + local_irq_restore(flags); /* Re-raise the RCU softirq if there are callbacks remaining. */ @@ -963,6 +1103,8 @@ static void rcu_do_batch(struct rcu_data *rdp) */ 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))) { @@ -971,17 +1113,16 @@ void rcu_check_callbacks(int cpu, int user) * Get here if this CPU took its interrupt from user * mode or from the idle loop, and if this is not a * nested interrupt. In this case, the CPU is in - * a quiescent state, so count it. + * a quiescent state, so note it. * * No memory barrier is required here because both - * rcu_qsctr_inc() and rcu_bh_qsctr_inc() reference - * only CPU-local variables that other CPUs neither - * access nor modify, at least not while the corresponding - * CPU is online. + * rcu_sched_qs() and rcu_bh_qs() reference only CPU-local + * variables that other CPUs neither access nor modify, + * at least not while the corresponding CPU is online. */ - rcu_qsctr_inc(cpu); - rcu_bh_qsctr_inc(cpu); + rcu_sched_qs(cpu); + rcu_bh_qs(cpu); } else if (!in_softirq()) { @@ -989,11 +1130,12 @@ void rcu_check_callbacks(int cpu, int user) * Get here if this CPU did not take its interrupt from * softirq, in other words, if it is not interrupting * a rcu_bh read-side critical section. This is an _bh - * critical section, so count it. + * critical section, so note it. */ - rcu_bh_qsctr_inc(cpu); + rcu_bh_qs(cpu); } + rcu_preempt_check_callbacks(cpu); raise_softirq(RCU_SOFTIRQ); } @@ -1012,33 +1154,32 @@ static int rcu_process_dyntick(struct rcu_state *rsp, long lastcomp, int cpu; unsigned long flags; unsigned long mask; - struct rcu_node *rnp_cur = rsp->level[NUM_RCU_LVLS - 1]; - struct rcu_node *rnp_end = &rsp->node[NUM_RCU_NODES]; + struct rcu_node *rnp; - for (; rnp_cur < rnp_end; rnp_cur++) { + rcu_for_each_leaf_node(rsp, rnp) { mask = 0; - spin_lock_irqsave(&rnp_cur->lock, flags); - if (rsp->completed != lastcomp) { - spin_unlock_irqrestore(&rnp_cur->lock, flags); + spin_lock_irqsave(&rnp->lock, flags); + if (rnp->completed != lastcomp) { + spin_unlock_irqrestore(&rnp->lock, flags); return 1; } - if (rnp_cur->qsmask == 0) { - spin_unlock_irqrestore(&rnp_cur->lock, flags); + if (rnp->qsmask == 0) { + spin_unlock_irqrestore(&rnp->lock, flags); continu |