/* * Read-Copy Update mechanism for mutual exclusion * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, you can access it online at * http://www.gnu.org/licenses/gpl-2.0.html. * * Copyright IBM Corporation, 2008 * * Authors: Dipankar Sarma * Manfred Spraul * Paul E. McKenney Hierarchical version * * Based on the original work by Paul McKenney * and inputs from Rusty Russell, Andrea Arcangeli and Andi Kleen. * * For detailed explanation of Read-Copy Update mechanism see - * Documentation/RCU */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "tree.h" #include "rcu.h" MODULE_ALIAS("rcutree"); #ifdef MODULE_PARAM_PREFIX #undef MODULE_PARAM_PREFIX #endif #define MODULE_PARAM_PREFIX "rcutree." /* Data structures. */ static struct lock_class_key rcu_node_class[RCU_NUM_LVLS]; static struct lock_class_key rcu_fqs_class[RCU_NUM_LVLS]; /* * In order to export the rcu_state name to the tracing tools, it * needs to be added in the __tracepoint_string section. * This requires defining a separate variable tp__varname * that points to the string being used, and this will allow * the tracing userspace tools to be able to decipher the string * address to the matching string. */ #ifdef CONFIG_TRACING # define DEFINE_RCU_TPS(sname) \ static char sname##_varname[] = #sname; \ static const char *tp_##sname##_varname __used __tracepoint_string = sname##_varname; # define RCU_STATE_NAME(sname) sname##_varname #else # define DEFINE_RCU_TPS(sname) # define RCU_STATE_NAME(sname) __stringify(sname) #endif #define RCU_STATE_INITIALIZER(sname, sabbr, cr) \ DEFINE_RCU_TPS(sname) \ struct rcu_state sname##_state = { \ .level = { &sname##_state.node[0] }, \ .call = cr, \ .fqs_state = RCU_GP_IDLE, \ .gpnum = 0UL - 300UL, \ .completed = 0UL - 300UL, \ .orphan_lock = __RAW_SPIN_LOCK_UNLOCKED(&sname##_state.orphan_lock), \ .orphan_nxttail = &sname##_state.orphan_nxtlist, \ .orphan_donetail = &sname##_state.orphan_donelist, \ .barrier_mutex = __MUTEX_INITIALIZER(sname##_state.barrier_mutex), \ .onoff_mutex = __MUTEX_INITIALIZER(sname##_state.onoff_mutex), \ .name = RCU_STATE_NAME(sname), \ .abbr = sabbr, \ }; \ DEFINE_PER_CPU_SHARED_ALIGNED(struct rcu_data, sname##_data) RCU_STATE_INITIALIZER(rcu_sched, 's', call_rcu_sched); RCU_STATE_INITIALIZER(rcu_bh, 'b', call_rcu_bh); static struct rcu_state *rcu_state_p; LIST_HEAD(rcu_struct_flavors); /* Increase (but not decrease) the CONFIG_RCU_FANOUT_LEAF at boot time. */ static int rcu_fanout_leaf = CONFIG_RCU_FANOUT_LEAF; module_param(rcu_fanout_leaf, int, 0444); int rcu_num_lvls __read_mostly = RCU_NUM_LVLS; static int num_rcu_lvl[] = { /* Number of rcu_nodes at specified level. */ NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2, NUM_RCU_LVL_3, NUM_RCU_LVL_4, }; int rcu_num_nodes __read_mostly = NUM_RCU_NODES; /* Total # rcu_nodes in use. */ /* * The rcu_scheduler_active variable transitions from zero to one just * before the first task is spawned. So when this variable is zero, RCU * can assume that there is but one task, allowing RCU to (for example) * optimize synchronize_sched() to a simple barrier(). When this variable * is one, RCU must actually do all the hard work required to detect real * grace periods. This variable is also used to suppress boot-time false * positives from lockdep-RCU error checking. */ int rcu_scheduler_active __read_mostly; EXPORT_SYMBOL_GPL(rcu_scheduler_active); /* * The rcu_scheduler_fully_active variable transitions from zero to one * during the early_initcall() processing, which is after the scheduler * is capable of creating new tasks. So RCU processing (for example, * creating tasks for RCU priority boosting) must be delayed until after * rcu_scheduler_fully_active transitions from zero to one. We also * currently delay invocation of any RCU callbacks until after this point. * * It might later prove better for people registering RCU callbacks during * early boot to take responsibility for these callbacks, but one step at * a time. */ static int rcu_scheduler_fully_active __read_mostly; static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu); static void invoke_rcu_core(void); static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp); /* rcuc/rcub kthread realtime priority */ static int kthread_prio = CONFIG_RCU_KTHREAD_PRIO; module_param(kthread_prio, int, 0644); /* * Track the rcutorture test sequence number and the update version * number within a given test. The rcutorture_testseq is incremented * on every rcutorture module load and unload, so has an odd value * when a test is running. The rcutorture_vernum is set to zero * when rcutorture starts and is incremented on each rcutorture update. * These variables enable correlating rcutorture output with the * RCU tracing information. */ unsigned long rcutorture_testseq; unsigned long rcutorture_vernum; /* * 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); } /* * 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, this just sets a flag. * The caller must have disabled preemption. */ void rcu_sched_qs(void) { if (!__this_cpu_read(rcu_sched_data.passed_quiesce)) { trace_rcu_grace_period(TPS("rcu_sched"), __this_cpu_read(rcu_sched_data.gpnum), TPS("cpuqs")); __this_cpu_write(rcu_sched_data.passed_quiesce, 1); } } void rcu_bh_qs(void) { if (!__this_cpu_read(rcu_bh_data.passed_quiesce)) { trace_rcu_grace_period(TPS("rcu_bh"), __this_cpu_read(rcu_bh_data.gpnum), TPS("cpuqs")); __this_cpu_write(rcu_bh_data.passed_quiesce, 1); } } static DEFINE_PER_CPU(int, rcu_sched_qs_mask); static DEFINE_PER_CPU(struct rcu_dynticks, rcu_dynticks) = { .dynticks_nesting = DYNTICK_TASK_EXIT_IDLE, .dynticks = ATOMIC_INIT(1), #ifdef CONFIG_NO_HZ_FULL_SYSIDLE .dynticks_idle_nesting = DYNTICK_TASK_NEST_VALUE, .dynticks_idle = ATOMIC_INIT(1), #endif /* #ifdef CONFIG_NO_HZ_FULL_SYSIDLE */ }; DEFINE_PER_CPU_SHARED_ALIGNED(unsigned long, rcu_qs_ctr); EXPORT_PER_CPU_SYMBOL_GPL(rcu_qs_ctr); /* * Let the RCU core know that this CPU has gone through the scheduler, * which is a quiescent state. This is called when the need for a * quiescent state is urgent, so we burn an atomic operation and full * memory barriers to let the RCU core know about it, regardless of what * this CPU might (or might not) do in the near future. * * We inform the RCU core by emulating a zero-duration dyntick-idle * period, which we in turn do by incrementing the ->dynticks counter * by two. */ static void rcu_momentary_dyntick_idle(void) { unsigned long flags; struct rcu_data *rdp; struct rcu_dynticks *rdtp; int resched_mask; struct rcu_state *rsp; local_irq_save(flags); /* * Yes, we can lose flag-setting operations. This is OK, because * the flag will be set again after some delay. */ resched_mask = raw_cpu_read(rcu_sched_qs_mask); raw_cpu_write(rcu_sched_qs_mask, 0); /* Find the flavor that needs a quiescent state. */ for_each_rcu_flavor(rsp) { rdp = raw_cpu_ptr(rsp->rda); if (!(resched_mask & rsp->flavor_mask)) continue; smp_mb(); /* rcu_sched_qs_mask before cond_resched_completed. */ if (ACCESS_ONCE(rdp->mynode->completed) != ACCESS_ONCE(rdp->cond_resched_completed)) continue; /* * Pretend to be momentarily idle for the quiescent state. * This allows the grace-period kthread to record the * quiescent state, with no need for this CPU to do anything * further. */ rdtp = this_cpu_ptr(&rcu_dynticks); smp_mb__before_atomic(); /* Earlier stuff before QS. */ atomic_add(2, &rdtp->dynticks); /* QS. */ smp_mb__after_atomic(); /* Later stuff after QS. */ break; } local_irq_restore(flags); } /* * Note a context switch. This is a quiescent state for RCU-sched, * and requires special handling for preemptible RCU. * The caller must have disabled preemption. */ void rcu_note_context_switch(void) { trace_rcu_utilization(TPS("Start context switch")); rcu_sched_qs(); rcu_preempt_note_context_switch(); if (unlikely(raw_cpu_read(rcu_sched_qs_mask))) rcu_momentary_dyntick_idle(); trace_rcu_utilization(TPS("End context switch")); } EXPORT_SYMBOL_GPL(rcu_note_context_switch); /* * Register a quiesecent state for all RCU flavors. If there is an * emergency, invoke rcu_momentary_dyntick_idle() to do a heavy-weight * dyntick-idle quiescent state visible to other CPUs (but only for those * RCU flavors in desparate need of a quiescent state, which will normally * be none of them). Either way, do a lightweight quiescent state for * all RCU flavors. */ void rcu_all_qs(void) { if (unlikely(raw_cpu_read(rcu_sched_qs_mask))) rcu_momentary_dyntick_idle(); this_cpu_inc(rcu_qs_ctr); } EXPORT_SYMBOL_GPL(rcu_all_qs); static long blimit = 10; /* Maximum callbacks per rcu_do_batch. */ static long qhimark = 10000; /* If this many pending, ignore blimit. */ static long qlowmark = 100; /* Once only this many pending, use blimit. */ module_param(blimit, long, 0444); module_param(qhimark, long, 0444); module_param(qlowmark, long, 0444); static ulong jiffies_till_first_fqs = ULONG_MAX; static ulong jiffies_till_next_fqs = ULONG_MAX; module_param(jiffies_till_first_fqs, ulong, 0644); module_param(jiffies_till_next_fqs, ulong, 0644); /* * How long the grace period must be before we start recruiting * quiescent-state help from rcu_note_context_switch(). */ static ulong jiffies_till_sched_qs = HZ / 20; module_param(jiffies_till_sched_qs, ulong, 0644); static bool rcu_start_gp_advanced(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp); static void force_qs_rnp(struct rcu_state *rsp, int (*f)(struct rcu_data *rsp, bool *isidle, unsigned long *maxj), bool *isidle, unsigned long *maxj); static void force_quiescent_state(struct rcu_state *rsp); static int rcu_pending(void); /* * Return the number of RCU batches started thus far for debug & stats. */ unsigned long rcu_batches_started(void) { return rcu_state_p->gpnum; } EXPORT_SYMBOL_GPL(rcu_batches_started); /* * Return the number of RCU-sched batches started thus far for debug & stats. */ unsigned long rcu_batches_started_sched(void) { return rcu_sched_state.gpnum; } EXPORT_SYMBOL_GPL(rcu_batches_started_sched); /* * Return the number of RCU BH batches started thus far for debug & stats. */ unsigned long rcu_batches_started_bh(void) { return rcu_bh_state.gpnum; } EXPORT_SYMBOL_GPL(rcu_batches_started_bh); /* * Return the number of RCU batches completed thus far for debug & stats. */ unsigned long rcu_batches_completed(void) { return rcu_state_p->completed; } EXPORT_SYMBOL_GPL(rcu_batches_completed); /* * Return the number of RCU-sched batches completed thus far for debug & stats. */ unsigned long rcu_batches_completed_sched(void) { return rcu_sched_state.completed; } EXPORT_SYMBOL_GPL(rcu_batches_completed_sched); /* * Return the number of RCU BH batches completed thus far for debug & stats. */ unsigned long rcu_batches_completed_bh(void) { return rcu_bh_state.completed; } EXPORT_SYMBOL_GPL(rcu_batches_completed_bh); /* * Force a quiescent state. */ void rcu_force_quiescent_state(void) { force_quiescent_state(rcu_state_p); } EXPORT_SYMBOL_GPL(rcu_force_quiescent_state); /* * Force a quiescent state for RCU BH. */ void rcu_bh_force_quiescent_state(void) { force_quiescent_state(&rcu_bh_state); } EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state); /* * Show the state of the grace-period kthreads. */ void show_rcu_gp_kthreads(void) { struct rcu_state *rsp; for_each_rcu_flavor(rsp) { pr_info("%s: wait state: %d ->state: %#lx\n", rsp->name, rsp->gp_state, rsp->gp_kthread->state); /* sched_show_task(rsp->gp_kthread); */ } } EXPORT_SYMBOL_GPL(show_rcu_gp_kthreads); /* * Record the number of times rcutorture tests have been initiated and * terminated. This information allows the debugfs tracing stats to be * correlated to the rcutorture messages, even when the rcutorture module * is being repeatedly loaded and unloaded. In other words, we cannot * store this state in rcutorture itself. */ void rcutorture_record_test_transition(void) { rcutorture_testseq++; rcutorture_vernum = 0; } EXPORT_SYMBOL_GPL(rcutorture_record_test_transition); /* * Send along grace-period-related data for rcutorture diagnostics. */ void rcutorture_get_gp_data(enum rcutorture_type test_type, int *flags, unsigned long *gpnum, unsigned long *completed) { struct rcu_state *rsp = NULL; switch (test_type) { case RCU_FLAVOR: rsp = rcu_state_p; break; case RCU_BH_FLAVOR: rsp = &rcu_bh_state; break; case RCU_SCHED_FLAVOR: rsp = &rcu_sched_state; break; default: break; } if (rsp != NULL) { *flags = ACCESS_ONCE(rsp->gp_flags); *gpnum = ACCESS_ONCE(rsp->gpnum); *completed = ACCESS_ONCE(rsp->completed); return; } *flags = 0; *gpnum = 0; *completed = 0; } EXPORT_SYMBOL_GPL(rcutorture_get_gp_data); /* * Record the number of writer passes through the current rcutorture test. * This is also used to correlate debugfs tracing stats with the rcutorture * messages. */ void rcutorture_record_progress(unsigned long vernum) { rcutorture_vernum++; } EXPORT_SYMBOL_GPL(rcutorture_record_progress); /* * Force a quiescent state for RCU-sched. */ void rcu_sched_force_quiescent_state(void) { force_quiescent_state(&rcu_sched_state); } EXPORT_SYMBOL_GPL(rcu_sched_force_quiescent_state); /* * Does the CPU have callbacks ready to be invoked? */ static int cpu_has_callbacks_ready_to_invoke(struct rcu_data *rdp) { return &rdp->nxtlist != rdp->nxttail[RCU_DONE_TAIL] && rdp->nxttail[RCU_DONE_TAIL] != NULL; } /* * Return the root node of the specified rcu_state structure. */ static struct rcu_node *rcu_get_root(struct rcu_state *rsp) { return &rsp->node[0]; } /* * Is there any need for future grace periods? * Interrupts must be disabled. If the caller does not hold the root * rnp_node structure's ->lock, the results are advisory only. */ static int rcu_future_needs_gp(struct rcu_state *rsp) { struct rcu_node *rnp = rcu_get_root(rsp); int idx = (ACCESS_ONCE(rnp->completed) + 1) & 0x1; int *fp = &rnp->need_future_gp[idx]; return ACCESS_ONCE(*fp); } /* * Does the current CPU require a not-yet-started grace period? * The caller must have disabled interrupts to prevent races with * normal callback registry. */ static int cpu_needs_another_gp(struct rcu_state *rsp, struct rcu_data *rdp) { int i; if (rcu_gp_in_progress(rsp)) return 0; /* No, a grace period is already in progress. */ if (rcu_future_needs_gp(rsp)) return 1; /* Yes, a no-CBs CPU needs one. */ if (!rdp->nxttail[RCU_NEXT_TAIL]) return 0; /* No, this is a no-CBs (or offline) CPU. */ if (*rdp->nxttail[RCU_NEXT_READY_TAIL]) return 1; /* Yes, this CPU has newly registered callbacks. */ for (i = RCU_WAIT_TAIL; i < RCU_NEXT_TAIL; i++) if (rdp->nxttail[i - 1] != rdp->nxttail[i] && ULONG_CMP_LT(ACCESS_ONCE(rsp->completed), rdp->nxtcompleted[i])) return 1; /* Yes, CBs for future grace period. */ return 0; /* No grace period needed. */ } /* * rcu_eqs_enter_common - current CPU is moving towards extended quiescent state * * If the new value of the ->dynticks_nesting counter now is zero, * we really have entered idle, and must do the appropriate accounting. * The caller must have disabled interrupts. */ static void rcu_eqs_enter_common(long long oldval, bool user) { struct rcu_state *rsp; struct rcu_data *rdp; struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); trace_rcu_dyntick(TPS("Start"), oldval, rdtp->dynticks_nesting); if (!user && !is_idle_task(current)) { struct task_struct *idle __maybe_unused = idle_task(smp_processor_id()); trace_rcu_dyntick(TPS("Error on entry: not idle task"), oldval, 0); ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, idle->pid, idle->comm); /* must be idle task! */ } for_each_rcu_flavor(rsp) { rdp = this_cpu_ptr(rsp->rda); do_nocb_deferred_wakeup(rdp); } rcu_prepare_for_idle(); /* CPUs seeing atomic_inc() must see prior RCU read-side crit sects */ smp_mb__before_atomic(); /* See above. */ atomic_inc(&rdtp->dynticks); smp_mb__after_atomic(); /* Force ordering with next sojourn. */ WARN_ON_ONCE(atomic_read(&rdtp->dynticks) & 0x1); rcu_dynticks_task_enter(); /* * It is illegal to enter an extended quiescent state while * in an RCU read-side critical section. */ rcu_lockdep_assert(!lock_is_held(&rcu_lock_map), "Illegal idle entry in RCU read-side critical section."); rcu_lockdep_assert(!lock_is_held(&rcu_bh_lock_map), "Illegal idle entry in RCU-bh read-side critical section."); rcu_lockdep_assert(!lock_is_held(&rcu_sched_lock_map), "Illegal idle entry in RCU-sched read-side critical section."); } /* * Enter an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. */ static void rcu_eqs_enter(bool user) { long long oldval; struct rcu_dynticks *rdtp; rdtp = this_cpu_ptr(&rcu_dynticks); oldval = rdtp->dynticks_nesting; WARN_ON_ONCE((oldval & DYNTICK_TASK_NEST_MASK) == 0); if ((oldval & DYNTICK_TASK_NEST_MASK) == DYNTICK_TASK_NEST_VALUE) { rdtp->dynticks_nesting = 0; rcu_eqs_enter_common(oldval, user); } else { rdtp->dynticks_nesting -= DYNTICK_TASK_NEST_VALUE; } } /** * rcu_idle_enter - inform RCU that current CPU is entering idle * * Enter idle mode, in other words, -leave- the mode in which RCU * read-side critical sections can occur. (Though RCU read-side * critical sections can occur in irq handlers in idle, a possibility * handled by irq_enter() and irq_exit().) * * We crowbar the ->dynticks_nesting field to zero to allow for * the possibility of usermode upcalls having messed up our count * of interrupt nesting level during the prior busy period. */ void rcu_idle_enter(void) { unsigned long flags; local_irq_save(flags); rcu_eqs_enter(false); rcu_sysidle_enter(0); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(rcu_idle_enter); #ifdef CONFIG_RCU_USER_QS /** * rcu_user_enter - inform RCU that we are resuming userspace. * * Enter RCU idle mode right before resuming userspace. No use of RCU * is permitted between this call and rcu_user_exit(). This way the * CPU doesn't need to maintain the tick for RCU maintenance purposes * when the CPU runs in userspace. */ void rcu_user_enter(void) { rcu_eqs_enter(1); } #endif /* CONFIG_RCU_USER_QS */ /** * rcu_irq_exit - inform RCU that current CPU is exiting irq towards idle * * Exit from an interrupt handler, which might possibly result in entering * idle mode, in other words, leaving the mode in which read-side critical * sections can occur. * * This code assumes that the idle loop never does anything that might * result in unbalanced calls to irq_enter() and irq_exit(). If your * architecture violates this assumption, RCU will give you what you * deserve, good and hard. But very infrequently and irreproducibly. * * Use things like work queues to work around this limitation. * * You have been warned. */ void rcu_irq_exit(void) { unsigned long flags; long long oldval; struct rcu_dynticks *rdtp; local_irq_save(flags); rdtp = this_cpu_ptr(&rcu_dynticks); oldval = rdtp->dynticks_nesting; rdtp->dynticks_nesting--; WARN_ON_ONCE(rdtp->dynticks_nesting < 0); if (rdtp->dynticks_nesting) trace_rcu_dyntick(TPS("--="), oldval, rdtp->dynticks_nesting); else rcu_eqs_enter_common(oldval, true); rcu_sysidle_enter(1); local_irq_restore(flags); } /* * rcu_eqs_exit_common - current CPU moving away from extended quiescent state * * If the new value of the ->dynticks_nesting counter was previously zero, * we really have exited idle, and must do the appropriate accounting. * The caller must have disabled interrupts. */ static void rcu_eqs_exit_common(long long oldval, int user) { struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); rcu_dynticks_task_exit(); smp_mb__before_atomic(); /* Force ordering w/previous sojourn. */ atomic_inc(&rdtp->dynticks); /* CPUs seeing atomic_inc() must see later RCU read-side crit sects */ smp_mb__after_atomic(); /* See above. */ WARN_ON_ONCE(!(atomic_read(&rdtp->dynticks) & 0x1)); rcu_cleanup_after_idle(); trace_rcu_dyntick(TPS("End"), oldval, rdtp->dynticks_nesting); if (!user && !is_idle_task(current)) { struct task_struct *idle __maybe_unused = idle_task(smp_processor_id()); trace_rcu_dyntick(TPS("Error on exit: not idle task"), oldval, rdtp->dynticks_nesting); ftrace_dump(DUMP_ORIG); WARN_ONCE(1, "Current pid: %d comm: %s / Idle pid: %d comm: %s", current->pid, current->comm, idle->pid, idle->comm); /* must be idle task! */ } } /* * Exit an RCU extended quiescent state, which can be either the * idle loop or adaptive-tickless usermode execution. */ static void rcu_eqs_exit(bool user) { struct rcu_dynticks *rdtp; long long oldval; rdtp = this_cpu_ptr(&rcu_dynticks); oldval = rdtp->dynticks_nesting; WARN_ON_ONCE(oldval < 0); if (oldval & DYNTICK_TASK_NEST_MASK) { rdtp->dynticks_nesting += DYNTICK_TASK_NEST_VALUE; } else { rdtp->dynticks_nesting = DYNTICK_TASK_EXIT_IDLE; rcu_eqs_exit_common(oldval, user); } } /** * rcu_idle_exit - inform RCU that current CPU is leaving idle * * Exit idle mode, in other words, -enter- the mode in which RCU * read-side critical sections can occur. * * We crowbar the ->dynticks_nesting field to DYNTICK_TASK_NEST to * allow for the possibility of usermode upcalls messing up our count * of interrupt nesting level during the busy period that is just * now starting. */ void rcu_idle_exit(void) { unsigned long flags; local_irq_save(flags); rcu_eqs_exit(false); rcu_sysidle_exit(0); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(rcu_idle_exit); #ifdef CONFIG_RCU_USER_QS /** * rcu_user_exit - inform RCU that we are exiting userspace. * * Exit RCU idle mode while entering the kernel because it can * run a RCU read side critical section anytime. */ void rcu_user_exit(void) { rcu_eqs_exit(1); } #endif /* CONFIG_RCU_USER_QS */ /** * rcu_irq_enter - inform RCU that current CPU is entering irq away from idle * * Enter an interrupt handler, which might possibly result in exiting * idle mode, in other words, entering the mode in which read-side critical * sections can occur. * * Note that the Linux kernel is fully capable of entering an interrupt * handler that it never exits, for example when doing upcalls to * user mode! This code assumes that the idle loop never does upcalls to * user mode. If your architecture does do upcalls from the idle loop (or * does anything else that results in unbalanced calls to the irq_enter() * and irq_exit() functions), RCU will give you what you deserve, good * and hard. But very infrequently and irreproducibly. * * Use things like work queues to work around this limitation. * * You have been warned. */ void rcu_irq_enter(void) { unsigned long flags; struct rcu_dynticks *rdtp; long long oldval; local_irq_save(flags); rdtp = this_cpu_ptr(&rcu_dynticks); oldval = rdtp->dynticks_nesting; rdtp->dynticks_nesting++; WARN_ON_ONCE(rdtp->dynticks_nesting == 0); if (oldval) trace_rcu_dyntick(TPS("++="), oldval, rdtp->dynticks_nesting); else rcu_eqs_exit_common(oldval, true); rcu_sysidle_exit(1); local_irq_restore(flags); } /** * rcu_nmi_enter - inform RCU of entry to NMI context * * If the CPU was idle from RCU's viewpoint, update rdtp->dynticks and * rdtp->dynticks_nmi_nesting to let the RCU grace-period handling know * that the CPU is active. This implementation permits nested NMIs, as * long as the nesting level does not overflow an int. (You will probably * run out of stack space first.) */ void rcu_nmi_enter(void) { struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); int incby = 2; /* Complain about underflow. */ WARN_ON_ONCE(rdtp->dynticks_nmi_nesting < 0); /* * If idle from RCU viewpoint, atomically increment ->dynticks * to mark non-idle and increment ->dynticks_nmi_nesting by one. * Otherwise, increment ->dynticks_nmi_nesting by two. This means * if ->dynticks_nmi_nesting is equal to one, we are guaranteed * to be in the outermost NMI handler that interrupted an RCU-idle * period (observation due to Andy Lutomirski). */ if (!(atomic_read(&rdtp->dynticks) & 0x1)) { smp_mb__before_atomic(); /* Force delay from prior write. */ atomic_inc(&rdtp->dynticks); /* atomic_inc() before later RCU read-side crit sects */ smp_mb__after_atomic(); /* See above. */ WARN_ON_ONCE(!(atomic_read(&rdtp->dynticks) & 0x1)); incby = 1; } rdtp->dynticks_nmi_nesting += incby; barrier(); } /** * rcu_nmi_exit - inform RCU of exit from NMI context * * If we are returning from the outermost NMI handler that interrupted an * RCU-idle period, update rdtp->dynticks and rdtp->dynticks_nmi_nesting * to let the RCU grace-period handling know that the CPU is back to * being RCU-idle. */ void rcu_nmi_exit(void) { struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks); /* * Check for ->dynticks_nmi_nesting underflow and bad ->dynticks. * (We are exiting an NMI handler, so RCU better be paying attention * to us!) */ WARN_ON_ONCE(rdtp->dynticks_nmi_nesting <= 0); WARN_ON_ONCE(!(atomic_read(&rdtp->dynticks) & 0x1)); /* * If the nesting level is not 1, the CPU wasn't RCU-idle, so * leave it in non-RCU-idle state. */ if (rdtp->dynticks_nmi_nesting != 1) { rdtp->dynticks_nmi_nesting -= 2; return; } /* This NMI interrupted an RCU-idle CPU, restore RCU-idleness. */ rdtp->dynticks_nmi_nesting = 0; /* CPUs seeing atomic_inc() must see prior RCU read-side crit sects */ smp_mb__before_atomic(); /* See above. */ atomic_inc(&rdtp->dynticks); smp_mb__after_atomic(); /* Force delay to next write. */ WARN_ON_ONCE(atomic_read(&rdtp->dynticks) & 0x1); } /** * __rcu_is_watching - are RCU read-side critical sections safe? * * Return true if RCU is watching the running CPU, which means that * this CPU can safely enter RCU read-side critical sections. Unlike * rcu_is_watching(), the caller of __rcu_is_watching() must have at * least disabled preemption. */ bool notrace __rcu_is_watching(void) { return atomic_read(this_cpu_ptr(&rcu_dynticks.dynticks)) & 0x1; } /** * rcu_is_watching - see if RCU thinks that the current CPU is idle * * If the current CPU is in its idle loop and is neither in an interrupt * or NMI handler, return true. */ bool notrace rcu_is_watching(void) { bool ret; preempt_disable(); ret = __rcu_is_watching(); preempt_enable(); return ret; } EXPORT_SYMBOL_GPL(rcu_is_watching); #if defined(CONFIG_PROVE_RCU) && defined(CONFIG_HOTPLUG_CPU) /* * Is the current CPU online? Disable preemption to avoid false positives * that could otherwise happen due to the current CPU number being sampled, * this task being preempted, its old CPU being taken offline, resuming * on some other CPU, then determining that its old CPU is now offline. * It is OK to use RCU on an offline processor during initial boot, hence * the check for rcu_scheduler_fully_active. Note also that it is OK * for a CPU coming online to use RCU for one jiffy prior to marking itself * online in the cpu_online_mask. Similarly, it is OK for a CPU going * offline to continue to use RCU for one jiffy after marking itself * offline in the cpu_online_mask. This leniency is necessary given the * non-atomic nature of the online and offline processing, for example, * the fact that a CPU enters the scheduler after completing the CPU_DYING * notifiers. * * This is also why RCU internally marks CPUs online during the * CPU_UP_PREPARE phase and offline during the CPU_DEAD phase. * * Disable checking if in an NMI handler because we cannot safely report * errors from NMI handlers anyway. */ bool rcu_lockdep_current_cpu_online(void) { struct rcu_data *rdp; struct rcu_node *rnp; bool ret; if (in_nmi()) return true; preempt_disable(); rdp = this_cpu_ptr(&rcu_sched_data); rnp = rdp->mynode; ret = (rdp->grpmask & rnp->qsmaskinit) || !rcu_scheduler_fully_active; preempt_enable(); return ret; } EXPORT_SYMBOL_GPL(rcu_lockdep_current_cpu_online); #endif /* #if defined(CONFIG_PROVE_RCU) && defined(CONFIG_HOTPLUG_CPU) */ /** * rcu_is_cpu_rrupt_from_idle - see if idle or immediately interrupted from idle * * If the current CPU is idle or running at a first-level (not nested) * interrupt from idle, return true. The caller must have at least * disabled preemption. */ static int rcu_is_cpu_rrupt_from_idle(void) { return __this_cpu_read(rcu_dynticks.dynticks_nesting) <= 1; } /* * 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 in dynticks idle mode, which is an extended quiescent state. */ static int dyntick_save_progress_counter(struct rcu_data *rdp, bool *isidle, unsigned long *maxj) { rdp->dynticks_snap = atomic_add_return(0, &rdp->dynticks->dynticks); rcu_sysidle_check_cpu(rdp, isidle, maxj); if ((rdp->dynticks_snap & 0x1) == 0) { trace_rcu_fqs(rdp->rsp->name, rdp->gpnum, rdp->cpu, TPS("dti")); return 1; } else { if (ULONG_CMP_LT(ACCESS_ONCE(rdp->gpnum) + ULONG_MAX / 4, rdp->mynode->gpnum)) ACCESS_ONCE(rdp->gpwrap) = true; return 0; } } /* * Return true if the specified CPU has passed through a quiescent * state by virtue of being in or having passed through an dynticks * idle state since the last call to dyntick_save_progress_counter() * for this same CPU, or by virtue of having been offline. */ static int rcu_implicit_dynticks_qs(struct rcu_data *rdp, bool *isidle, unsigned long *maxj) { unsigned int curr; int *rcrmp; unsigned int snap; curr = (unsigned int)atomic_add_return(0, &rdp->dynticks->dynticks); snap = (unsigned int)rdp->dynticks_snap; /* * If the CPU passed through or entered a dynticks idle phase with * no active irq/NMI handlers, then we can safely pretend that the CPU * already acknowledged the request to pass through a quiescent * state. Either way, that CPU cannot possibly be in an RCU * read-side critical section that started before the beginning * of the current RCU grace period. */ if ((curr & 0x1) == 0 || UINT_CMP_GE(curr, snap + 2)) { trace_rcu_fqs(rdp->rsp->name, rdp->gpnum, rdp->cpu, TPS("dti")); rdp->dynticks_fqs++; return 1; } /* * Check for the CPU being offline, but only if the grace period * is old enough. We don't need to worry about the CPU changing * state: If we see it offline even once, it has been through a * quiescent state. * * The reason for insisting that the grace period be at least * one jiffy old is that CPUs that are not quite online and that * have just gone offline can still execute RCU read-side critical * sections. */ if (ULONG_CMP_GE(rdp->rsp->gp_start + 2, jiffies)) return 0; /* Grace period is not old enough. */ barrier(); if (cpu_is_offline(rdp->cpu)) { trace_rcu_fqs(rdp->rsp->name, rdp->gpnum, rdp->cpu, TPS("ofl")); rdp->offline_fqs++; return 1; } /* * A CPU running for an extended time within the kernel can * delay RCU grace periods. When the CPU is in NO_HZ_FULL mode, * even context-switching back and forth between a pair of * in-kernel CPU-bound tasks cannot advance grace periods. * So if the grace period is old enough, make the CPU pay attention. * Note that the unsynchronized assignments to the per-CPU * rcu_sched_qs_mask variable are safe. Yes, setting of * bits can be lost, but they will be set again on the next * force-quiescent-state pass. So lost bit sets do not result * in incorrect behavior, merely in a grace period lasting * a few jiffies longer than it might otherwise. Because * there are at most four threads involved, and because the * updates are only once every few jiffies, the probability of * lossage (and thus of slight grace-period extension) is * quite low. * * Note that if the jiffies_till_sched_qs boot/sysfs parameter * is set too high, we override with half of the RCU CPU stall * warning delay. */ rcrmp = &per_cpu(rcu_sched_qs_mask, rdp->cpu); if (ULONG_CMP_GE(jiffies, rdp->rsp->gp_start + jiffies_till_sched_qs) || ULONG_CMP_GE(jiffies, rdp->rsp->jiffies_resched)) { if (!(ACCESS_ONCE(*rcrmp) & rdp->rsp->flavor_mask)) { ACCESS_ONCE(rdp->cond_resched_completed) = ACCESS_ONCE(rdp->mynode->completed); smp_mb(); /* ->cond_resched_completed before *rcrmp. */ ACCESS_ONCE(*rcrmp) = ACCESS_ONCE(*rcrmp) + rdp->rsp->flavor_mask; resched_cpu(rdp->cpu); /* Force CPU into scheduler. */ rdp->rsp->jiffies_resched += 5; /* Enable beating. */ } else if (ULONG_CMP_GE(jiffies, rdp->rsp->jiffies_resched)) { /* Time to beat on that CPU again! */ resched_cpu(rdp->cpu); /* Force CPU into scheduler. */ rdp->rsp->jiffies_resched += 5; /* Re-enable beating. */ } } return 0; } static void record_gp_stall_check_time(struct rcu_state *rsp) { unsigned long j = jiffies; unsigned long j1; rsp->gp_start = j; smp_wmb(); /* Record start time before stall time. */ j1 = rcu_jiffies_till_stall_check(); ACCESS_ONCE(rsp->jiffies_stall) = j + j1; rsp->jiffies_resched = j + j1 / 2; rsp->n_force_qs_gpstart = ACCESS_ONCE(rsp->n_force_qs); } /* * Complain about starvation of grace-period kthread. */ static void rcu_check_gp_kthread_starvation(struct rcu_state *rsp) { unsigned long gpa; unsigned long j; j = jiffies; gpa = ACCESS_ONCE(rsp->gp_activity); if (j - gpa > 2 * HZ) pr_err("%s kthread starved for %ld jiffies!\n", rsp->name, j - gpa); } /* * Dump stacks of all tasks running on stalled CPUs. */ static void rcu_dump_cpu_stacks(struct rcu_state *rsp) { int cpu; unsigned long flags; struct rcu_node *rnp; rcu_for_each_leaf_node(rsp, rnp) { raw_spin_lock_irqsave(&rnp->lock, flags); if (rnp->qsmask != 0) { for (cpu = 0; cpu <= rnp->grphi - rnp->grplo; cpu++) if (rnp->qsmask & (1UL << cpu)) dump_cpu_task(rnp->grplo + cpu); } raw_spin_unlock_irqrestore(&rnp->lock, flags); } } static void print_other_cpu_stall(struct rcu_state *rsp, unsigned long gpnum) { int cpu; long delta; unsigned long flags; unsigned long gpa; unsigned long j; int ndetected = 0; struct rcu_node *rnp = rcu_get_root(rsp); long totqlen = 0; /* Only let one CPU complain about others per time interval. */ raw_spin_lock_irqsave(&rnp->lock, flags); delta = jiffies - ACCESS_ONCE(rsp->jiffies_stall); if (delta < RCU_STALL_RAT_DELAY || !rcu_gp_in_progress(rsp)) { raw_spin_unlock_irqrestore(&rnp->lock, flags); return; } ACCESS_ONCE(rsp->jiffies_stall) = jiffies + 3 * rcu_jiffies_till_stall_check() + 3; raw_spin_unlock_irqrestore(&rnp->lock, flags); /* * OK, time to rat on our buddy... * See Documentation/RCU/stallwarn.txt for info on how to debug * RCU CPU stall warnings. */ pr_err("INFO: %s detected stalls on CPUs/tasks:", rsp->name); print_cpu_stall_info_begin(); rcu_for_each_leaf_node(rsp, rnp) { raw_spin_lock_irqsave(&rnp->lock, flags); ndetected += rcu_print_task_stall(rnp); if (rnp->qsmask != 0) { for (cpu = 0; cpu <= rnp->grphi - rnp->grplo; cpu++) if (rnp->qsmask & (1UL << cpu)) { print_cpu_stall_info(rsp, rnp->grplo + cpu); ndetected++; } } raw_spin_unlock_irqrestore(&rnp->lock, flags); } print_cpu_stall_info_end(); for_each_possible_cpu(cpu) totqlen += per_cpu_ptr(rsp->rda, cpu)->qlen; pr_cont("(detected by %d, t=%ld jiffies, g=%ld, c=%ld, q=%lu)\n", smp_processor_id(), (long)(jiffies - rsp->gp_start), (long)rsp->gpnum, (long)rsp->completed, totqlen); if (ndetected) { rcu_dump_cpu_stacks(rsp); } else { if (ACCESS_ONCE(rsp->gpnum) != gpnum || ACCESS_ONCE(rsp->completed) == gpnum) { pr_err("INFO: Stall ended before state dump start\n"); } else { j = jiffies; gpa = ACCESS_ONCE(rsp->gp_activity); pr_err("All QSes seen, last %s kthread activity %ld (%ld-%ld), jiffies_till_next_fqs=%ld\n", rsp->name, j - gpa, j, gpa, jiffies_till_next_fqs); /* In this case, the current CPU might be at fault. */ sched_show_task(current); } } /* Complain about tasks blocking the grace period. */ rcu_print_detail_task_stall(rsp); rcu_check_gp_kthread_starvation(rsp); force_quiescent_state(rsp); /* Kick them all. */ } static void print_cpu_stall(struct rcu_state *rsp) { int cpu; unsigned long flags; struct rcu_node *rnp = rcu_get_root(rsp); long totqlen = 0; /* * OK, time to rat on ourselves... * See Documentation/RCU/stallwarn.txt for info on how to debug * RCU CPU stall warnings. */ pr_err("INFO: %s self-detected stall on CPU", rsp->name); print_cpu_stall_info_begin(); print_cpu_stall_info(rsp, smp_processor_id()); print_cpu_stall_info_end(); for_each_possible_cpu(cpu) totqlen += per_cpu_ptr(rsp->rda, cpu)->qlen; pr_cont(" (t=%lu jiffies g=%ld c=%ld q=%lu)\n", jiffies - rsp->gp_start, (long)rsp->gpnum, (long)rsp->completed, totqlen); rcu_check_gp_kthread_starvation(rsp); rcu_dump_cpu_stacks(rsp); raw_spin_lock_irqsave(&rnp->lock, flags); if (ULONG_CMP_GE(jiffies, ACCESS_ONCE(rsp->jiffies_stall))) ACCESS_ONCE(rsp->jiffies_stall) = jiffies + 3 * rcu_jiffies_till_stall_check() + 3; raw_spin_unlock_irqrestore(&rnp->lock, flags); /* * Attempt to revive the RCU machinery by forcing a context switch. * * A context switch would normally allow the RCU state machine to make * progress and it could be we're stuck in kernel space without context * switches for an entirely unreasonable amount of time. */ resched_cpu(smp_processor_id()); } static void check_cpu_stall(struct rcu_state *rsp, struct rcu_data *rdp) { unsigned long completed; unsigned long gpnum; unsigned long gps; unsigned long j; unsigned long js; struct rcu_node *rnp; if (rcu_cpu_stall_suppress || !rcu_gp_in_progress(rsp)) return; j = jiffies; /* * Lots of memory barriers to reject false positives. * * The idea is to pick up rsp->gpnum, then rsp->jiffies_stall, * then rsp->gp_start, and finally rsp->completed. These values * are updated in the opposite order with memory barriers (or * equivalent) during grace-period initialization and cleanup. * Now, a false positive can occur if we get an new value of * rsp->gp_start and a old value of rsp->jiffies_stall. But given * the memory barriers, the only way that this can happen is if one * grace period ends and another starts between these two fetches. * Detect this by comparing rsp->completed with the previous fetch * from rsp->gpnum. * * Given this check, comparisons of jiffies, rsp->jiffies_stall, * and rsp->gp_start suffice to forestall false positives. */ gpnum = ACCESS_ONCE(rsp->gpnum); smp_rmb(); /* Pick up ->gpnum first... */ js = ACCESS_ONCE(rsp->jiffies_stall); smp_rmb(); /* ...then ->jiffies_stall before the rest... */ gps = ACCESS_ONCE(rsp->gp_start); smp_rmb(); /* ...and finally ->gp_start before ->completed. */ completed = ACCESS_ONCE(rsp->completed); if (ULONG_CMP_GE(completed, gpnum) || ULONG_CMP_LT(j, js) || ULONG_CMP_GE(gps, js)) return; /* No stall or GP completed since entering function. */ rnp = rdp->mynode; if (rcu_gp_in_progress(rsp) && (ACCESS_ONCE(rnp->qsmask) & rdp->grpmask)) { /* We haven't checked in, so go dump stack. */ print_cpu_stall(rsp); } else if (rcu_gp_in_progress(rsp) && ULONG_CMP_GE(j, js + RCU_STALL_RAT_DELAY)) { /* They had a few time units to dump stack, so complain. */ print_other_cpu_stall(rsp, gpnum); } } /** * rcu_cpu_stall_reset - prevent further stall warnings in current grace period * * Set the stall-warning timeout way off into the future, thus preventing * any RCU CPU stall-warning messages from appearing in the current set of * RCU grace periods. * * The caller must disable hard irqs. */ void rcu_cpu_stall_reset(void) { struct rcu_state *rsp; for_each_rcu_flavor(rsp) ACCESS_ONCE(rsp->jiffies_stall) = jiffies + ULONG_MAX / 2; } /* * Initialize the specified rcu_data structure's callback list to empty. */ static void init_callback_list(struct rcu_data *rdp) { int i; if (init_nocb_callback_list(rdp)) return; rdp->nxtlist = NULL; for (i = 0; i < RCU_NEXT_SIZE; i++) rdp->nxttail[i] = &rdp->nxtlist; } /* * Determine the value that ->completed will have at the end of the * next subsequent grace period. This is used to tag callbacks so that * a CPU can invoke callbacks in a timely fashion even if that CPU has * been dyntick-idle for an extended period with callbacks under the * influence of RCU_FAST_NO_HZ. * * The caller must hold rnp->lock with interrupts disabled. */ static unsigned long rcu_cbs_completed(struct rcu_state *rsp, struct rcu_node *rnp) { /* * If RCU is idle, we just wait for the next grace period. * But we can only be sure that RCU is idle if we are looking * at the root rcu_node structure -- otherwise, a new grace * period might have started, but just not yet gotten around * to initializing the current non-root rcu_node structure. */ if (rcu_get_root(rsp) == rnp && rnp->gpnum == rnp->completed) return rnp->completed + 1; /* * Otherwise, wait for a possible partial grace period and * then the subsequent full grace period. */ return rnp->completed + 2; } /* * Trace-event helper function for rcu_start_future_gp() and * rcu_nocb_wait_gp(). */ static void trace_rcu_future_gp(struct rcu_node *rnp, struct rcu_data *rdp, unsigned long c, const char *s) { trace_rcu_future_grace_period(rdp->rsp->name, rnp->gpnum, rnp->completed, c, rnp->level, rnp->grplo, rnp->grphi, s); } /* * Start some future grace period, as needed to handle newly arrived * callbacks. The required future grace periods are recorded in each * rcu_node structure's ->need_future_gp field. Returns true if there * is reason to awaken the grace-period kthread. * * The caller must hold the specified rcu_node structure's ->lock. */ static bool __maybe_unused rcu_start_future_gp(struct rcu_node *rnp, struct rcu_data *rdp, unsigned long *c_out) { unsigned long c; int i; bool ret = false; struct rcu_node *rnp_root = rcu_get_root(rdp->rsp); /* * Pick up grace-period number for new callbacks. If this * grace period is already marked as needed, return to the caller. */ c = rcu_cbs_completed(rdp->rsp, rnp); trace_rcu_future_gp(rnp, rdp, c, TPS("Startleaf")); if (rnp->need_future_gp[c & 0x1]) { trace_rcu_future_gp(rnp, rdp, c, TPS("Prestartleaf")); goto out; } /* * If either this rcu_node structure or the root rcu_node structure * believe that a grace period is in progress, then we must wait * for the one following, which is in "c". Because our request * will be noticed at the end of the current grace period, we don't * need to explicitly start one. We only do the lockless check * of rnp_root's fields if the current rcu_node structure thinks * there is no grace period in flight, and because we hold rnp->lock, * the only possible change is when rnp_root's two fields are * equal, in which case rnp_root->gpnum might be concurrently * incremented. But that is OK, as it will just result in our * doing some extra useless work. */ if (rnp->gpnum != rnp->completed || ACCESS_ONCE(rnp_root->gpnum) != ACCESS_ONCE(rnp_root->completed)) { rnp->need_future_gp[c & 0x1]++; trace_rcu_future_gp(rnp, rdp, c, TPS("Startedleaf")); goto out; } /* * There might be no grace period in progress. If we don't already * hold it, acquire the root rcu_node structure's lock in order to * start one (if needed). */ if (rnp != rnp_root) { raw_spin_lock(&rnp_root->lock); smp_mb__after_unlock_lock(); } /* * Get a new grace-period number. If there really is no grace * period in progress, it will be smaller than the one we obtained * earlier. Adjust callbacks as needed. Note that even no-CBs * CPUs have a ->nxtcompleted[] array, so no no-CBs checks needed. */ c = rcu_cbs_completed(rdp->rsp, rnp_root); for (i = RCU_DONE_TAIL; i < RCU_NEXT_TAIL; i++) if (ULONG_CMP_LT(c, rdp->nxtcompleted[i])) rdp->nxtcompleted[i] = c; /* * If the needed for the required grace period is already * recorded, trace and leave. */ if (rnp_root->need_future_gp[c & 0x1]) { trace_rcu_future_gp(rnp, rdp, c, TPS("Prestartedroot")); goto unlock_out; } /* Record the need for the future grace period. */ rnp_root->need_future_gp[c & 0x1]++; /* If a grace period is not already in progress, start one. */ if (rnp_root->gpnum != rnp_root->completed) { trace_rcu_future_gp(rnp, rdp, c, TPS("Startedleafroot")); } else { trace_rcu_future_gp(rnp, rdp, c, TPS("Startedroot")); ret = rcu_start_gp_advanced(rdp->rsp, rnp_root, rdp); } unlock_out: if (rnp != rnp_root) raw_spin_unlock(&rnp_root->lock); out: if (c_out != NULL) *c_out = c; return ret; } /* * Clean up any old requests for the just-ended grace period. Also return * whether any additional grace periods have been requested. Also invoke * rcu_nocb_gp_cleanup() in order to wake up any no-callbacks kthreads * waiting for this grace period to complete. */ static int rcu_future_gp_cleanup(struct rcu_state *rsp, struct rcu_node *rnp) { int c = rnp->completed; int needmore; struct rcu_data *rdp = this_cpu_ptr(rsp->rda); rcu_nocb_gp_cleanup(rsp, rnp); rnp->need_future_gp[c & 0x1] = 0; needmore = rnp->need_future_gp[(c + 1) & 0x1]; trace_rcu_future_gp(rnp, rdp, c, needmore ? TPS("CleanupMore") : TPS("Cleanup")); return needmore; } /* * Awaken the grace-period kthread for the specified flavor of RCU. * Don't do a self-awaken, and don't bother awakening when there is * nothing for the grace-period kthread to do (as in several CPUs * raced to awaken, and we lost), and finally don't try to awaken * a kthread that has not yet been created. */ static void rcu_gp_kthread_wake(struct rcu_state *rsp) { if (current == rsp->gp_kthread || !ACCESS_ONCE(rsp->gp_flags) || !rsp->gp_kthread) return; wake_up(&rsp->gp_wq); } /* * If there is room, assign a ->completed number to any callbacks on * this CPU that have not already been assigned. Also accelerate any * callbacks that were previously assigned a ->completed number that has * since proven to be too conservative, which can happen if callbacks get * assigned a ->completed number while RCU is idle, but with reference to * a non-root rcu_node structure. This function is idempotent, so it does * not hurt to call it repeatedly. Returns an flag saying that we should * awaken the RCU grace-period kthread. * * The caller must hold rnp->lock with interrupts disabled. */ static bool rcu_accelerate_cbs(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) { unsigned long c; int i; bool ret; /* If the CPU has no callbacks, nothing to do. */ if (!rdp->nxttail[RCU_NEXT_TAIL] || !*rdp->nxttail[RCU_DONE_TAIL]) return false; /* * Starting from the sublist containing the callbacks most * recently assigned a ->completed number and working down, find the * first sublist that is not assignable to an upcoming grace period. * Such a sublist has something in it (first two tests) and has * a ->completed number assigned that will complete sooner than * the ->completed number for newly arrived callbacks (last test). * * The key point is that any later sublist can be assigned the * same ->completed number as the newly arrived callbacks, which * means that the callbacks in any of these later sublist can be * grouped into a single sublist, whether or not they have already * been assigned a ->completed number. */ c = rcu_cbs_completed(rsp, rnp); for (i = RCU_NEXT_TAIL - 1; i > RCU_DONE_TAIL; i--) if (rdp->nxttail[i] != rdp->nxttail[i - 1] && !ULONG_CMP_GE(rdp->nxtcompleted[i], c)) break; /* * If there are no sublist for unassigned callbacks, leave. * At the same time, advance "i" one sublist, so that "i" will * index into the sublist where all the remaining callbacks should * be grouped into. */ if (++i >= RCU_NEXT_TAIL) return false; /* * Assign all subsequent callbacks' ->completed number to the next * full grace period and group them all in the sublist initially * indexed by "i". */ for (; i <= RCU_NEXT_TAIL; i++) { rdp->nxttail[i] = rdp->nxttail[RCU_NEXT_TAIL]; rdp->nxtcompleted[i] = c; } /* Record any needed additional grace periods. */ ret = rcu_start_future_gp(rnp, rdp, NULL); /* Trace depending on how much we were able to accelerate. */ if (!*rdp->nxttail[RCU_WAIT_TAIL]) trace_rcu_grace_period(rsp->name, rdp->gpnum, TPS("AccWaitCB")); else trace_rcu_grace_period(rsp->name, rdp->gpnum, TPS("AccReadyCB")); return ret; } /* * Move any callbacks whose grace period has completed to the * RCU_DONE_TAIL sublist, then compact the remaining sublists and * assign ->completed numbers to any callbacks in the RCU_NEXT_TAIL * sublist. This function is idempotent, so it does not hurt to * invoke it repeatedly. As long as it is not invoked -too- often... * Returns true if the RCU grace-period kthread needs to be awakened. * * The caller must hold rnp->lock with interrupts disabled. */ static bool rcu_advance_cbs(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) { int i, j; /* If the CPU has no callbacks, nothing to do. */ if (!rdp->nxttail[RCU_NEXT_TAIL] || !*rdp->nxttail[RCU_DONE_TAIL]) return false; /* * Find all callbacks whose ->completed numbers indicate that they * are ready to invoke, and put them into the RCU_DONE_TAIL sublist. */ for (i = RCU_WAIT_TAIL; i < RCU_NEXT_TAIL; i++) { if (ULONG_CMP_LT(rnp->completed, rdp->nxtcompleted[i])) break; rdp->nxttail[RCU_DONE_TAIL] = rdp->nxttail[i]; } /* Clean up any sublist tail pointers that were misordered above. */ for (j = RCU_WAIT_TAIL; j < i; j++) rdp->nxttail[j] = rdp->nxttail[RCU_DONE_TAIL]; /* Copy down callbacks to fill in empty sublists. */ for (j = RCU_WAIT_TAIL; i < RCU_NEXT_TAIL; i++, j++) { if (rdp->nxttail[j] == rdp->nxttail[RCU_NEXT_TAIL]) break; rdp->nxttail[j] = rdp->nxttail[i]; rdp->nxtcompleted[j] = rdp->nxtcompleted[i]; } /* Classify any remaining callbacks. */ return rcu_accelerate_cbs(rsp, rnp, rdp); } /* * Update CPU-local rcu_data state to record the beginnings and ends of * grace periods. The caller must hold the ->lock of the leaf rcu_node * structure corresponding to the current CPU, and must have irqs disabled. * Returns true if the grace-period kthread needs to be awakened. */ static bool __note_gp_changes(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) { bool ret; /* Handle the ends of any preceding grace periods first. */ if (rdp->completed == rnp->completed && !unlikely(ACCESS_ONCE(rdp->gpwrap))) { /* No grace period end, so just accelerate recent callbacks. */ ret = rcu_accelerate_cbs(rsp, rnp, rdp); } else { /* Advance callbacks. */ ret = rcu_advance_cbs(rsp, rnp, rdp); /* Remember that we saw this grace-period completion. */ rdp->completed = rnp->completed; trace_rcu_grace_period(rsp->name, rdp->gpnum, TPS("cpuend")); } if (rdp->gpnum != rnp->gpnum || unlikely(ACCESS_ONCE(rdp->gpwrap))) { /* * If the current grace period is waiting for this CPU, * set up to detect a quiescent state, otherwise don't * go looking for one. */ rdp->gpnum = rnp->gpnum; trace_rcu_grace_period(rsp->name, rdp->gpnum, TPS("cpustart")); rdp->passed_quiesce = 0; rdp->rcu_qs_ctr_snap = __this_cpu_read(rcu_qs_ctr); rdp->qs_pending = !!(rnp->qsmask & rdp->grpmask); zero_cpu_stall_ticks(rdp); ACCESS_ONCE(rdp->gpwrap) = false; } return ret; } static void note_gp_changes(struct rcu_state *rsp, struct rcu_data *rdp) { unsigned long flags; bool needwake; struct rcu_node *rnp; local_irq_save(flags); rnp = rdp->mynode; if ((rdp->gpnum == ACCESS_ONCE(rnp->gpnum) && rdp->completed == ACCESS_ONCE(rnp->completed) && !unlikely(ACCESS_ONCE(rdp->gpwrap))) || /* w/out lock. */ !raw_spin_trylock(&rnp->lock)) { /* irqs already off, so later. */ local_irq_restore(flags); return; } smp_mb__after_unlock_lock(); needwake = __note_gp_changes(rsp, rnp, rdp); raw_spin_unlock_irqrestore(&rnp->lock, flags); if (needwake) rcu_gp_kthread_wake(rsp); } /* * Initialize a new grace period. Return 0 if no grace period required. */ static int rcu_gp_init(struct rcu_state *rsp) { struct rcu_data *rdp; struct rcu_node *rnp = rcu_get_root(rsp); ACCESS_ONCE(rsp->gp_activity) = jiffies; rcu_bind_gp_kthread(); raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); if (!ACCESS_ONCE(rsp->gp_flags)) { /* Spurious wakeup, tell caller to go back to sleep. */ raw_spin_unlock_irq(&rnp->lock); return 0; } ACCESS_ONCE(rsp->gp_flags) = 0; /* Clear all flags: New grace period. */ if (WARN_ON_ONCE(rcu_gp_in_progress(rsp))) { /* * Grace period already in progress, don't start another. * Not supposed to be able to happen. */ raw_spin_unlock_irq(&rnp->lock); return 0; } /* Advance to a new grace period and initialize state. */ record_gp_stall_check_time(rsp); /* Record GP times before starting GP, hence smp_store_release(). */ smp_store_release(&rsp->gpnum, rsp->gpnum + 1); trace_rcu_grace_period(rsp->name, rsp->gpnum, TPS("start")); raw_spin_unlock_irq(&rnp->lock); /* Exclude any concurrent CPU-hotplug operations. */ mutex_lock(&rsp->onoff_mutex); smp_mb__after_unlock_lock(); /* ->gpnum increment before GP! */ /* * 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, relying on the layout * of the tree within the rsp->node[] array. Note that other CPUs * will access only the leaves of the hierarchy, thus seeing 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. * * The grace period cannot complete until the initialization * process finishes, because this kthread handles both. */ rcu_for_each_node_breadth_first(rsp, rnp) { raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); rdp = this_cpu_ptr(rsp->rda); rcu_preempt_check_blocked_tasks(rnp); rnp->qsmask = rnp->qsmaskinit; ACCESS_ONCE(rnp->gpnum) = rsp->gpnum; WARN_ON_ONCE(rnp->completed != rsp->completed); ACCESS_ONCE(rnp->completed) = rsp->completed; if (rnp == rdp->mynode) (void)__note_gp_changes(rsp, rnp, rdp); rcu_preempt_boost_start_gp(rnp); trace_rcu_grace_period_init(rsp->name, rnp->gpnum, rnp->level, rnp->grplo, rnp->grphi, rnp->qsmask); raw_spin_unlock_irq(&rnp->lock); cond_resched_rcu_qs(); ACCESS_ONCE(rsp->gp_activity) = jiffies; } mutex_unlock(&rsp->onoff_mutex); return 1; } /* * Do one round of quiescent-state forcing. */ static int rcu_gp_fqs(struct rcu_state *rsp, int fqs_state_in) { int fqs_state = fqs_state_in; bool isidle = false; unsigned long maxj; struct rcu_node *rnp = rcu_get_root(rsp); ACCESS_ONCE(rsp->gp_activity) = jiffies; rsp->n_force_qs++; if (fqs_state == RCU_SAVE_DYNTICK) { /* Collect dyntick-idle snapshots. */ if (is_sysidle_rcu_state(rsp)) { isidle = true; maxj = jiffies - ULONG_MAX / 4; } force_qs_rnp(rsp, dyntick_save_progress_counter, &isidle, &maxj); rcu_sysidle_report_gp(rsp, isidle, maxj); fqs_state = RCU_FORCE_QS; } else { /* Handle dyntick-idle and offline CPUs. */ isidle = false; force_qs_rnp(rsp, rcu_implicit_dynticks_qs, &isidle, &maxj); } /* Clear flag to prevent immediate re-entry. */ if (ACCESS_ONCE(rsp->gp_flags) & RCU_GP_FLAG_FQS) { raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); ACCESS_ONCE(rsp->gp_flags) = ACCESS_ONCE(rsp->gp_flags) & ~RCU_GP_FLAG_FQS; raw_spin_unlock_irq(&rnp->lock); } return fqs_state; } /* * Clean up after the old grace period. */ static void rcu_gp_cleanup(struct rcu_state *rsp) { unsigned long gp_duration; bool needgp = false; int nocb = 0; struct rcu_data *rdp; struct rcu_node *rnp = rcu_get_root(rsp); ACCESS_ONCE(rsp->gp_activity) = jiffies; raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); gp_duration = jiffies - rsp->gp_start; if (gp_duration > rsp->gp_max) rsp->gp_max = gp_duration; /* * We know the grace period is complete, but to everyone else * it appears to still be ongoing. But it is also the case * that to everyone else it looks like there is nothing that * they can do to advance the grace period. It is therefore * safe for us to drop the lock in order to mark the grace * period as completed in all of the rcu_node structures. */ raw_spin_unlock_irq(&rnp->lock); /* * 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. This also avoids * some nasty RCU grace-period initialization races by forcing * the end of the current grace period to be completely recorded in * all of the rcu_node structures before the beginning of the next * grace period is recorded in any of the rcu_node structures. */ rcu_for_each_node_breadth_first(rsp, rnp) { raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); ACCESS_ONCE(rnp->completed) = rsp->gpnum; rdp = this_cpu_ptr(rsp->rda); if (rnp == rdp->mynode) needgp = __note_gp_changes(rsp, rnp, rdp) || needgp; /* smp_mb() provided by prior unlock-lock pair. */ nocb += rcu_future_gp_cleanup(rsp, rnp); raw_spin_unlock_irq(&rnp->lock); cond_resched_rcu_qs(); ACCESS_ONCE(rsp->gp_activity) = jiffies; } rnp = rcu_get_root(rsp); raw_spin_lock_irq(&rnp->lock); smp_mb__after_unlock_lock(); /* Order GP before ->completed update. */ rcu_nocb_gp_set(rnp, nocb); /* Declare grace period done. */ ACCESS_ONCE(rsp->completed) = rsp->gpnum; trace_rcu_grace_period(rsp->name, rsp->completed, TPS("end")); rsp->fqs_state = RCU_GP_IDLE; rdp = this_cpu_ptr(rsp->rda); /* Advance CBs to reduce false positives below. */ needgp = rcu_advance_cbs(rsp, rnp, rdp) || needgp; if (needgp || cpu_needs_another_gp(rsp, rdp)) { ACCESS_ONCE(rsp->gp_flags) = RCU_GP_FLAG_INIT; trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("newreq")); } raw_spin_unlock_irq(&rnp->lock); } /* * Body of kthread that handles grace periods. */ static int __noreturn rcu_gp_kthread(void *arg) { int fqs_state; int gf; unsigned long j; int ret; struct rcu_state *rsp = arg; struct rcu_node *rnp = rcu_get_root(rsp); for (;;) { /* Handle grace-period start. */ for (;;) { trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("reqwait")); rsp->gp_state = RCU_GP_WAIT_GPS; wait_event_interruptible(rsp->gp_wq, ACCESS_ONCE(rsp->gp_flags) & RCU_GP_FLAG_INIT); /* Locking provides needed memory barrier. */ if (rcu_gp_init(rsp)) break; cond_resched_rcu_qs(); ACCESS_ONCE(rsp->gp_activity) = jiffies; WARN_ON(signal_pending(current)); trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("reqwaitsig")); } /* Handle quiescent-state forcing. */ fqs_state = RCU_SAVE_DYNTICK; j = jiffies_till_first_fqs; if (j > HZ) { j = HZ; jiffies_till_first_fqs = HZ; } ret = 0; for (;;) { if (!ret) rsp->jiffies_force_qs = jiffies + j; trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("fqswait")); rsp->gp_state = RCU_GP_WAIT_FQS; ret = wait_event_interruptible_timeout(rsp->gp_wq, ((gf = ACCESS_ONCE(rsp->gp_flags)) & RCU_GP_FLAG_FQS) || (!ACCESS_ONCE(rnp->qsmask) && !rcu_preempt_blocked_readers_cgp(rnp)), j); /* Locking provides needed memory barriers. */ /* If grace period done, leave loop. */ if (!ACCESS_ONCE(rnp->qsmask) && !rcu_preempt_blocked_readers_cgp(rnp)) break; /* If time for quiescent-state forcing, do it. */ if (ULONG_CMP_GE(jiffies, rsp->jiffies_force_qs) || (gf & RCU_GP_FLAG_FQS)) { trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("fqsstart")); fqs_state = rcu_gp_fqs(rsp, fqs_state); trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("fqsend")); cond_resched_rcu_qs(); ACCESS_ONCE(rsp->gp_activity) = jiffies; } else { /* Deal with stray signal. */ cond_resched_rcu_qs(); ACCESS_ONCE(rsp->gp_activity) = jiffies; WARN_ON(signal_pending(current)); trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("fqswaitsig")); } j = jiffies_till_next_fqs; if (j > HZ) { j = HZ; jiffies_till_next_fqs = HZ; } else if (j < 1) { j = 1; jiffies_till_next_fqs = 1; } } /* Handle grace-period end. */ rcu_gp_cleanup(rsp); } } /* * 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 and hard irqs must be disabled. * * Note that it is legal for a dying CPU (which is marked as offline) to * invoke this function. This can happen when the dying CPU reports its * quiescent state. * * Returns true if the grace-period kthread must be awakened. */ static bool rcu_start_gp_advanced(struct rcu_state *rsp, struct rcu_node *rnp, struct rcu_data *rdp) { if (!rsp->gp_kthread || !cpu_needs_another_gp(rsp, rdp)) { /* * Either we have not yet spawned the grace-period * task, this CPU does not need another grace period, * or a grace period is already in progress. * Either way, don't start a new grace period. */ return false; } ACCESS_ONCE(rsp->gp_flags) = RCU_GP_FLAG_INIT; trace_rcu_grace_period(rsp->name, ACCESS_ONCE(rsp->gpnum), TPS("newreq")); /* * We can't do wakeups while holding the rnp->lock, as that * could cause possible deadlocks with the rq->lock. Defer * the wakeup to our caller. */ return true; } /* * Similar to rcu_start_gp_advanced(), but also advance the calling CPU's * callbacks. Note that rcu_start_gp_advanced() cannot do this because it * is invoked indirectly from rcu_advance_cbs(), which would result in * endless recursion -- or would do so if it wasn't for the self-deadlock * that is encountered beforehand. * * Returns true if the grace-period kthread needs to be awakened. */ static bool rcu_start_gp(struct rcu_state *rsp) { struct rcu_data *rdp = this_cpu_ptr(rsp->rda); struct rcu_node *rnp = rcu_get_root(rsp); bool ret = false; /* * If there is no grace period in progress right now, any * callbacks we have up to this point will be satisf