X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=kernel%2Fsched_fair.c;h=d3c03070872d9f335d3997e6e542ceb5006f2951;hb=907135aaa0cc120a347222c8f274ecc5ca0db641;hp=d8502ec569e4bc595c3f5d99ebb70cc0aa87155c;hpb=a03c9061d93822f66eb6287f8e9cf5833a12b49c;p=linux-2.6 diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index d8502ec569..d3c0307087 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -22,31 +22,34 @@ /* * Targeted preemption latency for CPU-bound tasks: - * (default: 20ms, units: nanoseconds) + * (default: 20ms * ilog(ncpus), units: nanoseconds) * * NOTE: this latency value is not the same as the concept of - * 'timeslice length' - timeslices in CFS are of variable length. - * (to see the precise effective timeslice length of your workload, - * run vmstat and monitor the context-switches field) + * 'timeslice length' - timeslices in CFS are of variable length + * and have no persistent notion like in traditional, time-slice + * based scheduling concepts. * - * On SMP systems the value of this is multiplied by the log2 of the - * number of CPUs. (i.e. factor 2x on 2-way systems, 3x on 4-way - * systems, 4x on 8-way systems, 5x on 16-way systems, etc.) - * Targeted preemption latency for CPU-bound tasks: + * (to see the precise effective timeslice length of your workload, + * run vmstat and monitor the context-switches (cs) field) */ -const_debug unsigned int sysctl_sched_latency = 20000000ULL; +unsigned int sysctl_sched_latency = 20000000ULL; /* - * After fork, child runs first. (default) If set to 0 then - * parent will (try to) run first. + * Minimal preemption granularity for CPU-bound tasks: + * (default: 1 msec * ilog(ncpus), units: nanoseconds) */ -const_debug unsigned int sysctl_sched_child_runs_first = 1; +unsigned int sysctl_sched_min_granularity = 1000000ULL; /* - * Minimal preemption granularity for CPU-bound tasks: - * (default: 2 msec, units: nanoseconds) + * is kept at sysctl_sched_latency / sysctl_sched_min_granularity + */ +unsigned int sched_nr_latency = 20; + +/* + * After fork, child runs first. (default) If set to 0 then + * parent will (try to) run first. */ -const_debug unsigned int sysctl_sched_nr_latency = 20; +const_debug unsigned int sysctl_sched_child_runs_first = 1; /* * sys_sched_yield() compat mode @@ -58,23 +61,25 @@ unsigned int __read_mostly sysctl_sched_compat_yield; /* * SCHED_BATCH wake-up granularity. - * (default: 10 msec, units: nanoseconds) + * (default: 10 msec * ilog(ncpus), units: nanoseconds) * * This option delays the preemption effects of decoupled workloads * and reduces their over-scheduling. Synchronous workloads will still * have immediate wakeup/sleep latencies. */ -const_debug unsigned int sysctl_sched_batch_wakeup_granularity = 10000000UL; +unsigned int sysctl_sched_batch_wakeup_granularity = 10000000UL; /* * SCHED_OTHER wake-up granularity. - * (default: 10 msec, units: nanoseconds) + * (default: 10 msec * ilog(ncpus), units: nanoseconds) * * This option delays the preemption effects of decoupled workloads * and reduces their over-scheduling. Synchronous workloads will still * have immediate wakeup/sleep latencies. */ -const_debug unsigned int sysctl_sched_wakeup_granularity = 10000000UL; +unsigned int sysctl_sched_wakeup_granularity = 10000000UL; + +const_debug unsigned int sysctl_sched_migration_cost = 500000UL; /************************************************************** * CFS operations on generic schedulable entities: @@ -112,8 +117,7 @@ static inline struct task_struct *task_of(struct sched_entity *se) * Scheduling class tree data structure manipulation methods: */ -static inline u64 -max_vruntime(u64 min_vruntime, u64 vruntime) +static inline u64 max_vruntime(u64 min_vruntime, u64 vruntime) { s64 delta = (s64)(vruntime - min_vruntime); if (delta > 0) @@ -122,8 +126,7 @@ max_vruntime(u64 min_vruntime, u64 vruntime) return min_vruntime; } -static inline u64 -min_vruntime(u64 min_vruntime, u64 vruntime) +static inline u64 min_vruntime(u64 min_vruntime, u64 vruntime) { s64 delta = (s64)(vruntime - min_vruntime); if (delta < 0) @@ -132,8 +135,7 @@ min_vruntime(u64 min_vruntime, u64 vruntime) return min_vruntime; } -static inline s64 -entity_key(struct cfs_rq *cfs_rq, struct sched_entity *se) +static inline s64 entity_key(struct cfs_rq *cfs_rq, struct sched_entity *se) { return se->vruntime - cfs_rq->min_vruntime; } @@ -141,8 +143,7 @@ entity_key(struct cfs_rq *cfs_rq, struct sched_entity *se) /* * Enqueue an entity into the rb-tree: */ -static void -__enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) +static void __enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) { struct rb_node **link = &cfs_rq->tasks_timeline.rb_node; struct rb_node *parent = NULL; @@ -179,8 +180,7 @@ __enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) rb_insert_color(&se->run_node, &cfs_rq->tasks_timeline); } -static void -__dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) +static void __dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) { if (cfs_rq->rb_leftmost == &se->run_node) cfs_rq->rb_leftmost = rb_next(&se->run_node); @@ -217,6 +217,22 @@ static inline struct sched_entity *__pick_last_entity(struct cfs_rq *cfs_rq) * Scheduling class statistics methods: */ +#ifdef CONFIG_SCHED_DEBUG +int sched_nr_latency_handler(struct ctl_table *table, int write, + struct file *filp, void __user *buffer, size_t *lenp, + loff_t *ppos) +{ + int ret = proc_dointvec_minmax(table, write, filp, buffer, lenp, ppos); + + if (ret || !write) + return ret; + + sched_nr_latency = DIV_ROUND_UP(sysctl_sched_latency, + sysctl_sched_min_granularity); + + return 0; +} +#endif /* * The idea is to set a period in which each task runs once. @@ -229,7 +245,7 @@ static inline struct sched_entity *__pick_last_entity(struct cfs_rq *cfs_rq) static u64 __sched_period(unsigned long nr_running) { u64 period = sysctl_sched_latency; - unsigned long nr_latency = sysctl_sched_nr_latency; + unsigned long nr_latency = sched_nr_latency; if (unlikely(nr_running > nr_latency)) { period *= nr_running; @@ -264,6 +280,7 @@ static u64 __sched_vslice(unsigned long rq_weight, unsigned long nr_running) { u64 vslice = __sched_period(nr_running); + vslice *= NICE_0_LOAD; do_div(vslice, rq_weight); return vslice; @@ -366,7 +383,6 @@ update_stats_wait_end(struct cfs_rq *cfs_rq, struct sched_entity *se) static inline void update_stats_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) { - update_curr(cfs_rq); /* * Mark the end of the wait period if dequeueing a * waiting task: @@ -387,15 +403,6 @@ update_stats_curr_start(struct cfs_rq *cfs_rq, struct sched_entity *se) se->exec_start = rq_of(cfs_rq)->clock; } -/* - * We are descheduling a task - update its stats: - */ -static inline void -update_stats_curr_end(struct cfs_rq *cfs_rq, struct sched_entity *se) -{ - se->exec_start = 0; -} - /************************************************** * Scheduling class queueing methods: */ @@ -478,7 +485,7 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) vruntime = cfs_rq->min_vruntime; - if (sched_feat(USE_TREE_AVG)) { + if (sched_feat(TREE_AVG)) { struct sched_entity *last = __pick_last_entity(cfs_rq); if (last) { vruntime += last->vruntime; @@ -487,25 +494,33 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) } else if (sched_feat(APPROX_AVG) && cfs_rq->nr_running) vruntime += sched_vslice(cfs_rq)/2; + /* + * The 'current' period is already promised to the current tasks, + * however the extra weight of the new task will slow them down a + * little, place the new task so that it fits in the slot that + * stays open at the end. + */ if (initial && sched_feat(START_DEBIT)) vruntime += sched_vslice_add(cfs_rq, se); if (!initial) { - if (sched_feat(NEW_FAIR_SLEEPERS)) + /* sleeps upto a single latency don't count. */ + if (sched_feat(NEW_FAIR_SLEEPERS) && entity_is_task(se) && + task_of(se)->policy != SCHED_BATCH) vruntime -= sysctl_sched_latency; - vruntime = max_t(s64, vruntime, se->vruntime); + /* ensure we never gain time by being placed backwards. */ + vruntime = max_vruntime(se->vruntime, vruntime); } se->vruntime = vruntime; - } static void enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int wakeup) { /* - * Update the fair clock. + * Update run-time statistics of the 'current'. */ update_curr(cfs_rq); @@ -524,6 +539,11 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int wakeup) static void dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep) { + /* + * Update run-time statistics of the 'current'. + */ + update_curr(cfs_rq); + update_stats_dequeue(cfs_rq, se); if (sleep) { #ifdef CONFIG_SCHEDSTATS @@ -608,8 +628,6 @@ static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev) if (prev->on_rq) update_curr(cfs_rq); - update_stats_curr_end(cfs_rq, prev); - check_spread(cfs_rq, prev); if (prev->on_rq) { update_stats_wait_start(cfs_rq, prev); @@ -626,7 +644,7 @@ static void entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr) */ update_curr(cfs_rq); - if (cfs_rq->nr_running > 1) + if (cfs_rq->nr_running > 1 || !sched_feat(WAKEUP_PREEMPT)) check_preempt_tick(cfs_rq, curr); } @@ -787,8 +805,7 @@ static void yield_task_fair(struct rq *rq) if (likely(!sysctl_sched_compat_yield)) { __update_rq_clock(rq); /* - * Dequeue and enqueue the task to update its - * position within the tree: + * Update run-time statistics of the 'current'. */ update_curr(cfs_rq); @@ -820,7 +837,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p) struct task_struct *curr = rq->curr; struct cfs_rq *cfs_rq = task_cfs_rq(curr); struct sched_entity *se = &curr->se, *pse = &p->se; - s64 delta; + unsigned long gran; if (unlikely(rt_prio(p->prio))) { update_rq_clock(rq); @@ -828,15 +845,26 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p) resched_task(curr); return; } + /* + * Batch tasks do not preempt (their preemption is driven by + * the tick): + */ + if (unlikely(p->policy == SCHED_BATCH)) + return; + + if (!sched_feat(WAKEUP_PREEMPT)) + return; while (!is_same_group(se, pse)) { se = parent_entity(se); pse = parent_entity(pse); } - delta = se->vruntime - pse->vruntime; + gran = sysctl_sched_wakeup_granularity; + if (unlikely(se->load.weight != NICE_0_LOAD)) + gran = calc_delta_fair(gran, &se->load); - if (delta > (s64)sysctl_sched_wakeup_granularity) + if (pse->vruntime + gran < se->vruntime) resched_task(curr); } @@ -870,6 +898,7 @@ static void put_prev_task_fair(struct rq *rq, struct task_struct *prev) } } +#ifdef CONFIG_SMP /************************************************** * Fair scheduling class load-balancing methods: */ @@ -930,12 +959,11 @@ static int cfs_rq_best_prio(struct cfs_rq *cfs_rq) static unsigned long load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, - unsigned long max_nr_move, unsigned long max_load_move, + unsigned long max_load_move, struct sched_domain *sd, enum cpu_idle_type idle, int *all_pinned, int *this_best_prio) { struct cfs_rq *busy_cfs_rq; - unsigned long load_moved, total_nr_moved = 0, nr_moved; long rem_load_move = max_load_move; struct rq_iterator cfs_rq_iterator; @@ -963,25 +991,48 @@ load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, #else # define maxload rem_load_move #endif - /* pass busy_cfs_rq argument into + /* + * pass busy_cfs_rq argument into * load_balance_[start|next]_fair iterators */ cfs_rq_iterator.arg = busy_cfs_rq; - nr_moved = balance_tasks(this_rq, this_cpu, busiest, - max_nr_move, maxload, sd, idle, all_pinned, - &load_moved, this_best_prio, &cfs_rq_iterator); + rem_load_move -= balance_tasks(this_rq, this_cpu, busiest, + maxload, sd, idle, all_pinned, + this_best_prio, + &cfs_rq_iterator); - total_nr_moved += nr_moved; - max_nr_move -= nr_moved; - rem_load_move -= load_moved; - - if (max_nr_move <= 0 || rem_load_move <= 0) + if (rem_load_move <= 0) break; } return max_load_move - rem_load_move; } +static int +move_one_task_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, + struct sched_domain *sd, enum cpu_idle_type idle) +{ + struct cfs_rq *busy_cfs_rq; + struct rq_iterator cfs_rq_iterator; + + cfs_rq_iterator.start = load_balance_start_fair; + cfs_rq_iterator.next = load_balance_next_fair; + + for_each_leaf_cfs_rq(busiest, busy_cfs_rq) { + /* + * pass busy_cfs_rq argument into + * load_balance_[start|next]_fair iterators + */ + cfs_rq_iterator.arg = busy_cfs_rq; + if (iter_move_one_task(this_rq, this_cpu, busiest, sd, idle, + &cfs_rq_iterator)) + return 1; + } + + return 0; +} +#endif + /* * scheduler tick hitting a task of our scheduling class: */ @@ -996,7 +1047,7 @@ static void task_tick_fair(struct rq *rq, struct task_struct *curr) } } -#define swap(a,b) do { typeof(a) tmp = (a); (a) = (b); (b) = tmp; } while (0) +#define swap(a, b) do { typeof(a) tmp = (a); (a) = (b); (b) = tmp; } while (0) /* * Share the fairness runtime between parent and child, thus the @@ -1009,14 +1060,16 @@ static void task_new_fair(struct rq *rq, struct task_struct *p) { struct cfs_rq *cfs_rq = task_cfs_rq(p); struct sched_entity *se = &p->se, *curr = cfs_rq->curr; + int this_cpu = smp_processor_id(); sched_info_queued(p); update_curr(cfs_rq); place_entity(cfs_rq, se, 1); - if (sysctl_sched_child_runs_first && - curr->vruntime < se->vruntime) { + /* 'curr' will be NULL if the child belongs to a different group */ + if (sysctl_sched_child_runs_first && this_cpu == task_cpu(p) && + curr && curr->vruntime < se->vruntime) { /* * Upon rescheduling, sched_class::put_prev_task() will place * 'current' within the tree based on its new key value. @@ -1024,11 +1077,7 @@ static void task_new_fair(struct rq *rq, struct task_struct *p) swap(curr->vruntime, se->vruntime); } - update_stats_enqueue(cfs_rq, se); - check_spread(cfs_rq, se); - check_spread(cfs_rq, curr); - __enqueue_entity(cfs_rq, se); - account_entity_enqueue(cfs_rq, se); + enqueue_task_fair(rq, p, 0); resched_task(rq->curr); } @@ -1059,7 +1108,10 @@ static const struct sched_class fair_sched_class = { .pick_next_task = pick_next_task_fair, .put_prev_task = put_prev_task_fair, +#ifdef CONFIG_SMP .load_balance = load_balance_fair, + .move_one_task = move_one_task_fair, +#endif .set_curr_task = set_curr_task_fair, .task_tick = task_tick_fair,