X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=kernel%2Fsched.c;h=a2be2d05529910594e24d19b2c1f1937df6a0ef4;hb=73d3ec5abad3f1730ac8530899d2c14d92f3ad63;hp=788ecce1e0e4d98701432bc6575cab35c3db3ad5;hpb=858119e159384308a5dde67776691a2ebf70df0f;p=safe%2Fjmp%2Flinux-2.6 diff --git a/kernel/sched.c b/kernel/sched.c index 788ecce..a2be2d0 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include @@ -49,6 +50,8 @@ #include #include #include +#include +#include #include #include @@ -144,7 +147,8 @@ (v1) * (v2_max) / (v1_max) #define DELTA(p) \ - (SCALE(TASK_NICE(p), 40, MAX_BONUS) + INTERACTIVE_DELTA) + (SCALE(TASK_NICE(p) + 20, 40, MAX_BONUS) - 20 * MAX_BONUS / 40 + \ + INTERACTIVE_DELTA) #define TASK_INTERACTIVE(p) \ ((p)->prio <= (p)->static_prio - DELTA(p)) @@ -166,36 +170,28 @@ */ #define SCALE_PRIO(x, prio) \ - max(x * (MAX_PRIO - prio) / (MAX_USER_PRIO/2), MIN_TIMESLICE) + max(x * (MAX_PRIO - prio) / (MAX_USER_PRIO / 2), MIN_TIMESLICE) -static unsigned int task_timeslice(task_t *p) +static unsigned int static_prio_timeslice(int static_prio) { - if (p->static_prio < NICE_TO_PRIO(0)) - return SCALE_PRIO(DEF_TIMESLICE*4, p->static_prio); + if (static_prio < NICE_TO_PRIO(0)) + return SCALE_PRIO(DEF_TIMESLICE * 4, static_prio); else - return SCALE_PRIO(DEF_TIMESLICE, p->static_prio); + return SCALE_PRIO(DEF_TIMESLICE, static_prio); } -#define task_hot(p, now, sd) ((long long) ((now) - (p)->last_ran) \ - < (long long) (sd)->cache_hot_time) -void __put_task_struct_cb(struct rcu_head *rhp) +static inline unsigned int task_timeslice(struct task_struct *p) { - __put_task_struct(container_of(rhp, struct task_struct, rcu)); + return static_prio_timeslice(p->static_prio); } -EXPORT_SYMBOL_GPL(__put_task_struct_cb); - /* * These are the runqueue data structures: */ -#define BITMAP_SIZE ((((MAX_PRIO+1+7)/8)+sizeof(long)-1)/sizeof(long)) - -typedef struct runqueue runqueue_t; - struct prio_array { unsigned int nr_active; - unsigned long bitmap[BITMAP_SIZE]; + DECLARE_BITMAP(bitmap, MAX_PRIO+1); /* include 1 bit for delimiter */ struct list_head queue[MAX_PRIO]; }; @@ -206,7 +202,7 @@ struct prio_array { * (such as the load balancing or the thread migration code), lock * acquire operations must be ordered by ascending &runqueue. */ -struct runqueue { +struct rq { spinlock_t lock; /* @@ -214,8 +210,8 @@ struct runqueue { * remote CPUs use both these fields when doing load calculation. */ unsigned long nr_running; + unsigned long raw_weighted_load; #ifdef CONFIG_SMP - unsigned long prio_bias; unsigned long cpu_load[3]; #endif unsigned long long nr_switches; @@ -230,9 +226,9 @@ struct runqueue { unsigned long expired_timestamp; unsigned long long timestamp_last_tick; - task_t *curr, *idle; + struct task_struct *curr, *idle; struct mm_struct *prev_mm; - prio_array_t *active, *expired, arrays[2]; + struct prio_array *active, *expired, arrays[2]; int best_expired_prio; atomic_t nr_iowait; @@ -243,7 +239,7 @@ struct runqueue { int active_balance; int push_cpu; - task_t *migration_thread; + struct task_struct *migration_thread; struct list_head migration_queue; #endif @@ -266,9 +262,10 @@ struct runqueue { unsigned long ttwu_cnt; unsigned long ttwu_local; #endif + struct lock_class_key rq_lock_key; }; -static DEFINE_PER_CPU(struct runqueue, runqueues); +static DEFINE_PER_CPU(struct rq, runqueues); /* * The domain tree (rq->sd) is protected by RCU's quiescent state transition. @@ -277,8 +274,8 @@ static DEFINE_PER_CPU(struct runqueue, runqueues); * The domain tree of any CPU may only be accessed from within * preempt-disabled sections. */ -#define for_each_domain(cpu, domain) \ -for (domain = rcu_dereference(cpu_rq(cpu)->sd); domain; domain = domain->parent) +#define for_each_domain(cpu, __sd) \ + for (__sd = rcu_dereference(cpu_rq(cpu)->sd); __sd; __sd = __sd->parent) #define cpu_rq(cpu) (&per_cpu(runqueues, (cpu))) #define this_rq() (&__get_cpu_var(runqueues)) @@ -293,26 +290,33 @@ for (domain = rcu_dereference(cpu_rq(cpu)->sd); domain; domain = domain->parent) #endif #ifndef __ARCH_WANT_UNLOCKED_CTXSW -static inline int task_running(runqueue_t *rq, task_t *p) +static inline int task_running(struct rq *rq, struct task_struct *p) { return rq->curr == p; } -static inline void prepare_lock_switch(runqueue_t *rq, task_t *next) +static inline void prepare_lock_switch(struct rq *rq, struct task_struct *next) { } -static inline void finish_lock_switch(runqueue_t *rq, task_t *prev) +static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) { #ifdef CONFIG_DEBUG_SPINLOCK /* this is a valid case when another task releases the spinlock */ rq->lock.owner = current; #endif + /* + * If we are tracking spinlock dependencies then we have to + * fix up the runqueue lock - which gets 'carried over' from + * prev into current: + */ + spin_acquire(&rq->lock.dep_map, 0, 0, _THIS_IP_); + spin_unlock_irq(&rq->lock); } #else /* __ARCH_WANT_UNLOCKED_CTXSW */ -static inline int task_running(runqueue_t *rq, task_t *p) +static inline int task_running(struct rq *rq, struct task_struct *p) { #ifdef CONFIG_SMP return p->oncpu; @@ -321,7 +325,7 @@ static inline int task_running(runqueue_t *rq, task_t *p) #endif } -static inline void prepare_lock_switch(runqueue_t *rq, task_t *next) +static inline void prepare_lock_switch(struct rq *rq, struct task_struct *next) { #ifdef CONFIG_SMP /* @@ -338,7 +342,7 @@ static inline void prepare_lock_switch(runqueue_t *rq, task_t *next) #endif } -static inline void finish_lock_switch(runqueue_t *rq, task_t *prev) +static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) { #ifdef CONFIG_SMP /* @@ -356,14 +360,33 @@ static inline void finish_lock_switch(runqueue_t *rq, task_t *prev) #endif /* __ARCH_WANT_UNLOCKED_CTXSW */ /* + * __task_rq_lock - lock the runqueue a given task resides on. + * Must be called interrupts disabled. + */ +static inline struct rq *__task_rq_lock(struct task_struct *p) + __acquires(rq->lock) +{ + struct rq *rq; + +repeat_lock_task: + rq = task_rq(p); + spin_lock(&rq->lock); + if (unlikely(rq != task_rq(p))) { + spin_unlock(&rq->lock); + goto repeat_lock_task; + } + return rq; +} + +/* * task_rq_lock - lock the runqueue a given task resides on and disable * interrupts. Note the ordering: we can safely lookup the task_rq without * explicitly disabling preemption. */ -static inline runqueue_t *task_rq_lock(task_t *p, unsigned long *flags) +static struct rq *task_rq_lock(struct task_struct *p, unsigned long *flags) __acquires(rq->lock) { - struct runqueue *rq; + struct rq *rq; repeat_lock_task: local_irq_save(*flags); @@ -376,7 +399,13 @@ repeat_lock_task: return rq; } -static inline void task_rq_unlock(runqueue_t *rq, unsigned long *flags) +static inline void __task_rq_unlock(struct rq *rq) + __releases(rq->lock) +{ + spin_unlock(&rq->lock); +} + +static inline void task_rq_unlock(struct rq *rq, unsigned long *flags) __releases(rq->lock) { spin_unlock_irqrestore(&rq->lock, *flags); @@ -396,7 +425,7 @@ static int show_schedstat(struct seq_file *seq, void *v) seq_printf(seq, "version %d\n", SCHEDSTAT_VERSION); seq_printf(seq, "timestamp %lu\n", jiffies); for_each_online_cpu(cpu) { - runqueue_t *rq = cpu_rq(cpu); + struct rq *rq = cpu_rq(cpu); #ifdef CONFIG_SMP struct sched_domain *sd; int dcnt = 0; @@ -473,9 +502,36 @@ struct file_operations proc_schedstat_operations = { .release = single_release, }; +/* + * Expects runqueue lock to be held for atomicity of update + */ +static inline void +rq_sched_info_arrive(struct rq *rq, unsigned long delta_jiffies) +{ + if (rq) { + rq->rq_sched_info.run_delay += delta_jiffies; + rq->rq_sched_info.pcnt++; + } +} + +/* + * Expects runqueue lock to be held for atomicity of update + */ +static inline void +rq_sched_info_depart(struct rq *rq, unsigned long delta_jiffies) +{ + if (rq) + rq->rq_sched_info.cpu_time += delta_jiffies; +} # define schedstat_inc(rq, field) do { (rq)->field++; } while (0) # define schedstat_add(rq, field, amt) do { (rq)->field += (amt); } while (0) #else /* !CONFIG_SCHEDSTATS */ +static inline void +rq_sched_info_arrive(struct rq *rq, unsigned long delta_jiffies) +{} +static inline void +rq_sched_info_depart(struct rq *rq, unsigned long delta_jiffies) +{} # define schedstat_inc(rq, field) do { } while (0) # define schedstat_add(rq, field, amt) do { } while (0) #endif @@ -483,10 +539,10 @@ struct file_operations proc_schedstat_operations = { /* * rq_lock - lock a given runqueue and disable interrupts. */ -static inline runqueue_t *this_rq_lock(void) +static inline struct rq *this_rq_lock(void) __acquires(rq->lock) { - runqueue_t *rq; + struct rq *rq; local_irq_disable(); rq = this_rq(); @@ -495,7 +551,7 @@ static inline runqueue_t *this_rq_lock(void) return rq; } -#ifdef CONFIG_SCHEDSTATS +#if defined(CONFIG_SCHEDSTATS) || defined(CONFIG_TASK_DELAY_ACCT) /* * Called when a process is dequeued from the active array and given * the cpu. We should note that with the exception of interactive @@ -511,7 +567,7 @@ static inline runqueue_t *this_rq_lock(void) * long it was from the *first* time it was queued to the time that it * finally hit a cpu. */ -static inline void sched_info_dequeued(task_t *t) +static inline void sched_info_dequeued(struct task_struct *t) { t->sched_info.last_queued = 0; } @@ -521,23 +577,18 @@ static inline void sched_info_dequeued(task_t *t) * long it was waiting to run. We also note when it began so that we * can keep stats on how long its timeslice is. */ -static void sched_info_arrive(task_t *t) +static void sched_info_arrive(struct task_struct *t) { - unsigned long now = jiffies, diff = 0; - struct runqueue *rq = task_rq(t); + unsigned long now = jiffies, delta_jiffies = 0; if (t->sched_info.last_queued) - diff = now - t->sched_info.last_queued; + delta_jiffies = now - t->sched_info.last_queued; sched_info_dequeued(t); - t->sched_info.run_delay += diff; + t->sched_info.run_delay += delta_jiffies; t->sched_info.last_arrival = now; t->sched_info.pcnt++; - if (!rq) - return; - - rq->rq_sched_info.run_delay += diff; - rq->rq_sched_info.pcnt++; + rq_sched_info_arrive(task_rq(t), delta_jiffies); } /* @@ -555,25 +606,23 @@ static void sched_info_arrive(task_t *t) * the timestamp if it is already not set. It's assumed that * sched_info_dequeued() will clear that stamp when appropriate. */ -static inline void sched_info_queued(task_t *t) +static inline void sched_info_queued(struct task_struct *t) { - if (!t->sched_info.last_queued) - t->sched_info.last_queued = jiffies; + if (unlikely(sched_info_on())) + if (!t->sched_info.last_queued) + t->sched_info.last_queued = jiffies; } /* * Called when a process ceases being the active-running process, either * voluntarily or involuntarily. Now we can calculate how long we ran. */ -static inline void sched_info_depart(task_t *t) +static inline void sched_info_depart(struct task_struct *t) { - struct runqueue *rq = task_rq(t); - unsigned long diff = jiffies - t->sched_info.last_arrival; + unsigned long delta_jiffies = jiffies - t->sched_info.last_arrival; - t->sched_info.cpu_time += diff; - - if (rq) - rq->rq_sched_info.cpu_time += diff; + t->sched_info.cpu_time += delta_jiffies; + rq_sched_info_depart(task_rq(t), delta_jiffies); } /* @@ -581,9 +630,10 @@ static inline void sched_info_depart(task_t *t) * their time slice. (This may also be called when switching to or from * the idle task.) We are only called when prev != next. */ -static inline void sched_info_switch(task_t *prev, task_t *next) +static inline void +__sched_info_switch(struct task_struct *prev, struct task_struct *next) { - struct runqueue *rq = task_rq(prev); + struct rq *rq = task_rq(prev); /* * prev now departs the cpu. It's not interesting to record @@ -596,15 +646,21 @@ static inline void sched_info_switch(task_t *prev, task_t *next) if (next != rq->idle) sched_info_arrive(next); } +static inline void +sched_info_switch(struct task_struct *prev, struct task_struct *next) +{ + if (unlikely(sched_info_on())) + __sched_info_switch(prev, next); +} #else #define sched_info_queued(t) do { } while (0) #define sched_info_switch(t, next) do { } while (0) -#endif /* CONFIG_SCHEDSTATS */ +#endif /* CONFIG_SCHEDSTATS || CONFIG_TASK_DELAY_ACCT */ /* * Adding/removing a task to/from a priority array: */ -static void dequeue_task(struct task_struct *p, prio_array_t *array) +static void dequeue_task(struct task_struct *p, struct prio_array *array) { array->nr_active--; list_del(&p->run_list); @@ -612,7 +668,7 @@ static void dequeue_task(struct task_struct *p, prio_array_t *array) __clear_bit(p->prio, array->bitmap); } -static void enqueue_task(struct task_struct *p, prio_array_t *array) +static void enqueue_task(struct task_struct *p, struct prio_array *array) { sched_info_queued(p); list_add_tail(&p->run_list, array->queue + p->prio); @@ -625,12 +681,13 @@ static void enqueue_task(struct task_struct *p, prio_array_t *array) * Put task to the end of the run list without the overhead of dequeue * followed by enqueue. */ -static void requeue_task(struct task_struct *p, prio_array_t *array) +static void requeue_task(struct task_struct *p, struct prio_array *array) { list_move_tail(&p->run_list, array->queue + p->prio); } -static inline void enqueue_task_head(struct task_struct *p, prio_array_t *array) +static inline void +enqueue_task_head(struct task_struct *p, struct prio_array *array) { list_add(&p->run_list, array->queue + p->prio); __set_bit(p->prio, array->bitmap); @@ -639,7 +696,7 @@ static inline void enqueue_task_head(struct task_struct *p, prio_array_t *array) } /* - * effective_prio - return the priority that is based on the static + * __normal_prio - return the priority that is based on the static * priority but is modified by bonuses/penalties. * * We scale the actual sleep average [0 .... MAX_SLEEP_AVG] @@ -652,13 +709,11 @@ static inline void enqueue_task_head(struct task_struct *p, prio_array_t *array) * * Both properties are important to certain workloads. */ -static int effective_prio(task_t *p) + +static inline int __normal_prio(struct task_struct *p) { int bonus, prio; - if (rt_task(p)) - return p->prio; - bonus = CURRENT_BONUS(p) - MAX_BONUS / 2; prio = p->static_prio - bonus; @@ -669,124 +724,179 @@ static int effective_prio(task_t *p) return prio; } -#ifdef CONFIG_SMP -static inline void inc_prio_bias(runqueue_t *rq, int prio) -{ - rq->prio_bias += MAX_PRIO - prio; -} +/* + * To aid in avoiding the subversion of "niceness" due to uneven distribution + * of tasks with abnormal "nice" values across CPUs the contribution that + * each task makes to its run queue's load is weighted according to its + * scheduling class and "nice" value. For SCHED_NORMAL tasks this is just a + * scaled version of the new time slice allocation that they receive on time + * slice expiry etc. + */ -static inline void dec_prio_bias(runqueue_t *rq, int prio) -{ - rq->prio_bias -= MAX_PRIO - prio; -} +/* + * Assume: static_prio_timeslice(NICE_TO_PRIO(0)) == DEF_TIMESLICE + * If static_prio_timeslice() is ever changed to break this assumption then + * this code will need modification + */ +#define TIME_SLICE_NICE_ZERO DEF_TIMESLICE +#define LOAD_WEIGHT(lp) \ + (((lp) * SCHED_LOAD_SCALE) / TIME_SLICE_NICE_ZERO) +#define PRIO_TO_LOAD_WEIGHT(prio) \ + LOAD_WEIGHT(static_prio_timeslice(prio)) +#define RTPRIO_TO_LOAD_WEIGHT(rp) \ + (PRIO_TO_LOAD_WEIGHT(MAX_RT_PRIO) + LOAD_WEIGHT(rp)) -static inline void inc_nr_running(task_t *p, runqueue_t *rq) +static void set_load_weight(struct task_struct *p) { - rq->nr_running++; - if (rt_task(p)) { - if (p != rq->migration_thread) + if (has_rt_policy(p)) { +#ifdef CONFIG_SMP + if (p == task_rq(p)->migration_thread) /* - * The migration thread does the actual balancing. Do - * not bias by its priority as the ultra high priority - * will skew balancing adversely. + * The migration thread does the actual balancing. + * Giving its load any weight will skew balancing + * adversely. */ - inc_prio_bias(rq, p->prio); + p->load_weight = 0; + else +#endif + p->load_weight = RTPRIO_TO_LOAD_WEIGHT(p->rt_priority); } else - inc_prio_bias(rq, p->static_prio); + p->load_weight = PRIO_TO_LOAD_WEIGHT(p->static_prio); } -static inline void dec_nr_running(task_t *p, runqueue_t *rq) -{ - rq->nr_running--; - if (rt_task(p)) { - if (p != rq->migration_thread) - dec_prio_bias(rq, p->prio); - } else - dec_prio_bias(rq, p->static_prio); -} -#else -static inline void inc_prio_bias(runqueue_t *rq, int prio) +static inline void +inc_raw_weighted_load(struct rq *rq, const struct task_struct *p) { + rq->raw_weighted_load += p->load_weight; } -static inline void dec_prio_bias(runqueue_t *rq, int prio) +static inline void +dec_raw_weighted_load(struct rq *rq, const struct task_struct *p) { + rq->raw_weighted_load -= p->load_weight; } -static inline void inc_nr_running(task_t *p, runqueue_t *rq) +static inline void inc_nr_running(struct task_struct *p, struct rq *rq) { rq->nr_running++; + inc_raw_weighted_load(rq, p); } -static inline void dec_nr_running(task_t *p, runqueue_t *rq) +static inline void dec_nr_running(struct task_struct *p, struct rq *rq) { rq->nr_running--; + dec_raw_weighted_load(rq, p); +} + +/* + * Calculate the expected normal priority: i.e. priority + * without taking RT-inheritance into account. Might be + * boosted by interactivity modifiers. Changes upon fork, + * setprio syscalls, and whenever the interactivity + * estimator recalculates. + */ +static inline int normal_prio(struct task_struct *p) +{ + int prio; + + if (has_rt_policy(p)) + prio = MAX_RT_PRIO-1 - p->rt_priority; + else + prio = __normal_prio(p); + return prio; +} + +/* + * Calculate the current priority, i.e. the priority + * taken into account by the scheduler. This value might + * be boosted by RT tasks, or might be boosted by + * interactivity modifiers. Will be RT if the task got + * RT-boosted. If not then it returns p->normal_prio. + */ +static int effective_prio(struct task_struct *p) +{ + p->normal_prio = normal_prio(p); + /* + * If we are RT tasks or we were boosted to RT priority, + * keep the priority unchanged. Otherwise, update priority + * to the normal priority: + */ + if (!rt_prio(p->prio)) + return p->normal_prio; + return p->prio; } -#endif /* * __activate_task - move a task to the runqueue. */ -static inline void __activate_task(task_t *p, runqueue_t *rq) +static void __activate_task(struct task_struct *p, struct rq *rq) { - enqueue_task(p, rq->active); + struct prio_array *target = rq->active; + + if (batch_task(p)) + target = rq->expired; + enqueue_task(p, target); inc_nr_running(p, rq); } /* * __activate_idle_task - move idle task to the _front_ of runqueue. */ -static inline void __activate_idle_task(task_t *p, runqueue_t *rq) +static inline void __activate_idle_task(struct task_struct *p, struct rq *rq) { enqueue_task_head(p, rq->active); inc_nr_running(p, rq); } -static int recalc_task_prio(task_t *p, unsigned long long now) +/* + * Recalculate p->normal_prio and p->prio after having slept, + * updating the sleep-average too: + */ +static int recalc_task_prio(struct task_struct *p, unsigned long long now) { /* Caller must always ensure 'now >= p->timestamp' */ - unsigned long long __sleep_time = now - p->timestamp; - unsigned long sleep_time; + unsigned long sleep_time = now - p->timestamp; - if (unlikely(p->policy == SCHED_BATCH)) + if (batch_task(p)) sleep_time = 0; - else { - if (__sleep_time > NS_MAX_SLEEP_AVG) - sleep_time = NS_MAX_SLEEP_AVG; - else - sleep_time = (unsigned long)__sleep_time; - } if (likely(sleep_time > 0)) { /* - * User tasks that sleep a long time are categorised as - * idle and will get just interactive status to stay active & - * prevent them suddenly becoming cpu hogs and starving - * other processes. + * This ceiling is set to the lowest priority that would allow + * a task to be reinserted into the active array on timeslice + * completion. */ - if (p->mm && p->activated != -1 && - sleep_time > INTERACTIVE_SLEEP(p)) { - p->sleep_avg = JIFFIES_TO_NS(MAX_SLEEP_AVG - - DEF_TIMESLICE); - } else { + unsigned long ceiling = INTERACTIVE_SLEEP(p); + + if (p->mm && sleep_time > ceiling && p->sleep_avg < ceiling) { /* - * The lower the sleep avg a task has the more - * rapidly it will rise with sleep time. + * Prevents user tasks from achieving best priority + * with one single large enough sleep. */ - sleep_time *= (MAX_BONUS - CURRENT_BONUS(p)) ? : 1; - + p->sleep_avg = ceiling; + /* + * Using INTERACTIVE_SLEEP() as a ceiling places a + * nice(0) task 1ms sleep away from promotion, and + * gives it 700ms to round-robin with no chance of + * being demoted. This is more than generous, so + * mark this sleep as non-interactive to prevent the + * on-runqueue bonus logic from intervening should + * this task not receive cpu immediately. + */ + p->sleep_type = SLEEP_NONINTERACTIVE; + } else { /* * Tasks waking from uninterruptible sleep are * limited in their sleep_avg rise as they * are likely to be waiting on I/O */ - if (p->activated == -1 && p->mm) { - if (p->sleep_avg >= INTERACTIVE_SLEEP(p)) + if (p->sleep_type == SLEEP_NONINTERACTIVE && p->mm) { + if (p->sleep_avg >= ceiling) sleep_time = 0; else if (p->sleep_avg + sleep_time >= - INTERACTIVE_SLEEP(p)) { - p->sleep_avg = INTERACTIVE_SLEEP(p); - sleep_time = 0; + ceiling) { + p->sleep_avg = ceiling; + sleep_time = 0; } } @@ -800,9 +910,9 @@ static int recalc_task_prio(task_t *p, unsigned long long now) */ p->sleep_avg += sleep_time; - if (p->sleep_avg > NS_MAX_SLEEP_AVG) - p->sleep_avg = NS_MAX_SLEEP_AVG; } + if (p->sleep_avg > NS_MAX_SLEEP_AVG) + p->sleep_avg = NS_MAX_SLEEP_AVG; } return effective_prio(p); @@ -814,7 +924,7 @@ static int recalc_task_prio(task_t *p, unsigned long long now) * Update all the scheduling statistics stuff. (sleep average * calculation, priority modifiers, etc.) */ -static void activate_task(task_t *p, runqueue_t *rq, int local) +static void activate_task(struct task_struct *p, struct rq *rq, int local) { unsigned long long now; @@ -822,7 +932,7 @@ static void activate_task(task_t *p, runqueue_t *rq, int local) #ifdef CONFIG_SMP if (!local) { /* Compensate for drifting sched_clock */ - runqueue_t *this_rq = this_rq(); + struct rq *this_rq = this_rq(); now = (now - this_rq->timestamp_last_tick) + rq->timestamp_last_tick; } @@ -835,7 +945,7 @@ static void activate_task(task_t *p, runqueue_t *rq, int local) * This checks to make sure it's not an uninterruptible task * that is now waking up. */ - if (!p->activated) { + if (p->sleep_type == SLEEP_NORMAL) { /* * Tasks which were woken up by interrupts (ie. hw events) * are most likely of interactive nature. So we give them @@ -844,13 +954,13 @@ static void activate_task(task_t *p, runqueue_t *rq, int local) * on a CPU, first time around: */ if (in_interrupt()) - p->activated = 2; + p->sleep_type = SLEEP_INTERRUPTED; else { /* * Normal first-time wakeups get a credit too for * on-runqueue time, but it will be weighted down: */ - p->activated = 1; + p->sleep_type = SLEEP_INTERACTIVE; } } p->timestamp = now; @@ -861,7 +971,7 @@ static void activate_task(task_t *p, runqueue_t *rq, int local) /* * deactivate_task - remove a task from the runqueue. */ -static void deactivate_task(struct task_struct *p, runqueue_t *rq) +static void deactivate_task(struct task_struct *p, struct rq *rq) { dec_nr_running(p, rq); dequeue_task(p, p->array); @@ -876,7 +986,12 @@ static void deactivate_task(struct task_struct *p, runqueue_t *rq) * the target CPU. */ #ifdef CONFIG_SMP -static void resched_task(task_t *p) + +#ifndef tsk_is_polling +#define tsk_is_polling(t) test_tsk_thread_flag(t, TIF_POLLING_NRFLAG) +#endif + +static void resched_task(struct task_struct *p) { int cpu; @@ -891,13 +1006,13 @@ static void resched_task(task_t *p) if (cpu == smp_processor_id()) return; - /* NEED_RESCHED must be visible before we test POLLING_NRFLAG */ + /* NEED_RESCHED must be visible before we test polling */ smp_mb(); - if (!test_tsk_thread_flag(p, TIF_POLLING_NRFLAG)) + if (!tsk_is_polling(p)) smp_send_reschedule(cpu); } #else -static inline void resched_task(task_t *p) +static inline void resched_task(struct task_struct *p) { assert_spin_locked(&task_rq(p)->lock); set_tsk_need_resched(p); @@ -908,28 +1023,35 @@ static inline void resched_task(task_t *p) * task_curr - is this task currently executing on a CPU? * @p: the task in question. */ -inline int task_curr(const task_t *p) +inline int task_curr(const struct task_struct *p) { return cpu_curr(task_cpu(p)) == p; } +/* Used instead of source_load when we know the type == 0 */ +unsigned long weighted_cpuload(const int cpu) +{ + return cpu_rq(cpu)->raw_weighted_load; +} + #ifdef CONFIG_SMP -typedef struct { +struct migration_req { struct list_head list; - task_t *task; + struct task_struct *task; int dest_cpu; struct completion done; -} migration_req_t; +}; /* * The task's runqueue lock must be held. * Returns true if you have to wait for migration thread. */ -static int migrate_task(task_t *p, int dest_cpu, migration_req_t *req) +static int +migrate_task(struct task_struct *p, int dest_cpu, struct migration_req *req) { - runqueue_t *rq = task_rq(p); + struct rq *rq = task_rq(p); /* * If the task is not on a runqueue (and not running), then @@ -944,6 +1066,7 @@ static int migrate_task(task_t *p, int dest_cpu, migration_req_t *req) req->task = p; req->dest_cpu = dest_cpu; list_add(&req->list, &rq->migration_queue); + return 1; } @@ -956,10 +1079,10 @@ static int migrate_task(task_t *p, int dest_cpu, migration_req_t *req) * smp_call_function() if an IPI is sent by the same process we are * waiting to become inactive. */ -void wait_task_inactive(task_t *p) +void wait_task_inactive(struct task_struct *p) { unsigned long flags; - runqueue_t *rq; + struct rq *rq; int preempted; repeat: @@ -990,7 +1113,7 @@ repeat: * to another CPU then no harm is done and the purpose has been * achieved as well. */ -void kick_process(task_t *p) +void kick_process(struct task_struct *p) { int cpu; @@ -1002,66 +1125,45 @@ void kick_process(task_t *p) } /* - * Return a low guess at the load of a migration-source cpu. + * Return a low guess at the load of a migration-source cpu weighted + * according to the scheduling class and "nice" value. * * We want to under-estimate the load of migration sources, to * balance conservatively. */ -static unsigned long __source_load(int cpu, int type, enum idle_type idle) +static inline unsigned long source_load(int cpu, int type) { - runqueue_t *rq = cpu_rq(cpu); - unsigned long running = rq->nr_running; - unsigned long source_load, cpu_load = rq->cpu_load[type-1], - load_now = running * SCHED_LOAD_SCALE; + struct rq *rq = cpu_rq(cpu); if (type == 0) - source_load = load_now; - else - source_load = min(cpu_load, load_now); - - if (running > 1 || (idle == NOT_IDLE && running)) - /* - * If we are busy rebalancing the load is biased by - * priority to create 'nice' support across cpus. When - * idle rebalancing we should only bias the source_load if - * there is more than one task running on that queue to - * prevent idle rebalance from trying to pull tasks from a - * queue with only one running task. - */ - source_load = source_load * rq->prio_bias / running; + return rq->raw_weighted_load; - return source_load; -} - -static inline unsigned long source_load(int cpu, int type) -{ - return __source_load(cpu, type, NOT_IDLE); + return min(rq->cpu_load[type-1], rq->raw_weighted_load); } /* - * Return a high guess at the load of a migration-target cpu + * Return a high guess at the load of a migration-target cpu weighted + * according to the scheduling class and "nice" value. */ -static inline unsigned long __target_load(int cpu, int type, enum idle_type idle) +static inline unsigned long target_load(int cpu, int type) { - runqueue_t *rq = cpu_rq(cpu); - unsigned long running = rq->nr_running; - unsigned long target_load, cpu_load = rq->cpu_load[type-1], - load_now = running * SCHED_LOAD_SCALE; + struct rq *rq = cpu_rq(cpu); if (type == 0) - target_load = load_now; - else - target_load = max(cpu_load, load_now); + return rq->raw_weighted_load; - if (running > 1 || (idle == NOT_IDLE && running)) - target_load = target_load * rq->prio_bias / running; - - return target_load; + return max(rq->cpu_load[type-1], rq->raw_weighted_load); } -static inline unsigned long target_load(int cpu, int type) +/* + * Return the average load per task on the cpu's run queue + */ +static inline unsigned long cpu_avg_load_per_task(int cpu) { - return __target_load(cpu, type, NOT_IDLE); + struct rq *rq = cpu_rq(cpu); + unsigned long n = rq->nr_running; + + return n ? rq->raw_weighted_load / n : SCHED_LOAD_SCALE; } /* @@ -1134,7 +1236,7 @@ find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu) cpus_and(tmp, group->cpumask, p->cpus_allowed); for_each_cpu_mask(i, tmp) { - load = source_load(i, 0); + load = weighted_cpuload(i); if (load < min_load || (load == min_load && i == this_cpu)) { min_load = load; @@ -1161,9 +1263,15 @@ static int sched_balance_self(int cpu, int flag) struct task_struct *t = current; struct sched_domain *tmp, *sd = NULL; - for_each_domain(cpu, tmp) + for_each_domain(cpu, tmp) { + /* + * If power savings logic is enabled for a domain, stop there. + */ + if (tmp->flags & SD_POWERSAVINGS_BALANCE) + break; if (tmp->flags & flag) sd = tmp; + } while (sd) { cpumask_t span; @@ -1208,7 +1316,7 @@ nextlevel: * Returns the CPU we should wake onto. */ #if defined(ARCH_HAS_SCHED_WAKE_IDLE) -static int wake_idle(int cpu, task_t *p) +static int wake_idle(int cpu, struct task_struct *p) { cpumask_t tmp; struct sched_domain *sd; @@ -1231,7 +1339,7 @@ static int wake_idle(int cpu, task_t *p) return cpu; } #else -static inline int wake_idle(int cpu, task_t *p) +static inline int wake_idle(int cpu, struct task_struct *p) { return cpu; } @@ -1251,15 +1359,15 @@ static inline int wake_idle(int cpu, task_t *p) * * returns failure only if the task is already active. */ -static int try_to_wake_up(task_t *p, unsigned int state, int sync) +static int try_to_wake_up(struct task_struct *p, unsigned int state, int sync) { int cpu, this_cpu, success = 0; unsigned long flags; long old_state; - runqueue_t *rq; + struct rq *rq; #ifdef CONFIG_SMP - unsigned long load, this_load; struct sched_domain *sd, *this_sd = NULL; + unsigned long load, this_load; int new_cpu; #endif @@ -1294,9 +1402,6 @@ static int try_to_wake_up(task_t *p, unsigned int state, int sync) } } - if (p->last_waker_cpu != this_cpu) - goto out_set_cpu; - if (unlikely(!cpu_isset(this_cpu, p->cpus_allowed))) goto out_set_cpu; @@ -1316,17 +1421,19 @@ static int try_to_wake_up(task_t *p, unsigned int state, int sync) if (this_sd->flags & SD_WAKE_AFFINE) { unsigned long tl = this_load; + unsigned long tl_per_task = cpu_avg_load_per_task(this_cpu); + /* * If sync wakeup then subtract the (maximum possible) * effect of the currently running task from the load * of the current CPU: */ if (sync) - tl -= SCHED_LOAD_SCALE; + tl -= current->load_weight; if ((tl <= load && - tl + target_load(cpu, idx) <= SCHED_LOAD_SCALE) || - 100*(tl + SCHED_LOAD_SCALE) <= imbalance*load) { + tl + target_load(cpu, idx) <= tl_per_task) || + 100*(tl + p->load_weight) <= imbalance*load) { /* * This domain has SD_WAKE_AFFINE and * p is cache cold in this domain, and @@ -1367,8 +1474,6 @@ out_set_cpu: cpu = task_cpu(p); } - p->last_waker_cpu = this_cpu; - out_activate: #endif /* CONFIG_SMP */ if (old_state == TASK_UNINTERRUPTIBLE) { @@ -1377,19 +1482,19 @@ out_activate: * Tasks on involuntary sleep don't earn * sleep_avg beyond just interactive state. */ - p->activated = -1; - } + p->sleep_type = SLEEP_NONINTERACTIVE; + } else /* * Tasks that have marked their sleep as noninteractive get - * woken up without updating their sleep average. (i.e. their - * sleep is handled in a priority-neutral manner, no priority - * boost and no penalty.) + * woken up with their sleep average not weighted in an + * interactive way. */ - if (old_state & TASK_NONINTERACTIVE) - __activate_task(p, rq); - else - activate_task(p, rq, cpu == this_cpu); + if (old_state & TASK_NONINTERACTIVE) + p->sleep_type = SLEEP_NONINTERACTIVE; + + + activate_task(p, rq, cpu == this_cpu); /* * Sync wakeups (i.e. those types of wakeups where the waker * has indicated that it will leave the CPU in short order) @@ -1412,15 +1517,14 @@ out: return success; } -int fastcall wake_up_process(task_t *p) +int fastcall wake_up_process(struct task_struct *p) { return try_to_wake_up(p, TASK_STOPPED | TASK_TRACED | TASK_INTERRUPTIBLE | TASK_UNINTERRUPTIBLE, 0); } - EXPORT_SYMBOL(wake_up_process); -int fastcall wake_up_state(task_t *p, unsigned int state) +int fastcall wake_up_state(struct task_struct *p, unsigned int state) { return try_to_wake_up(p, state, 0); } @@ -1429,7 +1533,7 @@ int fastcall wake_up_state(task_t *p, unsigned int state) * Perform scheduler related setup for a newly forked process p. * p is forked by current. */ -void fastcall sched_fork(task_t *p, int clone_flags) +void fastcall sched_fork(struct task_struct *p, int clone_flags) { int cpu = get_cpu(); @@ -1445,17 +1549,21 @@ void fastcall sched_fork(task_t *p, int clone_flags) * event cannot wake it up and insert it on the runqueue either. */ p->state = TASK_RUNNING; + + /* + * Make sure we do not leak PI boosting priority to the child: + */ + p->prio = current->normal_prio; + INIT_LIST_HEAD(&p->run_list); p->array = NULL; -#ifdef CONFIG_SCHEDSTATS - memset(&p->sched_info, 0, sizeof(p->sched_info)); +#if defined(CONFIG_SCHEDSTATS) || defined(CONFIG_TASK_DELAY_ACCT) + if (unlikely(sched_info_on())) + memset(&p->sched_info, 0, sizeof(p->sched_info)); #endif -#if defined(CONFIG_SMP) - p->last_waker_cpu = cpu; -#if defined(__ARCH_WANT_UNLOCKED_CTXSW) +#if defined(CONFIG_SMP) && defined(__ARCH_WANT_UNLOCKED_CTXSW) p->oncpu = 0; #endif -#endif #ifdef CONFIG_PREEMPT /* Want to start with kernel preemption disabled. */ task_thread_info(p)->preempt_count = 1; @@ -1494,11 +1602,11 @@ void fastcall sched_fork(task_t *p, int clone_flags) * that must be done for every newly created context, then puts the task * on the runqueue and wakes it. */ -void fastcall wake_up_new_task(task_t *p, unsigned long clone_flags) +void fastcall wake_up_new_task(struct task_struct *p, unsigned long clone_flags) { + struct rq *rq, *this_rq; unsigned long flags; int this_cpu, cpu; - runqueue_t *rq, *this_rq; rq = task_rq_lock(p, &flags); BUG_ON(p->state != TASK_RUNNING); @@ -1527,6 +1635,7 @@ void fastcall wake_up_new_task(task_t *p, unsigned long clone_flags) __activate_task(p, rq); else { p->prio = current->prio; + p->normal_prio = current->normal_prio; list_add_tail(&p->run_list, ¤t->run_list); p->array = current->array; p->array->nr_active++; @@ -1577,10 +1686,10 @@ void fastcall wake_up_new_task(task_t *p, unsigned long clone_flags) * artificially, because any timeslice recovered here * was given away by the parent in the first place.) */ -void fastcall sched_exit(task_t *p) +void fastcall sched_exit(struct task_struct *p) { unsigned long flags; - runqueue_t *rq; + struct rq *rq; /* * If the child was a (relative-) CPU hog then decrease @@ -1611,7 +1720,7 @@ void fastcall sched_exit(task_t *p) * prepare_task_switch sets up locking and calls architecture specific * hooks. */ -static inline void prepare_task_switch(runqueue_t *rq, task_t *next) +static inline void prepare_task_switch(struct rq *rq, struct task_struct *next) { prepare_lock_switch(rq, next); prepare_arch_switch(next); @@ -1632,7 +1741,7 @@ static inline void prepare_task_switch(runqueue_t *rq, task_t *next) * with the lock held can cause deadlocks; see schedule() for * details.) */ -static inline void finish_task_switch(runqueue_t *rq, task_t *prev) +static inline void finish_task_switch(struct rq *rq, struct task_struct *prev) __releases(rq->lock) { struct mm_struct *mm = rq->prev_mm; @@ -1656,18 +1765,25 @@ static inline void finish_task_switch(runqueue_t *rq, task_t *prev) finish_lock_switch(rq, prev); if (mm) mmdrop(mm); - if (unlikely(prev_task_flags & PF_DEAD)) + if (unlikely(prev_task_flags & PF_DEAD)) { + /* + * Remove function-return probe instances associated with this + * task and put them back on the free list. + */ + kprobe_flush_task(prev); put_task_struct(prev); + } } /** * schedule_tail - first thing a freshly forked thread must call. * @prev: the thread we just switched away from. */ -asmlinkage void schedule_tail(task_t *prev) +asmlinkage void schedule_tail(struct task_struct *prev) __releases(rq->lock) { - runqueue_t *rq = this_rq(); + struct rq *rq = this_rq(); + finish_task_switch(rq, prev); #ifdef __ARCH_WANT_UNLOCKED_CTXSW /* In this case, finish_task_switch does not reenable preemption */ @@ -1681,8 +1797,9 @@ asmlinkage void schedule_tail(task_t *prev) * context_switch - switch to the new MM and the new * thread's register state. */ -static inline -task_t * context_switch(runqueue_t *rq, task_t *prev, task_t *next) +static inline struct task_struct * +context_switch(struct rq *rq, struct task_struct *prev, + struct task_struct *next) { struct mm_struct *mm = next->mm; struct mm_struct *oldmm = prev->active_mm; @@ -1699,6 +1816,15 @@ task_t * context_switch(runqueue_t *rq, task_t *prev, task_t *next) WARN_ON(rq->prev_mm); rq->prev_mm = oldmm; } + /* + * Since the runqueue lock will be released by the next + * task (which is an invalid locking op but in the case + * of the scheduler it's an obvious special-case), so we + * do an early lockdep release here: + */ +#ifndef __ARCH_WANT_UNLOCKED_CTXSW + spin_release(&rq->lock.dep_map, 1, _THIS_IP_); +#endif /* Here we just switch the register state and the stack. */ switch_to(prev, next, prev); @@ -1727,7 +1853,7 @@ unsigned long nr_uninterruptible(void) { unsigned long i, sum = 0; - for_each_cpu(i) + for_each_possible_cpu(i) sum += cpu_rq(i)->nr_uninterruptible; /* @@ -1742,9 +1868,10 @@ unsigned long nr_uninterruptible(void) unsigned long long nr_context_switches(void) { - unsigned long long i, sum = 0; + int i; + unsigned long long sum = 0; - for_each_cpu(i) + for_each_possible_cpu(i) sum += cpu_rq(i)->nr_switches; return sum; @@ -1754,21 +1881,45 @@ unsigned long nr_iowait(void) { unsigned long i, sum = 0; - for_each_cpu(i) + for_each_possible_cpu(i) sum += atomic_read(&cpu_rq(i)->nr_iowait); return sum; } +unsigned long nr_active(void) +{ + unsigned long i, running = 0, uninterruptible = 0; + + for_each_online_cpu(i) { + running += cpu_rq(i)->nr_running; + uninterruptible += cpu_rq(i)->nr_uninterruptible; + } + + if (unlikely((long)uninterruptible < 0)) + uninterruptible = 0; + + return running + uninterruptible; +} + #ifdef CONFIG_SMP /* + * Is this task likely cache-hot: + */ +static inline int +task_hot(struct task_struct *p, unsigned long long now, struct sched_domain *sd) +{ + return (long long)(now - p->last_ran) < (long long)sd->cache_hot_time; +} + +/* * double_rq_lock - safely lock two runqueues * * Note this does not disable interrupts like task_rq_lock, * you need to do so manually before calling. */ -static void double_rq_lock(runqueue_t *rq1, runqueue_t *rq2) +static void double_rq_lock(struct rq *rq1, struct rq *rq2) __acquires(rq1->lock) __acquires(rq2->lock) { @@ -1792,7 +1943,7 @@ static void double_rq_lock(runqueue_t *rq1, runqueue_t *rq2) * Note this does not restore interrupts like task_rq_unlock, * you need to do so manually after calling. */ -static void double_rq_unlock(runqueue_t *rq1, runqueue_t *rq2) +static void double_rq_unlock(struct rq *rq1, struct rq *rq2) __releases(rq1->lock) __releases(rq2->lock) { @@ -1806,7 +1957,7 @@ static void double_rq_unlock(runqueue_t *rq1, runqueue_t *rq2) /* * double_lock_balance - lock the busiest runqueue, this_rq is locked already. */ -static void double_lock_balance(runqueue_t *this_rq, runqueue_t *busiest) +static void double_lock_balance(struct rq *this_rq, struct rq *busiest) __releases(this_rq->lock) __acquires(busiest->lock) __acquires(this_rq->lock) @@ -1827,11 +1978,11 @@ static void double_lock_balance(runqueue_t *this_rq, runqueue_t *busiest) * allow dest_cpu, which will force the cpu onto dest_cpu. Then * the cpu_allowed mask is restored. */ -static void sched_migrate_task(task_t *p, int dest_cpu) +static void sched_migrate_task(struct task_struct *p, int dest_cpu) { - migration_req_t req; - runqueue_t *rq; + struct migration_req req; unsigned long flags; + struct rq *rq; rq = task_rq_lock(p, &flags); if (!cpu_isset(dest_cpu, p->cpus_allowed) @@ -1842,11 +1993,13 @@ static void sched_migrate_task(task_t *p, int dest_cpu) if (migrate_task(p, dest_cpu, &req)) { /* Need to wait for migration thread (might exit: take ref). */ struct task_struct *mt = rq->migration_thread; + get_task_struct(mt); task_rq_unlock(rq, &flags); wake_up_process(mt); put_task_struct(mt); wait_for_completion(&req.done); + return; } out: @@ -1870,9 +2023,9 @@ void sched_exec(void) * pull_task - move a task from a remote runqueue to the local runqueue. * Both runqueues must be locked. */ -static -void pull_task(runqueue_t *src_rq, prio_array_t *src_array, task_t *p, - runqueue_t *this_rq, prio_array_t *this_array, int this_cpu) +static void pull_task(struct rq *src_rq, struct prio_array *src_array, + struct task_struct *p, struct rq *this_rq, + struct prio_array *this_array, int this_cpu) { dequeue_task(p, src_array); dec_nr_running(p, src_rq); @@ -1893,7 +2046,7 @@ void pull_task(runqueue_t *src_rq, prio_array_t *src_array, task_t *p, * can_migrate_task - may task p from runqueue rq be migrated to this_cpu? */ static -int can_migrate_task(task_t *p, runqueue_t *rq, int this_cpu, +int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu, struct sched_domain *sd, enum idle_type idle, int *all_pinned) { @@ -1924,26 +2077,42 @@ int can_migrate_task(task_t *p, runqueue_t *rq, int this_cpu, return 1; } +#define rq_best_prio(rq) min((rq)->curr->prio, (rq)->best_expired_prio) + /* - * move_tasks tries to move up to max_nr_move tasks from busiest to this_rq, - * as part of a balancing operation within "domain". Returns the number of - * tasks moved. + * move_tasks tries to move up to max_nr_move tasks and max_load_move weighted + * load from busiest to this_rq, as part of a balancing operation within + * "domain". Returns the number of tasks moved. * * Called with both runqueues locked. */ -static int move_tasks(runqueue_t *this_rq, int this_cpu, runqueue_t *busiest, - unsigned long max_nr_move, struct sched_domain *sd, - enum idle_type idle, int *all_pinned) +static int move_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest, + unsigned long max_nr_move, unsigned long max_load_move, + struct sched_domain *sd, enum idle_type idle, + int *all_pinned) { - prio_array_t *array, *dst_array; + int idx, pulled = 0, pinned = 0, this_best_prio, best_prio, + best_prio_seen, skip_for_load; + struct prio_array *array, *dst_array; struct list_head *head, *curr; - int idx, pulled = 0, pinned = 0; - task_t *tmp; + struct task_struct *tmp; + long rem_load_move; - if (max_nr_move == 0) + if (max_nr_move == 0 || max_load_move == 0) goto out; + rem_load_move = max_load_move; pinned = 1; + this_best_prio = rq_best_prio(this_rq); + best_prio = rq_best_prio(busiest); + /* + * Enable handling of the case where there is more than one task + * with the best priority. If the current running task is one + * of those with prio==best_prio we know it won't be moved + * and therefore it's safe to override the skip (based on load) of + * any task we find with that prio. + */ + best_prio_seen = best_prio == busiest->curr->prio; /* * We first consider expired tasks. Those will likely not be @@ -1979,11 +2148,22 @@ skip_bitmap: head = array->queue + idx; curr = head->prev; skip_queue: - tmp = list_entry(curr, task_t, run_list); + tmp = list_entry(curr, struct task_struct, run_list); curr = curr->prev; - if (!can_migrate_task(tmp, busiest, this_cpu, sd, idle, &pinned)) { + /* + * To help distribute high priority tasks accross CPUs we don't + * skip a task if it will be the highest priority task (i.e. smallest + * prio value) on its new queue regardless of its load weight + */ + skip_for_load = tmp->load_weight > rem_load_move; + if (skip_for_load && idx < this_best_prio) + skip_for_load = !best_prio_seen && idx == best_prio; + if (skip_for_load || + !can_migrate_task(tmp, busiest, this_cpu, sd, idle, &pinned)) { + + best_prio_seen |= idx == best_prio; if (curr != head) goto skip_queue; idx++; @@ -1997,9 +2177,15 @@ skip_queue: pull_task(busiest, array, tmp, this_rq, dst_array, this_cpu); pulled++; + rem_load_move -= tmp->load_weight; - /* We only want to steal up to the prescribed number of tasks. */ - if (pulled < max_nr_move) { + /* + * We only want to steal up to the prescribed number of tasks + * and the prescribed amount of weighted load. + */ + if (pulled < max_nr_move && rem_load_move > 0) { + if (idx < this_best_prio) + this_best_prio = idx; if (curr != head) goto skip_queue; idx++; @@ -2020,8 +2206,8 @@ out: /* * find_busiest_group finds and returns the busiest CPU group within the - * domain. It calculates and returns the number of tasks which should be - * moved to restore balance via the imbalance parameter. + * domain. It calculates and returns the amount of weighted load which + * should be moved to restore balance via the imbalance parameter. */ static struct sched_group * find_busiest_group(struct sched_domain *sd, int this_cpu, @@ -2030,9 +2216,19 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, struct sched_group *busiest = NULL, *this = NULL, *group = sd->groups; unsigned long max_load, avg_load, total_load, this_load, total_pwr; unsigned long max_pull; + unsigned long busiest_load_per_task, busiest_nr_running; + unsigned long this_load_per_task, this_nr_running; int load_idx; +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) + int power_savings_balance = 1; + unsigned long leader_nr_running = 0, min_load_per_task = 0; + unsigned long min_nr_running = ULONG_MAX; + struct sched_group *group_min = NULL, *group_leader = NULL; +#endif max_load = this_load = total_load = total_pwr = 0; + busiest_load_per_task = busiest_nr_running = 0; + this_load_per_task = this_nr_running = 0; if (idle == NOT_IDLE) load_idx = sd->busy_idx; else if (idle == NEWLY_IDLE) @@ -2041,26 +2237,31 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, load_idx = sd->idle_idx; do { - unsigned long load; + unsigned long load, group_capacity; int local_group; int i; + unsigned long sum_nr_running, sum_weighted_load; local_group = cpu_isset(this_cpu, group->cpumask); /* Tally up the load of all CPUs in the group */ - avg_load = 0; + sum_weighted_load = sum_nr_running = avg_load = 0; for_each_cpu_mask(i, group->cpumask) { + struct rq *rq = cpu_rq(i); + if (*sd_idle && !idle_cpu(i)) *sd_idle = 0; /* Bias balancing toward cpus of our domain */ if (local_group) - load = __target_load(i, load_idx, idle); + load = target_load(i, load_idx); else - load = __source_load(i, load_idx, idle); + load = source_load(i, load_idx); avg_load += load; + sum_nr_running += rq->nr_running; + sum_weighted_load += rq->raw_weighted_load; } total_load += avg_load; @@ -2069,17 +2270,80 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, /* Adjust by relative CPU power of the group */ avg_load = (avg_load * SCHED_LOAD_SCALE) / group->cpu_power; + group_capacity = group->cpu_power / SCHED_LOAD_SCALE; + if (local_group) { this_load = avg_load; this = group; - } else if (avg_load > max_load) { + this_nr_running = sum_nr_running; + this_load_per_task = sum_weighted_load; + } else if (avg_load > max_load && + sum_nr_running > group_capacity) { max_load = avg_load; busiest = group; + busiest_nr_running = sum_nr_running; + busiest_load_per_task = sum_weighted_load; + } + +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) + /* + * Busy processors will not participate in power savings + * balance. + */ + if (idle == NOT_IDLE || !(sd->flags & SD_POWERSAVINGS_BALANCE)) + goto group_next; + + /* + * If the local group is idle or completely loaded + * no need to do power savings balance at this domain + */ + if (local_group && (this_nr_running >= group_capacity || + !this_nr_running)) + power_savings_balance = 0; + + /* + * If a group is already running at full capacity or idle, + * don't include that group in power savings calculations + */ + if (!power_savings_balance || sum_nr_running >= group_capacity + || !sum_nr_running) + goto group_next; + + /* + * Calculate the group which has the least non-idle load. + * This is the group from where we need to pick up the load + * for saving power + */ + if ((sum_nr_running < min_nr_running) || + (sum_nr_running == min_nr_running && + first_cpu(group->cpumask) < + first_cpu(group_min->cpumask))) { + group_min = group; + min_nr_running = sum_nr_running; + min_load_per_task = sum_weighted_load / + sum_nr_running; + } + + /* + * Calculate the group which is almost near its + * capacity but still has some space to pick up some load + * from other group and save more power + */ + if (sum_nr_running <= group_capacity - 1) { + if (sum_nr_running > leader_nr_running || + (sum_nr_running == leader_nr_running && + first_cpu(group->cpumask) > + first_cpu(group_leader->cpumask))) { + group_leader = group; + leader_nr_running = sum_nr_running; + } } +group_next: +#endif group = group->next; } while (group != sd->groups); - if (!busiest || this_load >= max_load || max_load <= SCHED_LOAD_SCALE) + if (!busiest || this_load >= max_load || busiest_nr_running == 0) goto out_balanced; avg_load = (SCHED_LOAD_SCALE * total_load) / total_pwr; @@ -2088,6 +2352,7 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, 100*max_load <= sd->imbalance_pct*this_load) goto out_balanced; + busiest_load_per_task /= busiest_nr_running; /* * We're trying to get all the cpus to the average_load, so we don't * want to push ourselves above the average load, nor do we wish to @@ -2099,21 +2364,49 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, * by pulling tasks to us. Be careful of negative numbers as they'll * appear as very large values with unsigned longs. */ + if (max_load <= busiest_load_per_task) + goto out_balanced; + + /* + * In the presence of smp nice balancing, certain scenarios can have + * max load less than avg load(as we skip the groups at or below + * its cpu_power, while calculating max_load..) + */ + if (max_load < avg_load) { + *imbalance = 0; + goto small_imbalance; + } /* Don't want to pull so many tasks that a group would go idle */ - max_pull = min(max_load - avg_load, max_load - SCHED_LOAD_SCALE); + max_pull = min(max_load - avg_load, max_load - busiest_load_per_task); /* How much load to actually move to equalise the imbalance */ *imbalance = min(max_pull * busiest->cpu_power, (avg_load - this_load) * this->cpu_power) / SCHED_LOAD_SCALE; - if (*imbalance < SCHED_LOAD_SCALE) { - unsigned long pwr_now = 0, pwr_move = 0; - unsigned long tmp; + /* + * if *imbalance is less than the average load per runnable task + * there is no gaurantee that any tasks will be moved so we'll have + * a think about bumping its value to force at least one task to be + * moved + */ + if (*imbalance < busiest_load_per_task) { + unsigned long tmp, pwr_now, pwr_move; + unsigned int imbn; + +small_imbalance: + pwr_move = pwr_now = 0; + imbn = 2; + if (this_nr_running) { + this_load_per_task /= this_nr_running; + if (busiest_load_per_task > this_load_per_task) + imbn = 1; + } else + this_load_per_task = SCHED_LOAD_SCALE; - if (max_load - this_load >= SCHED_LOAD_SCALE*2) { - *imbalance = 1; + if (max_load - this_load >= busiest_load_per_task * imbn) { + *imbalance = busiest_load_per_task; return busiest; } @@ -2123,39 +2416,47 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, * moving them. */ - pwr_now += busiest->cpu_power*min(SCHED_LOAD_SCALE, max_load); - pwr_now += this->cpu_power*min(SCHED_LOAD_SCALE, this_load); + pwr_now += busiest->cpu_power * + min(busiest_load_per_task, max_load); + pwr_now += this->cpu_power * + min(this_load_per_task, this_load); pwr_now /= SCHED_LOAD_SCALE; /* Amount of load we'd subtract */ - tmp = SCHED_LOAD_SCALE*SCHED_LOAD_SCALE/busiest->cpu_power; + tmp = busiest_load_per_task*SCHED_LOAD_SCALE/busiest->cpu_power; if (max_load > tmp) - pwr_move += busiest->cpu_power*min(SCHED_LOAD_SCALE, - max_load - tmp); + pwr_move += busiest->cpu_power * + min(busiest_load_per_task, max_load - tmp); /* Amount of load we'd add */ if (max_load*busiest->cpu_power < - SCHED_LOAD_SCALE*SCHED_LOAD_SCALE) + busiest_load_per_task*SCHED_LOAD_SCALE) tmp = max_load*busiest->cpu_power/this->cpu_power; else - tmp = SCHED_LOAD_SCALE*SCHED_LOAD_SCALE/this->cpu_power; - pwr_move += this->cpu_power*min(SCHED_LOAD_SCALE, this_load + tmp); + tmp = busiest_load_per_task*SCHED_LOAD_SCALE/this->cpu_power; + pwr_move += this->cpu_power*min(this_load_per_task, this_load + tmp); pwr_move /= SCHED_LOAD_SCALE; /* Move if we gain throughput */ if (pwr_move <= pwr_now) goto out_balanced; - *imbalance = 1; - return busiest; + *imbalance = busiest_load_per_task; } - /* Get rid of the scaling factor, rounding down as we divide */ - *imbalance = *imbalance / SCHED_LOAD_SCALE; return busiest; out_balanced: +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) + if (idle == NOT_IDLE || !(sd->flags & SD_POWERSAVINGS_BALANCE)) + goto ret; + if (this == group_leader && group_leader != group_min) { + *imbalance = min_load_per_task; + return group_min; + } +ret: +#endif *imbalance = 0; return NULL; } @@ -2163,19 +2464,23 @@ out_balanced: /* * find_busiest_queue - find the busiest runqueue among the cpus in group. */ -static runqueue_t *find_busiest_queue(struct sched_group *group, - enum idle_type idle) +static struct rq * +find_busiest_queue(struct sched_group *group, enum idle_type idle, + unsigned long imbalance) { - unsigned long load, max_load = 0; - runqueue_t *busiest = NULL; + struct rq *busiest = NULL, *rq; + unsigned long max_load = 0; int i; for_each_cpu_mask(i, group->cpumask) { - load = __source_load(i, 0, idle); + rq = cpu_rq(i); + + if (rq->nr_running == 1 && rq->raw_weighted_load > imbalance) + continue; - if (load > max_load) { - max_load = load; - busiest = cpu_rq(i); + if (rq->raw_weighted_load > max_load) { + max_load = rq->raw_weighted_load; + busiest = rq; } } @@ -2188,23 +2493,27 @@ static runqueue_t *find_busiest_queue(struct sched_group *group, */ #define MAX_PINNED_INTERVAL 512 +static inline unsigned long minus_1_or_zero(unsigned long n) +{ + return n > 0 ? n - 1 : 0; +} + /* * Check this_cpu to ensure it is balanced within domain. Attempt to move * tasks if there is an imbalance. * * Called with this_rq unlocked. */ -static int load_balance(int this_cpu, runqueue_t *this_rq, +static int load_balance(int this_cpu, struct rq *this_rq, struct sched_domain *sd, enum idle_type idle) { + int nr_moved, all_pinned = 0, active_balance = 0, sd_idle = 0; struct sched_group *group; - runqueue_t *busiest; unsigned long imbalance; - int nr_moved, all_pinned = 0; - int active_balance = 0; - int sd_idle = 0; + struct rq *busiest; - if (idle != NOT_IDLE && sd->flags & SD_SHARE_CPUPOWER) + if (idle != NOT_IDLE && sd->flags & SD_SHARE_CPUPOWER && + !sched_smt_power_savings) sd_idle = 1; schedstat_inc(sd, lb_cnt[idle]); @@ -2215,7 +2524,7 @@ static int load_balance(int this_cpu, runqueue_t *this_rq, goto out_balanced; } - busiest = find_busiest_queue(group, idle); + busiest = find_busiest_queue(group, idle, imbalance); if (!busiest) { schedstat_inc(sd, lb_nobusyq[idle]); goto out_balanced; @@ -2235,7 +2544,8 @@ static int load_balance(int this_cpu, runqueue_t *this_rq, */ double_rq_lock(this_rq, busiest); nr_moved = move_tasks(this_rq, this_cpu, busiest, - imbalance, sd, idle, &all_pinned); + minus_1_or_zero(busiest->nr_running), + imbalance, sd, idle, &all_pinned); double_rq_unlock(this_rq, busiest); /* All tasks on this runqueue were pinned by CPU affinity */ @@ -2292,7 +2602,8 @@ static int load_balance(int this_cpu, runqueue_t *this_rq, sd->balance_interval *= 2; } - if (!nr_moved && !sd_idle && sd->flags & SD_SHARE_CPUPOWER) + if (!nr_moved && !sd_idle && sd->flags & SD_SHARE_CPUPOWER && + !sched_smt_power_savings) return -1; return nr_moved; @@ -2307,7 +2618,8 @@ out_one_pinned: (sd->balance_interval < sd->max_interval)) sd->balance_interval *= 2; - if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER) + if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER && + !sched_smt_power_savings) return -1; return 0; } @@ -2319,16 +2631,16 @@ out_one_pinned: * Called from schedule when this_rq is about to become idle (NEWLY_IDLE). * this_rq is locked. */ -static int load_balance_newidle(int this_cpu, runqueue_t *this_rq, - struct sched_domain *sd) +static int +load_balance_newidle(int this_cpu, struct rq *this_rq, struct sched_domain *sd) { struct sched_group *group; - runqueue_t *busiest = NULL; + struct rq *busiest = NULL; unsigned long imbalance; int nr_moved = 0; int sd_idle = 0; - if (sd->flags & SD_SHARE_CPUPOWER) + if (sd->flags & SD_SHARE_CPUPOWER && !sched_smt_power_savings) sd_idle = 1; schedstat_inc(sd, lb_cnt[NEWLY_IDLE]); @@ -2338,7 +2650,7 @@ static int load_balance_newidle(int this_cpu, runqueue_t *this_rq, goto out_balanced; } - busiest = find_busiest_queue(group, NEWLY_IDLE); + busiest = find_busiest_queue(group, NEWLY_IDLE, imbalance); if (!busiest) { schedstat_inc(sd, lb_nobusyq[NEWLY_IDLE]); goto out_balanced; @@ -2353,6 +2665,7 @@ static int load_balance_newidle(int this_cpu, runqueue_t *this_rq, /* Attempt to move tasks */ double_lock_balance(this_rq, busiest); nr_moved = move_tasks(this_rq, this_cpu, busiest, + minus_1_or_zero(busiest->nr_running), imbalance, sd, NEWLY_IDLE, NULL); spin_unlock(&busiest->lock); } @@ -2368,9 +2681,11 @@ static int load_balance_newidle(int this_cpu, runqueue_t *this_rq, out_balanced: schedstat_inc(sd, lb_balanced[NEWLY_IDLE]); - if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER) + if (!sd_idle && sd->flags & SD_SHARE_CPUPOWER && + !sched_smt_power_savings) return -1; sd->nr_balance_failed = 0; + return 0; } @@ -2378,16 +2693,15 @@ out_balanced: * idle_balance is called by schedule() if this_cpu is about to become * idle. Attempts to pull tasks from other CPUs. */ -static void idle_balance(int this_cpu, runqueue_t *this_rq) +static void idle_balance(int this_cpu, struct rq *this_rq) { struct sched_domain *sd; for_each_domain(this_cpu, sd) { if (sd->flags & SD_BALANCE_NEWIDLE) { - if (load_balance_newidle(this_cpu, this_rq, sd)) { - /* We've pulled tasks over so stop searching */ + /* If we've pulled tasks over stop searching: */ + if (load_balance_newidle(this_cpu, this_rq, sd)) break; - } } } } @@ -2400,14 +2714,14 @@ static void idle_balance(int this_cpu, runqueue_t *this_rq) * * Called with busiest_rq locked. */ -static void active_load_balance(runqueue_t *busiest_rq, int busiest_cpu) +static void active_load_balance(struct rq *busiest_rq, int busiest_cpu) { - struct sched_domain *sd; - runqueue_t *target_rq; int target_cpu = busiest_rq->push_cpu; + struct sched_domain *sd; + struct rq *target_rq; + /* Is there any task to move? */ if (busiest_rq->nr_running <= 1) - /* no task to move */ return; target_rq = cpu_rq(target_cpu); @@ -2423,21 +2737,22 @@ static void active_load_balance(runqueue_t *busiest_rq, int busiest_cpu) double_lock_balance(busiest_rq, target_rq); /* Search for an sd spanning us and the target CPU. */ - for_each_domain(target_cpu, sd) + for_each_domain(target_cpu, sd) { if ((sd->flags & SD_LOAD_BALANCE) && - cpu_isset(busiest_cpu, sd->span)) + cpu_isset(busiest_cpu, sd->span)) break; + } - if (unlikely(sd == NULL)) - goto out; - - schedstat_inc(sd, alb_cnt); + if (likely(sd)) { + schedstat_inc(sd, alb_cnt); - if (move_tasks(target_rq, target_cpu, busiest_rq, 1, sd, SCHED_IDLE, NULL)) - schedstat_inc(sd, alb_pushed); - else - schedstat_inc(sd, alb_failed); -out: + if (move_tasks(target_rq, target_cpu, busiest_rq, 1, + RTPRIO_TO_LOAD_WEIGHT(100), sd, SCHED_IDLE, + NULL)) + schedstat_inc(sd, alb_pushed); + else + schedstat_inc(sd, alb_failed); + } spin_unlock(&target_rq->lock); } @@ -2450,23 +2765,27 @@ out: * Balancing parameters are set up in arch_init_sched_domains. */ -/* Don't have all balancing operations going off at once */ -#define CPU_OFFSET(cpu) (HZ * cpu / NR_CPUS) +/* Don't have all balancing operations going off at once: */ +static inline unsigned long cpu_offset(int cpu) +{ + return jiffies + cpu * HZ / NR_CPUS; +} -static void rebalance_tick(int this_cpu, runqueue_t *this_rq, - enum idle_type idle) +static void +rebalance_tick(int this_cpu, struct rq *this_rq, enum idle_type idle) { - unsigned long old_load, this_load; - unsigned long j = jiffies + CPU_OFFSET(this_cpu); + unsigned long this_load, interval, j = cpu_offset(this_cpu); struct sched_domain *sd; - int i; + int i, scale; + + this_load = this_rq->raw_weighted_load; + + /* Update our load: */ + for (i = 0, scale = 1; i < 3; i++, scale <<= 1) { + unsigned long old_load, new_load; - this_load = this_rq->nr_running * SCHED_LOAD_SCALE; - /* Update our load */ - for (i = 0; i < 3; i++) { - unsigned long new_load = this_load; - int scale = 1 << i; old_load = this_rq->cpu_load[i]; + new_load = this_load; /* * Round up the averaging division if load is increasing. This * prevents us from getting stuck on 9 if the load is 10, for @@ -2478,8 +2797,6 @@ static void rebalance_tick(int this_cpu, runqueue_t *this_rq, } for_each_domain(this_cpu, sd) { - unsigned long interval; - if (!(sd->flags & SD_LOAD_BALANCE)) continue; @@ -2509,17 +2826,18 @@ static void rebalance_tick(int this_cpu, runqueue_t *this_rq, /* * on UP we do not need to balance between CPUs: */ -static inline void rebalance_tick(int cpu, runqueue_t *rq, enum idle_type idle) +static inline void rebalance_tick(int cpu, struct rq *rq, enum idle_type idle) { } -static inline void idle_balance(int cpu, runqueue_t *rq) +static inline void idle_balance(int cpu, struct rq *rq) { } #endif -static inline int wake_priority_sleeper(runqueue_t *rq) +static inline int wake_priority_sleeper(struct rq *rq) { int ret = 0; + #ifdef CONFIG_SCHED_SMT spin_lock(&rq->lock); /* @@ -2543,25 +2861,26 @@ EXPORT_PER_CPU_SYMBOL(kstat); * This is called on clock ticks and on context switches. * Bank in p->sched_time the ns elapsed since the last tick or switch. */ -static inline void update_cpu_clock(task_t *p, runqueue_t *rq, - unsigned long long now) +static inline void +update_cpu_clock(struct task_struct *p, struct rq *rq, unsigned long long now) { - unsigned long long last = max(p->timestamp, rq->timestamp_last_tick); - p->sched_time += now - last; + p->sched_time += now - max(p->timestamp, rq->timestamp_last_tick); } /* * Return current->sched_time plus any more ns on the sched_clock * that have not yet been banked. */ -unsigned long long current_sched_time(const task_t *tsk) +unsigned long long current_sched_time(const struct task_struct *p) { unsigned long long ns; unsigned long flags; + local_irq_save(flags); - ns = max(tsk->timestamp, task_rq(tsk)->timestamp_last_tick); - ns = tsk->sched_time + (sched_clock() - ns); + ns = max(p->timestamp, task_rq(p)->timestamp_last_tick); + ns = p->sched_time + sched_clock() - ns; local_irq_restore(flags); + return ns; } @@ -2575,11 +2894,16 @@ unsigned long long current_sched_time(const task_t *tsk) * increasing number of running tasks. We also ignore the interactivity * if a better static_prio task has expired: */ -#define EXPIRED_STARVING(rq) \ - ((STARVATION_LIMIT && ((rq)->expired_timestamp && \ - (jiffies - (rq)->expired_timestamp >= \ - STARVATION_LIMIT * ((rq)->nr_running) + 1))) || \ - ((rq)->curr->static_prio > (rq)->best_expired_prio)) +static inline int expired_starving(struct rq *rq) +{ + if (rq->curr->static_prio > rq->best_expired_prio) + return 1; + if (!STARVATION_LIMIT || !rq->expired_timestamp) + return 0; + if (jiffies - rq->expired_timestamp > STARVATION_LIMIT * rq->nr_running) + return 1; + return 0; +} /* * Account user cpu time to a process. @@ -2612,7 +2936,7 @@ void account_system_time(struct task_struct *p, int hardirq_offset, cputime_t cputime) { struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; - runqueue_t *rq = this_rq(); + struct rq *rq = this_rq(); cputime64_t tmp; p->stime = cputime_add(p->stime, cputime); @@ -2642,7 +2966,7 @@ void account_steal_time(struct task_struct *p, cputime_t steal) { struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; cputime64_t tmp = cputime_to_cputime64(steal); - runqueue_t *rq = this_rq(); + struct rq *rq = this_rq(); if (p == rq->idle) { p->stime = cputime_add(p->stime, steal); @@ -2663,10 +2987,10 @@ void account_steal_time(struct task_struct *p, cputime_t steal) */ void scheduler_tick(void) { - int cpu = smp_processor_id(); - runqueue_t *rq = this_rq(); - task_t *p = current; unsigned long long now = sched_clock(); + struct task_struct *p = current; + int cpu = smp_processor_id(); + struct rq *rq = cpu_rq(cpu); update_cpu_clock(p, rq, now); @@ -2716,7 +3040,7 @@ void scheduler_tick(void) if (!rq->expired_timestamp) rq->expired_timestamp = jiffies; - if (!TASK_INTERACTIVE(p) || EXPIRED_STARVING(rq)) { + if (!TASK_INTERACTIVE(p) || expired_starving(rq)) { enqueue_task(p, rq->expired); if (p->static_prio < rq->best_expired_prio) rq->best_expired_prio = p->static_prio; @@ -2755,55 +3079,42 @@ out: } #ifdef CONFIG_SCHED_SMT -static inline void wakeup_busy_runqueue(runqueue_t *rq) +static inline void wakeup_busy_runqueue(struct rq *rq) { /* If an SMT runqueue is sleeping due to priority reasons wake it up */ if (rq->curr == rq->idle && rq->nr_running) resched_task(rq->idle); } -static void wake_sleeping_dependent(int this_cpu, runqueue_t *this_rq) +/* + * Called with interrupt disabled and this_rq's runqueue locked. + */ +static void wake_sleeping_dependent(int this_cpu) { struct sched_domain *tmp, *sd = NULL; - cpumask_t sibling_map; int i; - for_each_domain(this_cpu, tmp) - if (tmp->flags & SD_SHARE_CPUPOWER) + for_each_domain(this_cpu, tmp) { + if (tmp->flags & SD_SHARE_CPUPOWER) { sd = tmp; + break; + } + } if (!sd) return; - /* - * Unlock the current runqueue because we have to lock in - * CPU order to avoid deadlocks. Caller knows that we might - * unlock. We keep IRQs disabled. - */ - spin_unlock(&this_rq->lock); - - sibling_map = sd->span; + for_each_cpu_mask(i, sd->span) { + struct rq *smt_rq = cpu_rq(i); - for_each_cpu_mask(i, sibling_map) - spin_lock(&cpu_rq(i)->lock); - /* - * We clear this CPU from the mask. This both simplifies the - * inner loop and keps this_rq locked when we exit: - */ - cpu_clear(this_cpu, sibling_map); - - for_each_cpu_mask(i, sibling_map) { - runqueue_t *smt_rq = cpu_rq(i); + if (i == this_cpu) + continue; + if (unlikely(!spin_trylock(&smt_rq->lock))) + continue; wakeup_busy_runqueue(smt_rq); + spin_unlock(&smt_rq->lock); } - - for_each_cpu_mask(i, sibling_map) - spin_unlock(&cpu_rq(i)->lock); - /* - * We exit with this_cpu's rq still held and IRQs - * still disabled: - */ } /* @@ -2811,57 +3122,53 @@ static void wake_sleeping_dependent(int this_cpu, runqueue_t *this_rq) * utilize, if another task runs on a sibling. This models the * slowdown effect of other tasks running on siblings: */ -static inline unsigned long smt_slice(task_t *p, struct sched_domain *sd) +static inline unsigned long +smt_slice(struct task_struct *p, struct sched_domain *sd) { return p->time_slice * (100 - sd->per_cpu_gain) / 100; } -static int dependent_sleeper(int this_cpu, runqueue_t *this_rq) +/* + * To minimise lock contention and not have to drop this_rq's runlock we only + * trylock the sibling runqueues and bypass those runqueues if we fail to + * acquire their lock. As we only trylock the normal locking order does not + * need to be obeyed. + */ +static int +dependent_sleeper(int this_cpu, struct rq *this_rq, struct task_struct *p) { struct sched_domain *tmp, *sd = NULL; - cpumask_t sibling_map; - prio_array_t *array; int ret = 0, i; - task_t *p; - for_each_domain(this_cpu, tmp) - if (tmp->flags & SD_SHARE_CPUPOWER) + /* kernel/rt threads do not participate in dependent sleeping */ + if (!p->mm || rt_task(p)) + return 0; + + for_each_domain(this_cpu, tmp) { + if (tmp->flags & SD_SHARE_CPUPOWER) { sd = tmp; + break; + } + } if (!sd) return 0; - /* - * The same locking rules and details apply as for - * wake_sleeping_dependent(): - */ - spin_unlock(&this_rq->lock); - sibling_map = sd->span; - for_each_cpu_mask(i, sibling_map) - spin_lock(&cpu_rq(i)->lock); - cpu_clear(this_cpu, sibling_map); + for_each_cpu_mask(i, sd->span) { + struct task_struct *smt_curr; + struct rq *smt_rq; - /* - * Establish next task to be run - it might have gone away because - * we released the runqueue lock above: - */ - if (!this_rq->nr_running) - goto out_unlock; - array = this_rq->active; - if (!array->nr_active) - array = this_rq->expired; - BUG_ON(!array->nr_active); + if (i == this_cpu) + continue; - p = list_entry(array->queue[sched_find_first_bit(array->bitmap)].next, - task_t, run_list); + smt_rq = cpu_rq(i); + if (unlikely(!spin_trylock(&smt_rq->lock))) + continue; - for_each_cpu_mask(i, sibling_map) { - runqueue_t *smt_rq = cpu_rq(i); - task_t *smt_curr = smt_rq->curr; + smt_curr = smt_rq->curr; - /* Kernel threads do not participate in dependent sleeping */ - if (!p->mm || !smt_curr->mm || rt_task(p)) - goto check_smt_task; + if (!smt_curr->mm) + goto unlock; /* * If a user task with lower static priority than the @@ -2879,49 +3186,23 @@ static int dependent_sleeper(int this_cpu, runqueue_t *this_rq) if ((jiffies % DEF_TIMESLICE) > (sd->per_cpu_gain * DEF_TIMESLICE / 100)) ret = 1; - } else + } else { if (smt_curr->static_prio < p->static_prio && !TASK_PREEMPTS_CURR(p, smt_rq) && smt_slice(smt_curr, sd) > task_timeslice(p)) ret = 1; - -check_smt_task: - if ((!smt_curr->mm && smt_curr != smt_rq->idle) || - rt_task(smt_curr)) - continue; - if (!p->mm) { - wakeup_busy_runqueue(smt_rq); - continue; - } - - /* - * Reschedule a lower priority task on the SMT sibling for - * it to be put to sleep, or wake it up if it has been put to - * sleep for priority reasons to see if it should run now. - */ - if (rt_task(p)) { - if ((jiffies % DEF_TIMESLICE) > - (sd->per_cpu_gain * DEF_TIMESLICE / 100)) - resched_task(smt_curr); - } else { - if (TASK_PREEMPTS_CURR(p, smt_rq) && - smt_slice(p, sd) > task_timeslice(smt_curr)) - resched_task(smt_curr); - else - wakeup_busy_runqueue(smt_rq); } +unlock: + spin_unlock(&smt_rq->lock); } -out_unlock: - for_each_cpu_mask(i, sibling_map) - spin_unlock(&cpu_rq(i)->lock); return ret; } #else -static inline void wake_sleeping_dependent(int this_cpu, runqueue_t *this_rq) +static inline void wake_sleeping_dependent(int this_cpu) { } - -static inline int dependent_sleeper(int this_cpu, runqueue_t *this_rq) +static inline int +dependent_sleeper(int this_cpu, struct rq *this_rq, struct task_struct *p) { return 0; } @@ -2934,12 +3215,13 @@ void fastcall add_preempt_count(int val) /* * Underflow? */ - BUG_ON((preempt_count() < 0)); + if (DEBUG_LOCKS_WARN_ON((preempt_count() < 0))) + return; preempt_count() += val; /* * Spinlock count overflowing soon? */ - BUG_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK-10); + DEBUG_LOCKS_WARN_ON((preempt_count() & PREEMPT_MASK) >= PREEMPT_MASK-10); } EXPORT_SYMBOL(add_preempt_count); @@ -2948,43 +3230,51 @@ void fastcall sub_preempt_count(int val) /* * Underflow? */ - BUG_ON(val > preempt_count()); + if (DEBUG_LOCKS_WARN_ON(val > preempt_count())) + return; /* * Is the spinlock portion underflowing? */ - BUG_ON((val < PREEMPT_MASK) && !(preempt_count() & PREEMPT_MASK)); + if (DEBUG_LOCKS_WARN_ON((val < PREEMPT_MASK) && + !(preempt_count() & PREEMPT_MASK))) + return; + preempt_count() -= val; } EXPORT_SYMBOL(sub_preempt_count); #endif +static inline int interactive_sleep(enum sleep_type sleep_type) +{ + return (sleep_type == SLEEP_INTERACTIVE || + sleep_type == SLEEP_INTERRUPTED); +} + /* * schedule() is the main scheduler function. */ asmlinkage void __sched schedule(void) { - long *switch_count; - task_t *prev, *next; - runqueue_t *rq; - prio_array_t *array; + struct task_struct *prev, *next; + struct prio_array *array; struct list_head *queue; unsigned long long now; unsigned long run_time; int cpu, idx, new_prio; + long *switch_count; + struct rq *rq; /* * Test if we are atomic. Since do_exit() needs to call into * schedule() atomically, we ignore that path for now. * Otherwise, whine if we are scheduling when we should not be. */ - if (likely(!current->exit_state)) { - if (unlikely(in_atomic())) { - printk(KERN_ERR "scheduling while atomic: " - "%s/0x%08x/%d\n", - current->comm, preempt_count(), current->pid); - dump_stack(); - } + if (unlikely(in_atomic() && !current->exit_state)) { + printk(KERN_ERR "BUG: scheduling while atomic: " + "%s/0x%08x/%d\n", + current->comm, preempt_count(), current->pid); + dump_stack(); } profile_hit(SCHED_PROFILING, __builtin_return_address(0)); @@ -3039,32 +3329,13 @@ need_resched_nonpreemptible: cpu = smp_processor_id(); if (unlikely(!rq->nr_running)) { -go_idle: idle_balance(cpu, rq); if (!rq->nr_running) { next = rq->idle; rq->expired_timestamp = 0; - wake_sleeping_dependent(cpu, rq); - /* - * wake_sleeping_dependent() might have released - * the runqueue, so break out if we got new - * tasks meanwhile: - */ - if (!rq->nr_running) - goto switch_tasks; - } - } else { - if (dependent_sleeper(cpu, rq)) { - next = rq->idle; + wake_sleeping_dependent(cpu); goto switch_tasks; } - /* - * dependent_sleeper() releases and reacquires the runqueue - * lock, hence go into the idle loop if the rq went - * empty meanwhile: - */ - if (unlikely(!rq->nr_running)) - goto go_idle; } array = rq->active; @@ -3082,14 +3353,14 @@ go_idle: idx = sched_find_first_bit(array->bitmap); queue = array->queue + idx; - next = list_entry(queue->next, task_t, run_list); + next = list_entry(queue->next, struct task_struct, run_list); - if (!rt_task(next) && next->activated > 0) { + if (!rt_task(next) && interactive_sleep(next->sleep_type)) { unsigned long long delta = now - next->timestamp; if (unlikely((long long)(now - next->timestamp) < 0)) delta = 0; - if (next->activated == 1) + if (next->sleep_type == SLEEP_INTERACTIVE) delta = delta * (ON_RUNQUEUE_WEIGHT * 128 / 100) / 128; array = next->array; @@ -3099,10 +3370,11 @@ go_idle: dequeue_task(next, array); next->prio = new_prio; enqueue_task(next, array); - } else - requeue_task(next, array); + } } - next->activated = 0; + next->sleep_type = SLEEP_NORMAL; + if (dependent_sleeper(cpu, rq, next)) + next = rq->idle; switch_tasks: if (next == rq->idle) schedstat_inc(rq, sched_goidle); @@ -3144,12 +3416,11 @@ switch_tasks: if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) goto need_resched; } - EXPORT_SYMBOL(schedule); #ifdef CONFIG_PREEMPT /* - * this is is the entry point to schedule() from in-kernel preemption + * this is the entry point to schedule() from in-kernel preemption * off of preempt_enable. Kernel preemptions off return from interrupt * occur there and call schedule directly. */ @@ -3189,11 +3460,10 @@ need_resched: if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) goto need_resched; } - EXPORT_SYMBOL(preempt_schedule); /* - * this is is the entry point to schedule() from kernel preemption + * this is the entry point to schedule() from kernel preemption * off of irq context. * Note, that this is called and return with irqs disabled. This will * protect us against recursive calling from irq. @@ -3205,7 +3475,7 @@ asmlinkage void __sched preempt_schedule_irq(void) struct task_struct *task = current; int saved_lock_depth; #endif - /* Catch callers which need to be fixed*/ + /* Catch callers which need to be fixed */ BUG_ON(ti->preempt_count || !irqs_disabled()); need_resched: @@ -3238,10 +3508,8 @@ need_resched: int default_wake_function(wait_queue_t *curr, unsigned mode, int sync, void *key) { - task_t *p = curr->private; - return try_to_wake_up(p, mode, sync); + return try_to_wake_up(curr->private, mode, sync); } - EXPORT_SYMBOL(default_wake_function); /* @@ -3259,13 +3527,11 @@ static void __wake_up_common(wait_queue_head_t *q, unsigned int mode, struct list_head *tmp, *next; list_for_each_safe(tmp, next, &q->task_list) { - wait_queue_t *curr; - unsigned flags; - curr = list_entry(tmp, wait_queue_t, task_list); - flags = curr->flags; + wait_queue_t *curr = list_entry(tmp, wait_queue_t, task_list); + unsigned flags = curr->flags; + if (curr->func(curr, mode, sync, key) && - (flags & WQ_FLAG_EXCLUSIVE) && - !--nr_exclusive) + (flags & WQ_FLAG_EXCLUSIVE) && !--nr_exclusive) break; } } @@ -3286,7 +3552,6 @@ void fastcall __wake_up(wait_queue_head_t *q, unsigned int mode, __wake_up_common(q, mode, nr_exclusive, 0, key); spin_unlock_irqrestore(&q->lock, flags); } - EXPORT_SYMBOL(__wake_up); /* @@ -3355,6 +3620,7 @@ EXPORT_SYMBOL(complete_all); void fastcall __sched wait_for_completion(struct completion *x) { might_sleep(); + spin_lock_irq(&x->wait.lock); if (!x->done) { DECLARE_WAITQUEUE(wait, current); @@ -3499,7 +3765,6 @@ void fastcall __sched interruptible_sleep_on(wait_queue_head_t *q) schedule(); SLEEP_ON_TAIL } - EXPORT_SYMBOL(interruptible_sleep_on); long fastcall __sched @@ -3515,7 +3780,6 @@ interruptible_sleep_on_timeout(wait_queue_head_t *q, long timeout) return timeout; } - EXPORT_SYMBOL(interruptible_sleep_on_timeout); void fastcall __sched sleep_on(wait_queue_head_t *q) @@ -3528,7 +3792,6 @@ void fastcall __sched sleep_on(wait_queue_head_t *q) schedule(); SLEEP_ON_TAIL } - EXPORT_SYMBOL(sleep_on); long fastcall __sched sleep_on_timeout(wait_queue_head_t *q, long timeout) @@ -3546,12 +3809,65 @@ long fastcall __sched sleep_on_timeout(wait_queue_head_t *q, long timeout) EXPORT_SYMBOL(sleep_on_timeout); -void set_user_nice(task_t *p, long nice) +#ifdef CONFIG_RT_MUTEXES + +/* + * rt_mutex_setprio - set the current priority of a task + * @p: task + * @prio: prio value (kernel-internal form) + * + * This function changes the 'effective' priority of a task. It does + * not touch ->normal_prio like __setscheduler(). + * + * Used by the rt_mutex code to implement priority inheritance logic. + */ +void rt_mutex_setprio(struct task_struct *p, int prio) { + struct prio_array *array; unsigned long flags; - prio_array_t *array; - runqueue_t *rq; - int old_prio, new_prio, delta; + struct rq *rq; + int oldprio; + + BUG_ON(prio < 0 || prio > MAX_PRIO); + + rq = task_rq_lock(p, &flags); + + oldprio = p->prio; + array = p->array; + if (array) + dequeue_task(p, array); + p->prio = prio; + + if (array) { + /* + * If changing to an RT priority then queue it + * in the active array! + */ + if (rt_task(p)) + array = rq->active; + enqueue_task(p, array); + /* + * Reschedule if we are currently running on this runqueue and + * our priority decreased, or if we are not currently running on + * this runqueue and our priority is higher than the current's + */ + if (task_running(rq, p)) { + if (p->prio > oldprio) + resched_task(rq->curr); + } else if (TASK_PREEMPTS_CURR(p, rq)) + resched_task(rq->curr); + } + task_rq_unlock(rq, &flags); +} + +#endif + +void set_user_nice(struct task_struct *p, long nice) +{ + struct prio_array *array; + int old_prio, delta; + unsigned long flags; + struct rq *rq; if (TASK_NICE(p) == nice || nice < -20 || nice > 19) return; @@ -3566,25 +3882,25 @@ void set_user_nice(task_t *p, long nice) * it wont have any effect on scheduling until the task is * not SCHED_NORMAL/SCHED_BATCH: */ - if (rt_task(p)) { + if (has_rt_policy(p)) { p->static_prio = NICE_TO_PRIO(nice); goto out_unlock; } array = p->array; if (array) { dequeue_task(p, array); - dec_prio_bias(rq, p->static_prio); + dec_raw_weighted_load(rq, p); } - old_prio = p->prio; - new_prio = NICE_TO_PRIO(nice); - delta = new_prio - old_prio; p->static_prio = NICE_TO_PRIO(nice); - p->prio += delta; + set_load_weight(p); + old_prio = p->prio; + p->prio = effective_prio(p); + delta = p->prio - old_prio; if (array) { enqueue_task(p, array); - inc_prio_bias(rq, p->static_prio); + inc_raw_weighted_load(rq, p); /* * If the task increased its priority or is running and * lowered its priority, then reschedule its CPU: @@ -3595,7 +3911,6 @@ void set_user_nice(task_t *p, long nice) out_unlock: task_rq_unlock(rq, &flags); } - EXPORT_SYMBOL(set_user_nice); /* @@ -3603,10 +3918,11 @@ EXPORT_SYMBOL(set_user_nice); * @p: task * @nice: nice value */ -int can_nice(const task_t *p, const int nice) +int can_nice(const struct task_struct *p, const int nice) { /* convert nice value [19,-20] to rlimit style value [1,40] */ int nice_rlim = 20 - nice; + return (nice_rlim <= p->signal->rlim[RLIMIT_NICE].rlim_cur || capable(CAP_SYS_NICE)); } @@ -3622,8 +3938,7 @@ int can_nice(const task_t *p, const int nice) */ asmlinkage long sys_nice(int increment) { - int retval; - long nice; + long nice, retval; /* * Setpriority might change our priority at the same moment. @@ -3662,7 +3977,7 @@ asmlinkage long sys_nice(int increment) * RT tasks are offset by -200. Normal tasks are centered * around 0, value goes from -16 to +15. */ -int task_prio(const task_t *p) +int task_prio(const struct task_struct *p) { return p->prio - MAX_RT_PRIO; } @@ -3671,7 +3986,7 @@ int task_prio(const task_t *p) * task_nice - return the nice value of a given task. * @p: the task in question. */ -int task_nice(const task_t *p) +int task_nice(const struct task_struct *p) { return TASK_NICE(p); } @@ -3690,7 +4005,7 @@ int idle_cpu(int cpu) * idle_task - return the idle task for a given cpu. * @cpu: the processor in question. */ -task_t *idle_task(int cpu) +struct task_struct *idle_task(int cpu) { return cpu_rq(cpu)->idle; } @@ -3699,7 +4014,7 @@ task_t *idle_task(int cpu) * find_process_by_pid - find a process with a matching PID value. * @pid: the pid in question. */ -static inline task_t *find_process_by_pid(pid_t pid) +static inline struct task_struct *find_process_by_pid(pid_t pid) { return pid ? find_task_by_pid(pid) : current; } @@ -3708,18 +4023,18 @@ static inline task_t *find_process_by_pid(pid_t pid) static void __setscheduler(struct task_struct *p, int policy, int prio) { BUG_ON(p->array); + p->policy = policy; p->rt_priority = prio; - if (policy != SCHED_NORMAL && policy != SCHED_BATCH) { - p->prio = MAX_RT_PRIO-1 - p->rt_priority; - } else { - p->prio = p->static_prio; - /* - * SCHED_BATCH tasks are treated as perpetual CPU hogs: - */ - if (policy == SCHED_BATCH) - p->sleep_avg = 0; - } + p->normal_prio = normal_prio(p); + /* we are holding p->pi_lock already */ + p->prio = rt_mutex_getprio(p); + /* + * SCHED_BATCH tasks are treated as perpetual CPU hogs: + */ + if (policy == SCHED_BATCH) + p->sleep_avg = 0; + set_load_weight(p); } /** @@ -3732,12 +4047,13 @@ static void __setscheduler(struct task_struct *p, int policy, int prio) int sched_setscheduler(struct task_struct *p, int policy, struct sched_param *param) { - int retval; - int oldprio, oldpolicy = -1; - prio_array_t *array; + int retval, oldprio, oldpolicy = -1; + struct prio_array *array; unsigned long flags; - runqueue_t *rq; + struct rq *rq; + /* may grab non-irq protected spin_locks */ + BUG_ON(in_interrupt()); recheck: /* double check policy once rq lock held */ if (policy < 0) @@ -3786,14 +4102,20 @@ recheck: if (retval) return retval; /* + * make sure no PI-waiters arrive (or leave) while we are + * changing the priority of the task: + */ + spin_lock_irqsave(&p->pi_lock, flags); + /* * To be able to change p->policy safely, the apropriate * runqueue lock must be held. */ - rq = task_rq_lock(p, &flags); + rq = __task_rq_lock(p); /* recheck policy now with rq lock held */ if (unlikely(oldpolicy != -1 && oldpolicy != p->policy)) { policy = oldpolicy = -1; - task_rq_unlock(rq, &flags); + __task_rq_unlock(rq); + spin_unlock_irqrestore(&p->pi_lock, flags); goto recheck; } array = p->array; @@ -3814,7 +4136,11 @@ recheck: } else if (TASK_PREEMPTS_CURR(p, rq)) resched_task(rq->curr); } - task_rq_unlock(rq, &flags); + __task_rq_unlock(rq); + spin_unlock_irqrestore(&p->pi_lock, flags); + + rt_mutex_adjust_pi(p); + return 0; } EXPORT_SYMBOL_GPL(sched_setscheduler); @@ -3822,9 +4148,9 @@ EXPORT_SYMBOL_GPL(sched_setscheduler); static int do_sched_setscheduler(pid_t pid, int policy, struct sched_param __user *param) { - int retval; struct sched_param lparam; struct task_struct *p; + int retval; if (!param || pid < 0) return -EINVAL; @@ -3836,8 +4162,11 @@ do_sched_setscheduler(pid_t pid, int policy, struct sched_param __user *param) read_unlock_irq(&tasklist_lock); return -ESRCH; } - retval = sched_setscheduler(p, policy, &lparam); + get_task_struct(p); read_unlock_irq(&tasklist_lock); + retval = sched_setscheduler(p, policy, &lparam); + put_task_struct(p); + return retval; } @@ -3850,6 +4179,10 @@ do_sched_setscheduler(pid_t pid, int policy, struct sched_param __user *param) asmlinkage long sys_sched_setscheduler(pid_t pid, int policy, struct sched_param __user *param) { + /* negative values for policy are not valid */ + if (policy < 0) + return -EINVAL; + return do_sched_setscheduler(pid, policy, param); } @@ -3869,8 +4202,8 @@ asmlinkage long sys_sched_setparam(pid_t pid, struct sched_param __user *param) */ asmlinkage long sys_sched_getscheduler(pid_t pid) { + struct task_struct *p; int retval = -EINVAL; - task_t *p; if (pid < 0) goto out_nounlock; @@ -3897,8 +4230,8 @@ out_nounlock: asmlinkage long sys_sched_getparam(pid_t pid, struct sched_param __user *param) { struct sched_param lp; + struct task_struct *p; int retval = -EINVAL; - task_t *p; if (!param || pid < 0) goto out_nounlock; @@ -3931,9 +4264,9 @@ out_unlock: long sched_setaffinity(pid_t pid, cpumask_t new_mask) { - task_t *p; - int retval; cpumask_t cpus_allowed; + struct task_struct *p; + int retval; lock_cpu_hotplug(); read_lock(&tasklist_lock); @@ -3958,6 +4291,10 @@ long sched_setaffinity(pid_t pid, cpumask_t new_mask) !capable(CAP_SYS_NICE)) goto out_unlock; + retval = security_task_setscheduler(p, 0, NULL); + if (retval) + goto out_unlock; + cpus_allowed = cpuset_cpus_allowed(p); cpus_and(new_mask, new_mask, cpus_allowed); retval = set_cpus_allowed(p, new_mask); @@ -4015,8 +4352,8 @@ cpumask_t cpu_possible_map __read_mostly = CPU_MASK_ALL; long sched_getaffinity(pid_t pid, cpumask_t *mask) { + struct task_struct *p; int retval; - task_t *p; lock_cpu_hotplug(); read_lock(&tasklist_lock); @@ -4026,8 +4363,11 @@ long sched_getaffinity(pid_t pid, cpumask_t *mask) if (!p) goto out_unlock; - retval = 0; - cpus_and(*mask, p->cpus_allowed, cpu_possible_map); + retval = security_task_getscheduler(p); + if (retval) + goto out_unlock; + + cpus_and(*mask, p->cpus_allowed, cpu_online_map); out_unlock: read_unlock(&tasklist_lock); @@ -4072,9 +4412,8 @@ asmlinkage long sys_sched_getaffinity(pid_t pid, unsigned int len, */ asmlinkage long sys_sched_yield(void) { - runqueue_t *rq = this_rq_lock(); - prio_array_t *array = current->array; - prio_array_t *target = rq->expired; + struct rq *rq = this_rq_lock(); + struct prio_array *array = current->array, *target = rq->expired; schedstat_inc(rq, yld_cnt); /* @@ -4108,6 +4447,7 @@ asmlinkage long sys_sched_yield(void) * no need to preempt or enable interrupts: */ __release(rq->lock); + spin_release(&rq->lock.dep_map, 1, _THIS_IP_); _raw_spin_unlock(&rq->lock); preempt_enable_no_resched(); @@ -4116,15 +4456,25 @@ asmlinkage long sys_sched_yield(void) return 0; } -static inline void __cond_resched(void) +static inline int __resched_legal(int expected_preempt_count) { + if (unlikely(preempt_count() != expected_preempt_count)) + return 0; + if (unlikely(system_state != SYSTEM_RUNNING)) + return 0; + return 1; +} + +static void __cond_resched(void) +{ +#ifdef CONFIG_DEBUG_SPINLOCK_SLEEP + __might_sleep(__FILE__, __LINE__); +#endif /* * The BKS might be reacquired before we have dropped * PREEMPT_ACTIVE, which could trigger a second * cond_resched() call. */ - if (unlikely(preempt_count())) - return; do { add_preempt_count(PREEMPT_ACTIVE); schedule(); @@ -4134,13 +4484,12 @@ static inline void __cond_resched(void) int __sched cond_resched(void) { - if (need_resched()) { + if (need_resched() && __resched_legal(0)) { __cond_resched(); return 1; } return 0; } - EXPORT_SYMBOL(cond_resched); /* @@ -4161,7 +4510,8 @@ int cond_resched_lock(spinlock_t *lock) ret = 1; spin_lock(lock); } - if (need_resched()) { + if (need_resched() && __resched_legal(1)) { + spin_release(&lock->dep_map, 1, _THIS_IP_); _raw_spin_unlock(lock); preempt_enable_no_resched(); __cond_resched(); @@ -4170,25 +4520,24 @@ int cond_resched_lock(spinlock_t *lock) } return ret; } - EXPORT_SYMBOL(cond_resched_lock); int __sched cond_resched_softirq(void) { BUG_ON(!in_softirq()); - if (need_resched()) { - __local_bh_enable(); + if (need_resched() && __resched_legal(0)) { + raw_local_irq_disable(); + _local_bh_enable(); + raw_local_irq_enable(); __cond_resched(); local_bh_disable(); return 1; } return 0; } - EXPORT_SYMBOL(cond_resched_softirq); - /** * yield - yield the current processor to other threads. * @@ -4200,7 +4549,6 @@ void __sched yield(void) set_current_state(TASK_RUNNING); sys_sched_yield(); } - EXPORT_SYMBOL(yield); /* @@ -4212,23 +4560,26 @@ EXPORT_SYMBOL(yield); */ void __sched io_schedule(void) { - struct runqueue *rq = &per_cpu(runqueues, raw_smp_processor_id()); + struct rq *rq = &__raw_get_cpu_var(runqueues); + delayacct_blkio_start(); atomic_inc(&rq->nr_iowait); schedule(); atomic_dec(&rq->nr_iowait); + delayacct_blkio_end(); } - EXPORT_SYMBOL(io_schedule); long __sched io_schedule_timeout(long timeout) { - struct runqueue *rq = &per_cpu(runqueues, raw_smp_processor_id()); + struct rq *rq = &__raw_get_cpu_var(runqueues); long ret; + delayacct_blkio_start(); atomic_inc(&rq->nr_iowait); ret = schedule_timeout(timeout); atomic_dec(&rq->nr_iowait); + delayacct_blkio_end(); return ret; } @@ -4290,9 +4641,9 @@ asmlinkage long sys_sched_get_priority_min(int policy) asmlinkage long sys_sched_rr_get_interval(pid_t pid, struct timespec __user *interval) { + struct task_struct *p; int retval = -EINVAL; struct timespec t; - task_t *p; if (pid < 0) goto out_nounlock; @@ -4307,7 +4658,7 @@ long sys_sched_rr_get_interval(pid_t pid, struct timespec __user *interval) if (retval) goto out_unlock; - jiffies_to_timespec(p->policy & SCHED_FIFO ? + jiffies_to_timespec(p->policy == SCHED_FIFO ? 0 : task_timeslice(p), &t); read_unlock(&tasklist_lock); retval = copy_to_user(interval, &t, sizeof(t)) ? -EFAULT : 0; @@ -4320,35 +4671,36 @@ out_unlock: static inline struct task_struct *eldest_child(struct task_struct *p) { - if (list_empty(&p->children)) return NULL; + if (list_empty(&p->children)) + return NULL; return list_entry(p->children.next,struct task_struct,sibling); } static inline struct task_struct *older_sibling(struct task_struct *p) { - if (p->sibling.prev==&p->parent->children) return NULL; + if (p->sibling.prev==&p->parent->children) + return NULL; return list_entry(p->sibling.prev,struct task_struct,sibling); } static inline struct task_struct *younger_sibling(struct task_struct *p) { - if (p->sibling.next==&p->parent->children) return NULL; + if (p->sibling.next==&p->parent->children) + return NULL; return list_entry(p->sibling.next,struct task_struct,sibling); } -static void show_task(task_t *p) +static const char stat_nam[] = "RSDTtZX"; + +static void show_task(struct task_struct *p) { - task_t *relative; - unsigned state; + struct task_struct *relative; unsigned long free = 0; - static const char *stat_nam[] = { "R", "S", "D", "T", "t", "Z", "X" }; + unsigned state; - printk("%-13.13s ", p->comm); state = p->state ? __ffs(p->state) + 1 : 0; - if (state < ARRAY_SIZE(stat_nam)) - printk(stat_nam[state]); - else - printk("?"); + printk("%-13.13s %c", p->comm, + state < sizeof(stat_nam) - 1 ? stat_nam[state] : '?'); #if (BITS_PER_LONG == 32) if (state == TASK_RUNNING) printk(" running "); @@ -4392,7 +4744,7 @@ static void show_task(task_t *p) void show_state(void) { - task_t *g, *p; + struct task_struct *g, *p; #if (BITS_PER_LONG == 32) printk("\n" @@ -4414,7 +4766,7 @@ void show_state(void) } while_each_thread(g, p); read_unlock(&tasklist_lock); - mutex_debug_show_all_locks(); + debug_show_all_locks(); } /** @@ -4425,14 +4777,15 @@ void show_state(void) * NOTE: this function does not set the idle thread's NEED_RESCHED * flag, to make booting more robust. */ -void __devinit init_idle(task_t *idle, int cpu) +void __devinit init_idle(struct task_struct *idle, int cpu) { - runqueue_t *rq = cpu_rq(cpu); + struct rq *rq = cpu_rq(cpu); unsigned long flags; + idle->timestamp = sched_clock(); idle->sleep_avg = 0; idle->array = NULL; - idle->prio = MAX_PRIO; + idle->prio = idle->normal_prio = MAX_PRIO; idle->state = TASK_RUNNING; idle->cpus_allowed = cpumask_of_cpu(cpu); set_task_cpu(idle, cpu); @@ -4465,7 +4818,7 @@ cpumask_t nohz_cpu_mask = CPU_MASK_NONE; /* * This is how migration works: * - * 1) we queue a migration_req_t structure in the source CPU's + * 1) we queue a struct migration_req structure in the source CPU's * runqueue and wake up that CPU's migration thread. * 2) we down() the locked semaphore => thread blocks. * 3) migration thread wakes up (implicitly it forces the migrated @@ -4487,12 +4840,12 @@ cpumask_t nohz_cpu_mask = CPU_MASK_NONE; * task must not exit() & deallocate itself prematurely. The * call is not atomic; no spinlocks may be held. */ -int set_cpus_allowed(task_t *p, cpumask_t new_mask) +int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask) { + struct migration_req req; unsigned long flags; + struct rq *rq; int ret = 0; - migration_req_t req; - runqueue_t *rq; rq = task_rq_lock(p, &flags); if (!cpus_intersects(new_mask, cpu_online_map)) { @@ -4515,9 +4868,9 @@ int set_cpus_allowed(task_t *p, cpumask_t new_mask) } out: task_rq_unlock(rq, &flags); + return ret; } - EXPORT_SYMBOL_GPL(set_cpus_allowed); /* @@ -4528,13 +4881,16 @@ EXPORT_SYMBOL_GPL(set_cpus_allowed); * * So we race with normal scheduler movements, but that's OK, as long * as the task is no longer on this CPU. + * + * Returns non-zero if task was successfully migrated. */ -static void __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu) +static int __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu) { - runqueue_t *rq_dest, *rq_src; + struct rq *rq_dest, *rq_src; + int ret = 0; if (unlikely(cpu_is_offline(dest_cpu))) - return; + return ret; rq_src = cpu_rq(src_cpu); rq_dest = cpu_rq(dest_cpu); @@ -4558,13 +4914,14 @@ static void __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu) p->timestamp = p->timestamp - rq_src->timestamp_last_tick + rq_dest->timestamp_last_tick; deactivate_task(p, rq_src); - activate_task(p, rq_dest, 0); + __activate_task(p, rq_dest); if (TASK_PREEMPTS_CURR(p, rq_dest)) resched_task(rq_dest->curr); } - + ret = 1; out: double_rq_unlock(rq_src, rq_dest); + return ret; } /* @@ -4574,16 +4931,16 @@ out: */ static int migration_thread(void *data) { - runqueue_t *rq; int cpu = (long)data; + struct rq *rq; rq = cpu_rq(cpu); BUG_ON(rq->migration_thread != current); set_current_state(TASK_INTERRUPTIBLE); while (!kthread_should_stop()) { + struct migration_req *req; struct list_head *head; - migration_req_t *req; try_to_freeze(); @@ -4607,7 +4964,7 @@ static int migration_thread(void *data) set_current_state(TASK_INTERRUPTIBLE); continue; } - req = list_entry(head->next, migration_req_t, list); + req = list_entry(head->next, struct migration_req, list); list_del_init(head->next); spin_unlock(&rq->lock); @@ -4632,36 +4989,42 @@ wait_to_die: #ifdef CONFIG_HOTPLUG_CPU /* Figure out where task on dead CPU should go, use force if neccessary. */ -static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *tsk) +static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p) { - int dest_cpu; + unsigned long flags; cpumask_t mask; + struct rq *rq; + int dest_cpu; +restart: /* On same node? */ mask = node_to_cpumask(cpu_to_node(dead_cpu)); - cpus_and(mask, mask, tsk->cpus_allowed); + cpus_and(mask, mask, p->cpus_allowed); dest_cpu = any_online_cpu(mask); /* On any allowed CPU? */ if (dest_cpu == NR_CPUS) - dest_cpu = any_online_cpu(tsk->cpus_allowed); + dest_cpu = any_online_cpu(p->cpus_allowed); /* No more Mr. Nice Guy. */ if (dest_cpu == NR_CPUS) { - cpus_setall(tsk->cpus_allowed); - dest_cpu = any_online_cpu(tsk->cpus_allowed); + rq = task_rq_lock(p, &flags); + cpus_setall(p->cpus_allowed); + dest_cpu = any_online_cpu(p->cpus_allowed); + task_rq_unlock(rq, &flags); /* * Don't tell them about moving exiting tasks or * kernel threads (both mm NULL), since they never * leave kernel. */ - if (tsk->mm && printk_ratelimit()) + if (p->mm && printk_ratelimit()) printk(KERN_INFO "process %d (%s) no " "longer affine to cpu%d\n", - tsk->pid, tsk->comm, dead_cpu); + p->pid, p->comm, dead_cpu); } - __migrate_task(tsk, dead_cpu, dest_cpu); + if (!__migrate_task(p, dead_cpu, dest_cpu)) + goto restart; } /* @@ -4671,9 +5034,9 @@ static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *tsk) * their home CPUs. So we just add the counter to another CPU's counter, * to keep the global sum constant after CPU-down: */ -static void migrate_nr_uninterruptible(runqueue_t *rq_src) +static void migrate_nr_uninterruptible(struct rq *rq_src) { - runqueue_t *rq_dest = cpu_rq(any_online_cpu(CPU_MASK_ALL)); + struct rq *rq_dest = cpu_rq(any_online_cpu(CPU_MASK_ALL)); unsigned long flags; local_irq_save(flags); @@ -4687,48 +5050,51 @@ static void migrate_nr_uninterruptible(runqueue_t *rq_src) /* Run through task list and migrate tasks from the dead cpu. */ static void migrate_live_tasks(int src_cpu) { - struct task_struct *tsk, *t; + struct task_struct *p, *t; write_lock_irq(&tasklist_lock); - do_each_thread(t, tsk) { - if (tsk == current) + do_each_thread(t, p) { + if (p == current) continue; - if (task_cpu(tsk) == src_cpu) - move_task_off_dead_cpu(src_cpu, tsk); - } while_each_thread(t, tsk); + if (task_cpu(p) == src_cpu) + move_task_off_dead_cpu(src_cpu, p); + } while_each_thread(t, p); write_unlock_irq(&tasklist_lock); } /* Schedules idle task to be the next runnable task on current CPU. * It does so by boosting its priority to highest possible and adding it to - * the _front_ of runqueue. Used by CPU offline code. + * the _front_ of the runqueue. Used by CPU offline code. */ void sched_idle_next(void) { - int cpu = smp_processor_id(); - runqueue_t *rq = this_rq(); + int this_cpu = smp_processor_id(); + struct rq *rq = cpu_rq(this_cpu); struct task_struct *p = rq->idle; unsigned long flags; /* cpu has to be offline */ - BUG_ON(cpu_online(cpu)); + BUG_ON(cpu_online(this_cpu)); - /* Strictly not necessary since rest of the CPUs are stopped by now - * and interrupts disabled on current cpu. + /* + * Strictly not necessary since rest of the CPUs are stopped by now + * and interrupts disabled on the current cpu. */ spin_lock_irqsave(&rq->lock, flags); __setscheduler(p, SCHED_FIFO, MAX_RT_PRIO-1); - /* Add idle task to _front_ of it's priority queue */ + + /* Add idle task to the _front_ of its priority queue: */ __activate_idle_task(p, rq); spin_unlock_irqrestore(&rq->lock, flags); } -/* Ensures that the idle task is using init_mm right before its cpu goes +/* + * Ensures that the idle task is using init_mm right before its cpu goes * offline. */ void idle_task_exit(void) @@ -4742,17 +5108,17 @@ void idle_task_exit(void) mmdrop(mm); } -static void migrate_dead(unsigned int dead_cpu, task_t *tsk) +static void migrate_dead(unsigned int dead_cpu, struct task_struct *p) { - struct runqueue *rq = cpu_rq(dead_cpu); + struct rq *rq = cpu_rq(dead_cpu); /* Must be exiting, otherwise would be on tasklist. */ - BUG_ON(tsk->exit_state != EXIT_ZOMBIE && tsk->exit_state != EXIT_DEAD); + BUG_ON(p->exit_state != EXIT_ZOMBIE && p->exit_state != EXIT_DEAD); /* Cannot have done final schedule yet: would have vanished. */ - BUG_ON(tsk->flags & PF_DEAD); + BUG_ON(p->flags & PF_DEAD); - get_task_struct(tsk); + get_task_struct(p); /* * Drop lock around migration; if someone else moves it, @@ -4760,25 +5126,25 @@ static void migrate_dead(unsigned int dead_cpu, task_t *tsk) * fine. */ spin_unlock_irq(&rq->lock); - move_task_off_dead_cpu(dead_cpu, tsk); + move_task_off_dead_cpu(dead_cpu, p); spin_lock_irq(&rq->lock); - put_task_struct(tsk); + put_task_struct(p); } /* release_task() removes task from tasklist, so we won't find dead tasks. */ static void migrate_dead_tasks(unsigned int dead_cpu) { - unsigned arr, i; - struct runqueue *rq = cpu_rq(dead_cpu); + struct rq *rq = cpu_rq(dead_cpu); + unsigned int arr, i; for (arr = 0; arr < 2; arr++) { for (i = 0; i < MAX_PRIO; i++) { struct list_head *list = &rq->arrays[arr].queue[i]; + while (!list_empty(list)) - migrate_dead(dead_cpu, - list_entry(list->next, task_t, - run_list)); + migrate_dead(dead_cpu, list_entry(list->next, + struct task_struct, run_list)); } } } @@ -4788,13 +5154,13 @@ static void migrate_dead_tasks(unsigned int dead_cpu) * migration_call - callback that gets triggered when a CPU is added. * Here we can start up the necessary migration thread for the new CPU. */ -static int migration_call(struct notifier_block *nfb, unsigned long action, - void *hcpu) +static int __cpuinit +migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) { - int cpu = (long)hcpu; struct task_struct *p; - struct runqueue *rq; + int cpu = (long)hcpu; unsigned long flags; + struct rq *rq; switch (action) { case CPU_UP_PREPARE: @@ -4809,18 +5175,23 @@ static int migration_call(struct notifier_block *nfb, unsigned long action, task_rq_unlock(rq, &flags); cpu_rq(cpu)->migration_thread = p; break; + case CPU_ONLINE: /* Strictly unneccessary, as first user will wake it. */ wake_up_process(cpu_rq(cpu)->migration_thread); break; + #ifdef CONFIG_HOTPLUG_CPU case CPU_UP_CANCELED: + if (!cpu_rq(cpu)->migration_thread) + break; /* Unbind it from offline cpu so it can run. Fall thru. */ kthread_bind(cpu_rq(cpu)->migration_thread, any_online_cpu(cpu_online_map)); kthread_stop(cpu_rq(cpu)->migration_thread); cpu_rq(cpu)->migration_thread = NULL; break; + case CPU_DEAD: migrate_live_tasks(cpu); rq = cpu_rq(cpu); @@ -4841,9 +5212,10 @@ static int migration_call(struct notifier_block *nfb, unsigned long action, * the requestors. */ spin_lock_irq(&rq->lock); while (!list_empty(&rq->migration_queue)) { - migration_req_t *req; + struct migration_req *req; + req = list_entry(rq->migration_queue.next, - migration_req_t, list); + struct migration_req, list); list_del_init(&req->list); complete(&req->done); } @@ -4857,7 +5229,7 @@ static int migration_call(struct notifier_block *nfb, unsigned long action, /* Register at highest priority so that task migration (migrate_all_tasks) * happens before everything else. */ -static struct notifier_block __devinitdata migration_notifier = { +static struct notifier_block __cpuinitdata migration_notifier = { .notifier_call = migration_call, .priority = 10 }; @@ -4865,10 +5237,12 @@ static struct notifier_block __devinitdata migration_notifier = { int __init migration_init(void) { void *cpu = (void *)(long)smp_processor_id(); - /* Start one for boot CPU. */ + + /* Start one for the boot CPU: */ migration_call(&migration_notifier, CPU_UP_PREPARE, cpu); migration_call(&migration_notifier, CPU_ONLINE, cpu); register_cpu_notifier(&migration_notifier); + return 0; } #endif @@ -4964,7 +5338,7 @@ static void sched_domain_debug(struct sched_domain *sd, int cpu) } while (sd); } #else -#define sched_domain_debug(sd, cpu) {} +# define sched_domain_debug(sd, cpu) do { } while (0) #endif static int sd_degenerate(struct sched_domain *sd) @@ -4990,8 +5364,8 @@ static int sd_degenerate(struct sched_domain *sd) return 1; } -static int sd_parent_degenerate(struct sched_domain *sd, - struct sched_domain *parent) +static int +sd_parent_degenerate(struct sched_domain *sd, struct sched_domain *parent) { unsigned long cflags = sd->flags, pflags = parent->flags; @@ -5024,7 +5398,7 @@ static int sd_parent_degenerate(struct sched_domain *sd, */ static void cpu_attach_domain(struct sched_domain *sd, int cpu) { - runqueue_t *rq = cpu_rq(cpu); + struct rq *rq = cpu_rq(cpu); struct sched_domain *tmp; /* Remove the sched domains which do not contribute to scheduling. */ @@ -5137,7 +5511,7 @@ static void init_sched_build_groups(struct sched_group groups[], cpumask_t span, #define SEARCH_SCOPE 2 #define MIN_CACHE_SIZE (64*1024U) #define DEFAULT_CACHE_SIZE (5*1024*1024U) -#define ITERATIONS 2 +#define ITERATIONS 1 #define SIZE_THRESH 130 #define COST_THRESH 130 @@ -5155,7 +5529,18 @@ static void init_sched_build_groups(struct sched_group groups[], cpumask_t span, #define MAX_DOMAIN_DISTANCE 32 static unsigned long long migration_cost[MAX_DOMAIN_DISTANCE] = - { [ 0 ... MAX_DOMAIN_DISTANCE-1 ] = -1LL }; + { [ 0 ... MAX_DOMAIN_DISTANCE-1 ] = +/* + * Architectures may override the migration cost and thus avoid + * boot-time calibration. Unit is nanoseconds. Mostly useful for + * virtualized hardware: + */ +#ifdef CONFIG_DEFAULT_MIGRATION_COST + CONFIG_DEFAULT_MIGRATION_COST +#else + -1LL +#endif +}; /* * Allow override of migration cost - in units of microseconds. @@ -5275,8 +5660,8 @@ static void touch_cache(void *__cache, unsigned long __size) /* * Measure the cache-cost of one task migration. Returns in units of nsec. */ -static unsigned long long measure_one(void *cache, unsigned long size, - int source, int target) +static unsigned long long +measure_one(void *cache, unsigned long size, int source, int target) { cpumask_t mask, saved_mask; unsigned long long t0, t1, t2, t3, cost; @@ -5428,7 +5813,7 @@ static unsigned long long measure_migration_cost(int cpu1, int cpu2) cache = vmalloc(max_size); if (!cache) { printk("could not vmalloc %d bytes for cache!\n", 2*max_size); - return 1000000; // return 1 msec on very small boxen + return 1000000; /* return 1 msec on very small boxen */ } while (size <= max_size) { @@ -5476,9 +5861,9 @@ static unsigned long long measure_migration_cost(int cpu1, int cpu2) break; } /* - * Increase the cachesize in 5% steps: + * Increase the cachesize in 10% steps: */ - size = size * 20 / 19; + size = size * 10 / 9; } if (migration_debug) @@ -5547,13 +5932,15 @@ static void calibrate_migration_costs(const cpumask_t *cpu_map) -1 #endif ); - printk("migration_cost="); - for (distance = 0; distance <= max_distance; distance++) { - if (distance) - printk(","); - printk("%ld", (long)migration_cost[distance] / 1000); + if (system_state == SYSTEM_BOOTING) { + printk("migration_cost="); + for (distance = 0; distance <= max_distance; distance++) { + if (distance) + printk(","); + printk("%ld", (long)migration_cost[distance] / 1000); + } + printk("\n"); } - printk("\n"); j1 = jiffies; if (migration_debug) printk("migration: %ld seconds\n", (j1-j0)/HZ); @@ -5624,9 +6011,9 @@ static int find_next_best_node(int node, unsigned long *used_nodes) */ static cpumask_t sched_domain_node_span(int node) { - int i; - cpumask_t span, nodemask; DECLARE_BITMAP(used_nodes, MAX_NUMNODES); + cpumask_t span, nodemask; + int i; cpus_clear(span); bitmap_zero(used_nodes, MAX_NUMNODES); @@ -5637,6 +6024,7 @@ static cpumask_t sched_domain_node_span(int node) for (i = 1; i < SD_NODES_PER_DOMAIN; i++) { int next_node = find_next_best_node(node, used_nodes); + nodemask = node_to_cpumask(next_node); cpus_or(span, span, nodemask); } @@ -5645,24 +6033,50 @@ static cpumask_t sched_domain_node_span(int node) } #endif +int sched_smt_power_savings = 0, sched_mc_power_savings = 0; + /* - * At the moment, CONFIG_SCHED_SMT is never defined, but leave it in so we - * can switch it on easily if needed. + * SMT sched-domains: */ #ifdef CONFIG_SCHED_SMT static DEFINE_PER_CPU(struct sched_domain, cpu_domains); static struct sched_group sched_group_cpus[NR_CPUS]; + static int cpu_to_cpu_group(int cpu) { return cpu; } #endif +/* + * multi-core sched-domains: + */ +#ifdef CONFIG_SCHED_MC +static DEFINE_PER_CPU(struct sched_domain, core_domains); +static struct sched_group *sched_group_core_bycpu[NR_CPUS]; +#endif + +#if defined(CONFIG_SCHED_MC) && defined(CONFIG_SCHED_SMT) +static int cpu_to_core_group(int cpu) +{ + return first_cpu(cpu_sibling_map[cpu]); +} +#elif defined(CONFIG_SCHED_MC) +static int cpu_to_core_group(int cpu) +{ + return cpu; +} +#endif + static DEFINE_PER_CPU(struct sched_domain, phys_domains); -static struct sched_group sched_group_phys[NR_CPUS]; +static struct sched_group *sched_group_phys_bycpu[NR_CPUS]; + static int cpu_to_phys_group(int cpu) { -#ifdef CONFIG_SCHED_SMT +#ifdef CONFIG_SCHED_MC + cpumask_t mask = cpu_coregroup_map(cpu); + return first_cpu(mask); +#elif defined(CONFIG_SCHED_SMT) return first_cpu(cpu_sibling_map[cpu]); #else return cpu; @@ -5685,15 +6099,102 @@ static int cpu_to_allnodes_group(int cpu) { return cpu_to_node(cpu); } +static void init_numa_sched_groups_power(struct sched_group *group_head) +{ + struct sched_group *sg = group_head; + int j; + + if (!sg) + return; +next_sg: + for_each_cpu_mask(j, sg->cpumask) { + struct sched_domain *sd; + + sd = &per_cpu(phys_domains, j); + if (j != first_cpu(sd->groups->cpumask)) { + /* + * Only add "power" once for each + * physical package. + */ + continue; + } + + sg->cpu_power += sd->groups->cpu_power; + } + sg = sg->next; + if (sg != group_head) + goto next_sg; +} +#endif + +/* Free memory allocated for various sched_group structures */ +static void free_sched_groups(const cpumask_t *cpu_map) +{ + int cpu; +#ifdef CONFIG_NUMA + int i; + + for_each_cpu_mask(cpu, *cpu_map) { + struct sched_group *sched_group_allnodes + = sched_group_allnodes_bycpu[cpu]; + struct sched_group **sched_group_nodes + = sched_group_nodes_bycpu[cpu]; + + if (sched_group_allnodes) { + kfree(sched_group_allnodes); + sched_group_allnodes_bycpu[cpu] = NULL; + } + + if (!sched_group_nodes) + continue; + + for (i = 0; i < MAX_NUMNODES; i++) { + cpumask_t nodemask = node_to_cpumask(i); + struct sched_group *oldsg, *sg = sched_group_nodes[i]; + + cpus_and(nodemask, nodemask, *cpu_map); + if (cpus_empty(nodemask)) + continue; + + if (sg == NULL) + continue; + sg = sg->next; +next_sg: + oldsg = sg; + sg = sg->next; + kfree(oldsg); + if (oldsg != sched_group_nodes[i]) + goto next_sg; + } + kfree(sched_group_nodes); + sched_group_nodes_bycpu[cpu] = NULL; + } +#endif + for_each_cpu_mask(cpu, *cpu_map) { + if (sched_group_phys_bycpu[cpu]) { + kfree(sched_group_phys_bycpu[cpu]); + sched_group_phys_bycpu[cpu] = NULL; + } +#ifdef CONFIG_SCHED_MC + if (sched_group_core_bycpu[cpu]) { + kfree(sched_group_core_bycpu[cpu]); + sched_group_core_bycpu[cpu] = NULL; + } #endif + } +} /* * Build sched domains for a given set of cpus and attach the sched domains * to the individual cpus */ -void build_sched_domains(const cpumask_t *cpu_map) +static int build_sched_domains(const cpumask_t *cpu_map) { int i; + struct sched_group *sched_group_phys = NULL; +#ifdef CONFIG_SCHED_MC + struct sched_group *sched_group_core = NULL; +#endif #ifdef CONFIG_NUMA struct sched_group **sched_group_nodes = NULL; struct sched_group *sched_group_allnodes = NULL; @@ -5701,11 +6202,11 @@ void build_sched_domains(const cpumask_t *cpu_map) /* * Allocate the per-node list of sched groups */ - sched_group_nodes = kmalloc(sizeof(struct sched_group*)*MAX_NUMNODES, - GFP_ATOMIC); + sched_group_nodes = kzalloc(sizeof(struct sched_group*)*MAX_NUMNODES, + GFP_KERNEL); if (!sched_group_nodes) { printk(KERN_WARNING "Can not alloc sched group node list\n"); - return; + return -ENOMEM; } sched_group_nodes_bycpu[first_cpu(*cpu_map)] = sched_group_nodes; #endif @@ -5731,7 +6232,7 @@ void build_sched_domains(const cpumask_t *cpu_map) if (!sched_group_allnodes) { printk(KERN_WARNING "Can not alloc allnodes sched group\n"); - break; + goto error; } sched_group_allnodes_bycpu[i] = sched_group_allnodes; @@ -5752,6 +6253,18 @@ void build_sched_domains(const cpumask_t *cpu_map) cpus_and(sd->span, sd->span, *cpu_map); #endif + if (!sched_group_phys) { + sched_group_phys + = kmalloc(sizeof(struct sched_group) * NR_CPUS, + GFP_KERNEL); + if (!sched_group_phys) { + printk (KERN_WARNING "Can not alloc phys sched" + "group\n"); + goto error; + } + sched_group_phys_bycpu[i] = sched_group_phys; + } + p = sd; sd = &per_cpu(phys_domains, i); group = cpu_to_phys_group(i); @@ -5760,6 +6273,29 @@ void build_sched_domains(const cpumask_t *cpu_map) sd->parent = p; sd->groups = &sched_group_phys[group]; +#ifdef CONFIG_SCHED_MC + if (!sched_group_core) { + sched_group_core + = kmalloc(sizeof(struct sched_group) * NR_CPUS, + GFP_KERNEL); + if (!sched_group_core) { + printk (KERN_WARNING "Can not alloc core sched" + "group\n"); + goto error; + } + sched_group_core_bycpu[i] = sched_group_core; + } + + p = sd; + sd = &per_cpu(core_domains, i); + group = cpu_to_core_group(i); + *sd = SD_MC_INIT; + sd->span = cpu_coregroup_map(i); + cpus_and(sd->span, sd->span, *cpu_map); + sd->parent = p; + sd->groups = &sched_group_core[group]; +#endif + #ifdef CONFIG_SCHED_SMT p = sd; sd = &per_cpu(cpu_domains, i); @@ -5785,6 +6321,19 @@ void build_sched_domains(const cpumask_t *cpu_map) } #endif +#ifdef CONFIG_SCHED_MC + /* Set up multi-core groups */ + for_each_cpu_mask(i, *cpu_map) { + cpumask_t this_core_map = cpu_coregroup_map(i); + cpus_and(this_core_map, this_core_map, *cpu_map); + if (i != first_cpu(this_core_map)) + continue; + init_sched_build_groups(sched_group_core, this_core_map, + &cpu_to_core_group); + } +#endif + + /* Set up physical groups */ for (i = 0; i < MAX_NUMNODES; i++) { cpumask_t nodemask = node_to_cpumask(i); @@ -5820,24 +6369,21 @@ void build_sched_domains(const cpumask_t *cpu_map) domainspan = sched_domain_node_span(i); cpus_and(domainspan, domainspan, *cpu_map); - sg = kmalloc(sizeof(struct sched_group), GFP_KERNEL); + sg = kmalloc_node(sizeof(struct sched_group), GFP_KERNEL, i); + if (!sg) { + printk(KERN_WARNING "Can not alloc domain group for " + "node %d\n", i); + goto error; + } sched_group_nodes[i] = sg; for_each_cpu_mask(j, nodemask) { struct sched_domain *sd; sd = &per_cpu(node_domains, j); sd->groups = sg; - if (sd->groups == NULL) { - /* Turn off balancing if we have no groups */ - sd->flags = 0; - } - } - if (!sg) { - printk(KERN_WARNING - "Can not alloc domain group for node %d\n", i); - continue; } sg->cpu_power = 0; sg->cpumask = nodemask; + sg->next = sg; cpus_or(covered, covered, nodemask); prev = sg; @@ -5856,75 +6402,103 @@ void build_sched_domains(const cpumask_t *cpu_map) if (cpus_empty(tmp)) continue; - sg = kmalloc(sizeof(struct sched_group), GFP_KERNEL); + sg = kmalloc_node(sizeof(struct sched_group), + GFP_KERNEL, i); if (!sg) { printk(KERN_WARNING "Can not alloc domain group for node %d\n", j); - break; + goto error; } sg->cpu_power = 0; sg->cpumask = tmp; + sg->next = prev->next; cpus_or(covered, covered, tmp); prev->next = sg; prev = sg; } - prev->next = sched_group_nodes[i]; } #endif /* Calculate CPU power for physical packages and nodes */ +#ifdef CONFIG_SCHED_SMT for_each_cpu_mask(i, *cpu_map) { - int power; struct sched_domain *sd; -#ifdef CONFIG_SCHED_SMT sd = &per_cpu(cpu_domains, i); - power = SCHED_LOAD_SCALE; + sd->groups->cpu_power = SCHED_LOAD_SCALE; + } +#endif +#ifdef CONFIG_SCHED_MC + for_each_cpu_mask(i, *cpu_map) { + int power; + struct sched_domain *sd; + sd = &per_cpu(core_domains, i); + if (sched_smt_power_savings) + power = SCHED_LOAD_SCALE * cpus_weight(sd->groups->cpumask); + else + power = SCHED_LOAD_SCALE + (cpus_weight(sd->groups->cpumask)-1) + * SCHED_LOAD_SCALE / 10; sd->groups->cpu_power = power; + } #endif + for_each_cpu_mask(i, *cpu_map) { + struct sched_domain *sd; +#ifdef CONFIG_SCHED_MC sd = &per_cpu(phys_domains, i); - power = SCHED_LOAD_SCALE + SCHED_LOAD_SCALE * - (cpus_weight(sd->groups->cpumask)-1) / 10; - sd->groups->cpu_power = power; + if (i != first_cpu(sd->groups->cpumask)) + continue; -#ifdef CONFIG_NUMA - sd = &per_cpu(allnodes_domains, i); - if (sd->groups) { - power = SCHED_LOAD_SCALE + SCHED_LOAD_SCALE * - (cpus_weight(sd->groups->cpumask)-1) / 10; - sd->groups->cpu_power = power; - } + sd->groups->cpu_power = 0; + if (sched_mc_power_savings || sched_smt_power_savings) { + int j; + + for_each_cpu_mask(j, sd->groups->cpumask) { + struct sched_domain *sd1; + sd1 = &per_cpu(core_domains, j); + /* + * for each core we will add once + * to the group in physical domain + */ + if (j != first_cpu(sd1->groups->cpumask)) + continue; + + if (sched_smt_power_savings) + sd->groups->cpu_power += sd1->groups->cpu_power; + else + sd->groups->cpu_power += SCHED_LOAD_SCALE; + } + } else + /* + * This has to be < 2 * SCHED_LOAD_SCALE + * Lets keep it SCHED_LOAD_SCALE, so that + * while calculating NUMA group's cpu_power + * we can simply do + * numa_group->cpu_power += phys_group->cpu_power; + * + * See "only add power once for each physical pkg" + * comment below + */ + sd->groups->cpu_power = SCHED_LOAD_SCALE; +#else + int power; + sd = &per_cpu(phys_domains, i); + if (sched_smt_power_savings) + power = SCHED_LOAD_SCALE * cpus_weight(sd->groups->cpumask); + else + power = SCHED_LOAD_SCALE; + sd->groups->cpu_power = power; #endif } #ifdef CONFIG_NUMA - for (i = 0; i < MAX_NUMNODES; i++) { - struct sched_group *sg = sched_group_nodes[i]; - int j; + for (i = 0; i < MAX_NUMNODES; i++) + init_numa_sched_groups_power(sched_group_nodes[i]); - if (sg == NULL) - continue; -next_sg: - for_each_cpu_mask(j, sg->cpumask) { - struct sched_domain *sd; - int power; + if (sched_group_allnodes) { + int group = cpu_to_allnodes_group(first_cpu(*cpu_map)); + struct sched_group *sg = &sched_group_allnodes[group]; - sd = &per_cpu(phys_domains, j); - if (j != first_cpu(sd->groups->cpumask)) { - /* - * Only add "power" once for each - * physical package. - */ - continue; - } - power = SCHED_LOAD_SCALE + SCHED_LOAD_SCALE * - (cpus_weight(sd->groups->cpumask)-1) / 10; - - sg->cpu_power += power; - } - sg = sg->next; - if (sg != sched_group_nodes[i]) - goto next_sg; + init_numa_sched_groups_power(sg); } #endif @@ -5933,6 +6507,8 @@ next_sg: struct sched_domain *sd; #ifdef CONFIG_SCHED_SMT sd = &per_cpu(cpu_domains, i); +#elif defined(CONFIG_SCHED_MC) + sd = &per_cpu(core_domains, i); #else sd = &per_cpu(phys_domains, i); #endif @@ -5942,13 +6518,20 @@ next_sg: * Tune cache-hot values: */ calibrate_migration_costs(cpu_map); + + return 0; + +error: + free_sched_groups(cpu_map); + return -ENOMEM; } /* * Set up scheduler domains and groups. Callers must hold the hotplug lock. */ -static void arch_init_sched_domains(const cpumask_t *cpu_map) +static int arch_init_sched_domains(const cpumask_t *cpu_map) { cpumask_t cpu_default_map; + int err; /* * Setup mask for cpus without special case scheduling requirements. @@ -5957,51 +6540,14 @@ static void arch_init_sched_domains(const cpumask_t *cpu_map) */ cpus_andnot(cpu_default_map, *cpu_map, cpu_isolated_map); - build_sched_domains(&cpu_default_map); + err = build_sched_domains(&cpu_default_map); + + return err; } static void arch_destroy_sched_domains(const cpumask_t *cpu_map) { -#ifdef CONFIG_NUMA - int i; - int cpu; - - for_each_cpu_mask(cpu, *cpu_map) { - struct sched_group *sched_group_allnodes - = sched_group_allnodes_bycpu[cpu]; - struct sched_group **sched_group_nodes - = sched_group_nodes_bycpu[cpu]; - - if (sched_group_allnodes) { - kfree(sched_group_allnodes); - sched_group_allnodes_bycpu[cpu] = NULL; - } - - if (!sched_group_nodes) - continue; - - for (i = 0; i < MAX_NUMNODES; i++) { - cpumask_t nodemask = node_to_cpumask(i); - struct sched_group *oldsg, *sg = sched_group_nodes[i]; - - cpus_and(nodemask, nodemask, *cpu_map); - if (cpus_empty(nodemask)) - continue; - - if (sg == NULL) - continue; - sg = sg->next; -next_sg: - oldsg = sg; - sg = sg->next; - kfree(oldsg); - if (oldsg != sched_group_nodes[i]) - goto next_sg; - } - kfree(sched_group_nodes); - sched_group_nodes_bycpu[cpu] = NULL; - } -#endif + free_sched_groups(cpu_map); } /* @@ -6026,9 +6572,10 @@ static void detach_destroy_domains(const cpumask_t *cpu_map) * correct sched domains * Call with hotplug lock held */ -void partition_sched_domains(cpumask_t *partition1, cpumask_t *partition2) +int partition_sched_domains(cpumask_t *partition1, cpumask_t *partition2) { cpumask_t change_map; + int err = 0; cpus_and(*partition1, *partition1, cpu_online_map); cpus_and(*partition2, *partition2, cpu_online_map); @@ -6037,11 +6584,90 @@ void partition_sched_domains(cpumask_t *partition1, cpumask_t *partition2) /* Detach sched domains from all of the affected cpus */ detach_destroy_domains(&change_map); if (!cpus_empty(*partition1)) - build_sched_domains(partition1); - if (!cpus_empty(*partition2)) - build_sched_domains(partition2); + err = build_sched_domains(partition1); + if (!err && !cpus_empty(*partition2)) + err = build_sched_domains(partition2); + + return err; } +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) +int arch_reinit_sched_domains(void) +{ + int err; + + lock_cpu_hotplug(); + detach_destroy_domains(&cpu_online_map); + err = arch_init_sched_domains(&cpu_online_map); + unlock_cpu_hotplug(); + + return err; +} + +static ssize_t sched_power_savings_store(const char *buf, size_t count, int smt) +{ + int ret; + + if (buf[0] != '0' && buf[0] != '1') + return -EINVAL; + + if (smt) + sched_smt_power_savings = (buf[0] == '1'); + else + sched_mc_power_savings = (buf[0] == '1'); + + ret = arch_reinit_sched_domains(); + + return ret ? ret : count; +} + +int sched_create_sysfs_power_savings_entries(struct sysdev_class *cls) +{ + int err = 0; + +#ifdef CONFIG_SCHED_SMT + if (smt_capable()) + err = sysfs_create_file(&cls->kset.kobj, + &attr_sched_smt_power_savings.attr); +#endif +#ifdef CONFIG_SCHED_MC + if (!err && mc_capable()) + err = sysfs_create_file(&cls->kset.kobj, + &attr_sched_mc_power_savings.attr); +#endif + return err; +} +#endif + +#ifdef CONFIG_SCHED_MC +static ssize_t sched_mc_power_savings_show(struct sys_device *dev, char *page) +{ + return sprintf(page, "%u\n", sched_mc_power_savings); +} +static ssize_t sched_mc_power_savings_store(struct sys_device *dev, + const char *buf, size_t count) +{ + return sched_power_savings_store(buf, count, 0); +} +SYSDEV_ATTR(sched_mc_power_savings, 0644, sched_mc_power_savings_show, + sched_mc_power_savings_store); +#endif + +#ifdef CONFIG_SCHED_SMT +static ssize_t sched_smt_power_savings_show(struct sys_device *dev, char *page) +{ + return sprintf(page, "%u\n", sched_smt_power_savings); +} +static ssize_t sched_smt_power_savings_store(struct sys_device *dev, + const char *buf, size_t count) +{ + return sched_power_savings_store(buf, count, 1); +} +SYSDEV_ATTR(sched_smt_power_savings, 0644, sched_smt_power_savings_show, + sched_smt_power_savings_store); +#endif + + #ifdef CONFIG_HOTPLUG_CPU /* * Force a reinitialization of the sched domains hierarchy. The domains @@ -6095,6 +6721,7 @@ int in_sched_functions(unsigned long addr) { /* Linker adds these: start and end of __sched functions */ extern char __sched_text_start[], __sched_text_end[]; + return in_lock_functions(addr) || (addr >= (unsigned long)__sched_text_start && addr < (unsigned long)__sched_text_end); @@ -6102,14 +6729,15 @@ int in_sched_functions(unsigned long addr) void __init sched_init(void) { - runqueue_t *rq; int i, j, k; - for (i = 0; i < NR_CPUS; i++) { - prio_array_t *array; + for_each_possible_cpu(i) { + struct prio_array *array; + struct rq *rq; rq = cpu_rq(i); spin_lock_init(&rq->lock); + lockdep_set_class(&rq->lock, &rq->rq_lock_key); rq->nr_running = 0; rq->active = rq->arrays; rq->expired = rq->arrays + 1; @@ -6137,6 +6765,12 @@ void __init sched_init(void) } } + set_load_weight(&init_task); + +#ifdef CONFIG_RT_MUTEXES + plist_head_init(&init_task.pi_waiters, &init_task.pi_lock); +#endif + /* * The boot idle thread does lazy MMU switching as well: */ @@ -6155,7 +6789,7 @@ void __init sched_init(void) #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP void __might_sleep(char *file, int line) { -#if defined(in_atomic) +#ifdef in_atomic static unsigned long prev_jiffy; /* ratelimiting */ if ((in_atomic() || irqs_disabled()) && @@ -6163,7 +6797,7 @@ void __might_sleep(char *file, int line) if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy) return; prev_jiffy = jiffies; - printk(KERN_ERR "Debug: sleeping function called from invalid" + printk(KERN_ERR "BUG: sleeping function called from invalid" " context at %s:%d\n", file, line); printk("in_atomic():%d, irqs_disabled():%d\n", in_atomic(), irqs_disabled()); @@ -6177,17 +6811,18 @@ EXPORT_SYMBOL(__might_sleep); #ifdef CONFIG_MAGIC_SYSRQ void normalize_rt_tasks(void) { + struct prio_array *array; struct task_struct *p; - prio_array_t *array; unsigned long flags; - runqueue_t *rq; + struct rq *rq; read_lock_irq(&tasklist_lock); - for_each_process (p) { + for_each_process(p) { if (!rt_task(p)) continue; - rq = task_rq_lock(p, &flags); + spin_lock_irqsave(&p->pi_lock, flags); + rq = __task_rq_lock(p); array = p->array; if (array) @@ -6198,7 +6833,8 @@ void normalize_rt_tasks(void) resched_task(rq->curr); } - task_rq_unlock(rq, &flags); + __task_rq_unlock(rq); + spin_unlock_irqrestore(&p->pi_lock, flags); } read_unlock_irq(&tasklist_lock); } @@ -6222,7 +6858,7 @@ void normalize_rt_tasks(void) * * ONLY VALID WHEN THE WHOLE SYSTEM IS STOPPED! */ -task_t *curr_task(int cpu) +struct task_struct *curr_task(int cpu) { return cpu_curr(cpu); } @@ -6242,7 +6878,7 @@ task_t *curr_task(int cpu) * * ONLY VALID WHEN THE WHOLE SYSTEM IS STOPPED! */ -void set_curr_task(int cpu, task_t *p) +void set_curr_task(int cpu, struct task_struct *p) { cpu_curr(cpu) = p; }