/*
* Targeted preemption latency for CPU-bound tasks:
- * (default: 20ms, units: nanoseconds)
+ * (default: 20ms * (1 + ilog(ncpus)), units: nanoseconds)
*
* NOTE: this latency value is not the same as the concept of
* 'timeslice length' - timeslices in CFS are of variable length
* (to see the precise effective timeslice length of your workload,
* run vmstat and monitor the context-switches (cs) field)
*/
-const_debug unsigned int sysctl_sched_latency = 20000000ULL;
+unsigned int sysctl_sched_latency = 20000000ULL;
/*
- * After fork, child runs first. (default) If set to 0 then
- * parent will (try to) run first.
+ * Minimal preemption granularity for CPU-bound tasks:
+ * (default: 4 msec * (1 + ilog(ncpus)), units: nanoseconds)
*/
-const_debug unsigned int sysctl_sched_child_runs_first = 1;
+unsigned int sysctl_sched_min_granularity = 4000000ULL;
/*
- * Minimal preemption granularity for CPU-bound tasks:
- * (default: 2 msec, units: nanoseconds)
+ * is kept at sysctl_sched_latency / sysctl_sched_min_granularity
+ */
+static unsigned int sched_nr_latency = 5;
+
+/*
+ * After fork, child runs first. (default) If set to 0 then
+ * parent will (try to) run first.
*/
-const_debug unsigned int sysctl_sched_nr_latency = 20;
+const_debug unsigned int sysctl_sched_child_runs_first = 1;
/*
* sys_sched_yield() compat mode
/*
* SCHED_BATCH wake-up granularity.
- * (default: 10 msec, units: nanoseconds)
+ * (default: 10 msec * (1 + ilog(ncpus)), units: nanoseconds)
*
* This option delays the preemption effects of decoupled workloads
* and reduces their over-scheduling. Synchronous workloads will still
* have immediate wakeup/sleep latencies.
*/
-const_debug unsigned int sysctl_sched_batch_wakeup_granularity = 10000000UL;
+unsigned int sysctl_sched_batch_wakeup_granularity = 10000000UL;
/*
* SCHED_OTHER wake-up granularity.
- * (default: 10 msec, units: nanoseconds)
+ * (default: 10 msec * (1 + ilog(ncpus)), units: nanoseconds)
*
* This option delays the preemption effects of decoupled workloads
* and reduces their over-scheduling. Synchronous workloads will still
* have immediate wakeup/sleep latencies.
*/
-const_debug unsigned int sysctl_sched_wakeup_granularity = 10000000UL;
+unsigned int sysctl_sched_wakeup_granularity = 10000000UL;
const_debug unsigned int sysctl_sched_migration_cost = 500000UL;
* Scheduling class statistics methods:
*/
+#ifdef CONFIG_SCHED_DEBUG
+int sched_nr_latency_handler(struct ctl_table *table, int write,
+ struct file *filp, void __user *buffer, size_t *lenp,
+ loff_t *ppos)
+{
+ int ret = proc_dointvec_minmax(table, write, filp, buffer, lenp, ppos);
+
+ if (ret || !write)
+ return ret;
+
+ sched_nr_latency = DIV_ROUND_UP(sysctl_sched_latency,
+ sysctl_sched_min_granularity);
+
+ return 0;
+}
+#endif
/*
* The idea is to set a period in which each task runs once.
static u64 __sched_period(unsigned long nr_running)
{
u64 period = sysctl_sched_latency;
- unsigned long nr_latency = sysctl_sched_nr_latency;
+ unsigned long nr_latency = sched_nr_latency;
if (unlikely(nr_running > nr_latency)) {
period *= nr_running;
{
u64 vslice = __sched_period(nr_running);
+ vslice *= NICE_0_LOAD;
do_div(vslice, rq_weight);
return vslice;
__update_curr(cfs_rq, curr, delta_exec);
curr->exec_start = now;
+
+ if (entity_is_task(curr)) {
+ struct task_struct *curtask = task_of(curr);
+
+ cpuacct_charge(curtask, delta_exec);
+ }
}
static inline void
} else if (sched_feat(APPROX_AVG) && cfs_rq->nr_running)
vruntime += sched_vslice(cfs_rq)/2;
+ /*
+ * The 'current' period is already promised to the current tasks,
+ * however the extra weight of the new task will slow them down a
+ * little, place the new task so that it fits in the slot that
+ * stays open at the end.
+ */
if (initial && sched_feat(START_DEBIT))
vruntime += sched_vslice_add(cfs_rq, se);
if (!initial) {
- if (sched_feat(NEW_FAIR_SLEEPERS) && entity_is_task(se) &&
- task_of(se)->policy != SCHED_BATCH)
+ /* sleeps upto a single latency don't count. */
+ if (sched_feat(NEW_FAIR_SLEEPERS) && entity_is_task(se))
vruntime -= sysctl_sched_latency;
- vruntime = max_t(s64, vruntime, se->vruntime);
+ /* ensure we never gain time by being placed backwards. */
+ vruntime = max_vruntime(se->vruntime, vruntime);
}
se->vruntime = vruntime;
-
}
static void
update_stats_dequeue(cfs_rq, se);
if (sleep) {
- se->peer_preempt = 0;
#ifdef CONFIG_SCHEDSTATS
if (entity_is_task(se)) {
struct task_struct *tsk = task_of(se);
ideal_runtime = sched_slice(cfs_rq, curr);
delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime;
- if (delta_exec > ideal_runtime ||
- (sched_feat(PREEMPT_RESTRICT) && curr->peer_preempt))
+ if (delta_exec > ideal_runtime)
resched_task(rq_of(cfs_rq)->curr);
- curr->peer_preempt = 0;
}
static void
/* Iterate thr' all leaf cfs_rq's on a runqueue */
#define for_each_leaf_cfs_rq(rq, cfs_rq) \
- list_for_each_entry(cfs_rq, &rq->leaf_cfs_rq_list, leaf_cfs_rq_list)
+ list_for_each_entry_rcu(cfs_rq, &rq->leaf_cfs_rq_list, leaf_cfs_rq_list)
/* Do the two (enqueued) entities belong to the same group ? */
static inline int
return se->parent;
}
+#define GROUP_IMBALANCE_PCT 20
+
#else /* CONFIG_FAIR_GROUP_SCHED */
#define for_each_sched_entity(se) \
static void enqueue_task_fair(struct rq *rq, struct task_struct *p, int wakeup)
{
struct cfs_rq *cfs_rq;
- struct sched_entity *se = &p->se;
+ struct sched_entity *se = &p->se,
+ *topse = NULL; /* Highest schedulable entity */
+ int incload = 1;
for_each_sched_entity(se) {
- if (se->on_rq)
+ topse = se;
+ if (se->on_rq) {
+ incload = 0;
break;
+ }
cfs_rq = cfs_rq_of(se);
enqueue_entity(cfs_rq, se, wakeup);
wakeup = 1;
}
+ /* Increment cpu load if we just enqueued the first task of a group on
+ * 'rq->cpu'. 'topse' represents the group to which task 'p' belongs
+ * at the highest grouping level.
+ */
+ if (incload)
+ inc_cpu_load(rq, topse->load.weight);
}
/*
static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int sleep)
{
struct cfs_rq *cfs_rq;
- struct sched_entity *se = &p->se;
+ struct sched_entity *se = &p->se,
+ *topse = NULL; /* Highest schedulable entity */
+ int decload = 1;
for_each_sched_entity(se) {
+ topse = se;
cfs_rq = cfs_rq_of(se);
dequeue_entity(cfs_rq, se, sleep);
/* Don't dequeue parent if it has other entities besides us */
- if (cfs_rq->load.weight)
+ if (cfs_rq->load.weight) {
+ if (parent_entity(se))
+ decload = 0;
break;
+ }
sleep = 1;
}
+ /* Decrement cpu load if we just dequeued the last task of a group on
+ * 'rq->cpu'. 'topse' represents the group to which task 'p' belongs
+ * at the highest grouping level.
+ */
+ if (decload)
+ dec_cpu_load(rq, topse->load.weight);
}
/*
*/
static void yield_task_fair(struct rq *rq)
{
- struct cfs_rq *cfs_rq = task_cfs_rq(rq->curr);
- struct sched_entity *rightmost, *se = &rq->curr->se;
+ struct task_struct *curr = rq->curr;
+ struct cfs_rq *cfs_rq = task_cfs_rq(curr);
+ struct sched_entity *rightmost, *se = &curr->se;
/*
* Are we the only task in the tree?
if (unlikely(cfs_rq->nr_running == 1))
return;
- if (likely(!sysctl_sched_compat_yield)) {
+ if (likely(!sysctl_sched_compat_yield) && curr->policy != SCHED_BATCH) {
__update_rq_clock(rq);
/*
* Update run-time statistics of the 'current'.
struct task_struct *curr = rq->curr;
struct cfs_rq *cfs_rq = task_cfs_rq(curr);
struct sched_entity *se = &curr->se, *pse = &p->se;
- s64 delta, gran;
+ unsigned long gran;
if (unlikely(rt_prio(p->prio))) {
update_rq_clock(rq);
if (unlikely(p->policy == SCHED_BATCH))
return;
- if (sched_feat(WAKEUP_PREEMPT)) {
- while (!is_same_group(se, pse)) {
- se = parent_entity(se);
- pse = parent_entity(pse);
- }
+ if (!sched_feat(WAKEUP_PREEMPT))
+ return;
- delta = se->vruntime - pse->vruntime;
- gran = sysctl_sched_wakeup_granularity;
- if (unlikely(se->load.weight != NICE_0_LOAD))
- gran = calc_delta_fair(gran, &se->load);
+ while (!is_same_group(se, pse)) {
+ se = parent_entity(se);
+ pse = parent_entity(pse);
+ }
- if (delta > gran) {
- int now = !sched_feat(PREEMPT_RESTRICT);
+ gran = sysctl_sched_wakeup_granularity;
+ if (unlikely(se->load.weight != NICE_0_LOAD))
+ gran = calc_delta_fair(gran, &se->load);
- if (now || p->prio < curr->prio || !se->peer_preempt++)
- resched_task(curr);
- }
- }
+ if (pse->vruntime + gran < se->vruntime)
+ resched_task(curr);
}
static struct task_struct *pick_next_task_fair(struct rq *rq)
return __load_balance_iterator(cfs_rq, cfs_rq->rb_load_balance_curr);
}
-#ifdef CONFIG_FAIR_GROUP_SCHED
-static int cfs_rq_best_prio(struct cfs_rq *cfs_rq)
-{
- struct sched_entity *curr;
- struct task_struct *p;
-
- if (!cfs_rq->nr_running)
- return MAX_PRIO;
-
- curr = cfs_rq->curr;
- if (!curr)
- curr = __pick_next_entity(cfs_rq);
-
- p = task_of(curr);
-
- return p->prio;
-}
-#endif
-
static unsigned long
load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
unsigned long max_load_move,
struct cfs_rq *busy_cfs_rq;
long rem_load_move = max_load_move;
struct rq_iterator cfs_rq_iterator;
+ unsigned long load_moved;
cfs_rq_iterator.start = load_balance_start_fair;
cfs_rq_iterator.next = load_balance_next_fair;
for_each_leaf_cfs_rq(busiest, busy_cfs_rq) {
#ifdef CONFIG_FAIR_GROUP_SCHED
- struct cfs_rq *this_cfs_rq;
- long imbalance;
- unsigned long maxload;
+ struct cfs_rq *this_cfs_rq = busy_cfs_rq->tg->cfs_rq[this_cpu];
+ unsigned long maxload, task_load, group_weight;
+ unsigned long thisload, per_task_load;
+ struct sched_entity *se = busy_cfs_rq->tg->se[busiest->cpu];
+
+ task_load = busy_cfs_rq->load.weight;
+ group_weight = se->load.weight;
- this_cfs_rq = cpu_cfs_rq(busy_cfs_rq, this_cpu);
+ /*
+ * 'group_weight' is contributed by tasks of total weight
+ * 'task_load'. To move 'rem_load_move' worth of weight only,
+ * we need to move a maximum task load of:
+ *
+ * maxload = (remload / group_weight) * task_load;
+ */
+ maxload = (rem_load_move * task_load) / group_weight;
- imbalance = busy_cfs_rq->load.weight - this_cfs_rq->load.weight;
- /* Don't pull if this_cfs_rq has more load than busy_cfs_rq */
- if (imbalance <= 0)
+ if (!maxload || !task_load)
continue;
- /* Don't pull more than imbalance/2 */
- imbalance /= 2;
- maxload = min(rem_load_move, imbalance);
+ per_task_load = task_load / busy_cfs_rq->nr_running;
+ /*
+ * balance_tasks will try to forcibly move atleast one task if
+ * possible (because of SCHED_LOAD_SCALE_FUZZ). Avoid that if
+ * maxload is less than GROUP_IMBALANCE_FUZZ% the per_task_load.
+ */
+ if (100 * maxload < GROUP_IMBALANCE_PCT * per_task_load)
+ continue;
- *this_best_prio = cfs_rq_best_prio(this_cfs_rq);
+ /* Disable priority-based load balance */
+ *this_best_prio = 0;
+ thisload = this_cfs_rq->load.weight;
#else
# define maxload rem_load_move
#endif
* load_balance_[start|next]_fair iterators
*/
cfs_rq_iterator.arg = busy_cfs_rq;
- rem_load_move -= balance_tasks(this_rq, this_cpu, busiest,
+ load_moved = balance_tasks(this_rq, this_cpu, busiest,
maxload, sd, idle, all_pinned,
this_best_prio,
&cfs_rq_iterator);
+#ifdef CONFIG_FAIR_GROUP_SCHED
+ /*
+ * load_moved holds the task load that was moved. The
+ * effective (group) weight moved would be:
+ * load_moved_eff = load_moved/task_load * group_weight;
+ */
+ load_moved = (group_weight * load_moved) / task_load;
+
+ /* Adjust shares on both cpus to reflect load_moved */
+ group_weight -= load_moved;
+ set_se_shares(se, group_weight);
+
+ se = busy_cfs_rq->tg->se[this_cpu];
+ if (!thisload)
+ group_weight = load_moved;
+ else
+ group_weight = se->load.weight + load_moved;
+ set_se_shares(se, group_weight);
+#endif
+
+ rem_load_move -= load_moved;
+
if (rem_load_move <= 0)
break;
}
}
}
-#define swap(a,b) do { typeof(a) tmp = (a); (a) = (b); (b) = tmp; } while (0)
+#define swap(a, b) do { typeof(a) tmp = (a); (a) = (b); (b) = tmp; } while (0)
/*
* Share the fairness runtime between parent and child, thus the
update_curr(cfs_rq);
place_entity(cfs_rq, se, 1);
+ /* 'curr' will be NULL if the child belongs to a different group */
if (sysctl_sched_child_runs_first && this_cpu == task_cpu(p) &&
- curr->vruntime < se->vruntime) {
+ curr && curr->vruntime < se->vruntime) {
/*
* Upon rescheduling, sched_class::put_prev_task() will place
* 'current' within the tree based on its new key value.
swap(curr->vruntime, se->vruntime);
}
- se->peer_preempt = 0;
enqueue_task_fair(rq, p, 0);
resched_task(rq->curr);
}
#ifdef CONFIG_FAIR_GROUP_SCHED
print_cfs_rq(m, cpu, &cpu_rq(cpu)->cfs);
#endif
+ lock_task_group_list();
for_each_leaf_cfs_rq(cpu_rq(cpu), cfs_rq)
print_cfs_rq(m, cpu, cfs_rq);
+ unlock_task_group_list();
}
#endif
projects / linux-2.6-omap-h63xx.git / blobdiff