* Copyright (C) 2007 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
*/
-/*
- * Tunables that become constants when CONFIG_SCHED_DEBUG is off:
- */
-#ifdef CONFIG_SCHED_DEBUG
-# define const_debug __read_mostly
-#else
-# define const_debug static const
-#endif
-
/*
* Targeted preemption latency for CPU-bound tasks:
* (default: 20ms, units: nanoseconds)
* and reduces their over-scheduling. Synchronous workloads will still
* have immediate wakeup/sleep latencies.
*/
-const_debug unsigned int sysctl_sched_wakeup_granularity = 1000000UL;
+const_debug unsigned int sysctl_sched_wakeup_granularity = 2000000UL;
unsigned int sysctl_sched_runtime_limit __read_mostly;
-/*
- * Debugging: various feature bits
- */
-enum {
- SCHED_FEAT_FAIR_SLEEPERS = 1,
- SCHED_FEAT_NEW_FAIR_SLEEPERS = 2,
- SCHED_FEAT_SLEEPER_AVG = 4,
- SCHED_FEAT_SLEEPER_LOAD_AVG = 8,
- SCHED_FEAT_START_DEBIT = 16,
- SCHED_FEAT_SKIP_INITIAL = 32,
-};
-
-const_debug unsigned int sysctl_sched_features =
- SCHED_FEAT_FAIR_SLEEPERS *0 |
- SCHED_FEAT_NEW_FAIR_SLEEPERS *1 |
- SCHED_FEAT_SLEEPER_AVG *0 |
- SCHED_FEAT_SLEEPER_LOAD_AVG *1 |
- SCHED_FEAT_START_DEBIT *1 |
- SCHED_FEAT_SKIP_INITIAL *0;
-
-#define sched_feat(x) (sysctl_sched_features & SCHED_FEAT_##x)
-
extern struct sched_class fair_sched_class;
/**************************************************************
* Scheduling class statistics methods:
*/
-/*
- * Calculate the preemption granularity needed to schedule every
- * runnable task once per sysctl_sched_latency amount of time.
- * (down to a sensible low limit on granularity)
- *
- * For example, if there are 2 tasks running and latency is 10 msecs,
- * we switch tasks every 5 msecs. If we have 3 tasks running, we have
- * to switch tasks every 3.33 msecs to get a 10 msecs observed latency
- * for each task. We do finer and finer scheduling up to until we
- * reach the minimum granularity value.
- *
- * To achieve this we use the following dynamic-granularity rule:
- *
- * gran = lat/nr - lat/nr/nr
- *
- * This comes out of the following equations:
- *
- * kA1 + gran = kB1
- * kB2 + gran = kA2
- * kA2 = kA1
- * kB2 = kB1 - d + d/nr
- * lat = d * nr
- *
- * Where 'k' is key, 'A' is task A (waiting), 'B' is task B (running),
- * '1' is start of time, '2' is end of time, 'd' is delay between
- * 1 and 2 (during which task B was running), 'nr' is number of tasks
- * running, 'lat' is the the period of each task. ('lat' is the
- * sched_latency that we aim for.)
- */
-static long
-sched_granularity(struct cfs_rq *cfs_rq)
+static u64 __sched_period(unsigned long nr_running)
{
- unsigned int gran = sysctl_sched_latency;
- unsigned int nr = cfs_rq->nr_running;
+ u64 period = sysctl_sched_latency;
+ unsigned long nr_latency =
+ sysctl_sched_latency / sysctl_sched_min_granularity;
- if (nr > 1) {
- gran = gran/nr - gran/nr/nr;
- gran = max(gran, sysctl_sched_min_granularity);
+ if (unlikely(nr_running > nr_latency)) {
+ period *= nr_running;
+ do_div(period, nr_latency);
}
- return gran;
+ return period;
}
-/*
- * We rescale the rescheduling granularity of tasks according to their
- * nice level, but only linearly, not exponentially:
- */
-static long
-niced_granularity(struct sched_entity *curr, unsigned long granularity)
+static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se)
{
- u64 tmp;
+ u64 period = __sched_period(cfs_rq->nr_running);
- if (likely(curr->load.weight == NICE_0_LOAD))
- return granularity;
- /*
- * Positive nice levels get the same granularity as nice-0:
- */
- if (likely(curr->load.weight < NICE_0_LOAD)) {
- tmp = curr->load.weight * (u64)granularity;
- return (long) (tmp >> NICE_0_SHIFT);
- }
- /*
- * Negative nice level tasks get linearly finer
- * granularity:
- */
- tmp = curr->load.inv_weight * (u64)granularity;
+ period *= se->load.weight;
+ do_div(period, cfs_rq->load.weight);
- /*
- * It will always fit into 'long':
- */
- return (long) (tmp >> (WMULT_SHIFT-NICE_0_SHIFT));
+ return period;
}
static inline void
}
curr->vruntime += delta_exec_weighted;
+ if (!sched_feat(FAIR_SLEEPERS))
+ return;
+
if (unlikely(!load))
return;
* Preempt the current task with a newly woken task if needed:
*/
static void
-__check_preempt_curr_fair(struct cfs_rq *cfs_rq, struct sched_entity *se,
- struct sched_entity *curr, unsigned long granularity)
+check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr)
{
- s64 __delta = curr->fair_key - se->fair_key;
unsigned long ideal_runtime, delta_exec;
- /*
- * ideal_runtime is compared against sum_exec_runtime, which is
- * walltime, hence do not scale.
- */
- ideal_runtime = max(sysctl_sched_latency / cfs_rq->nr_running,
- (unsigned long)sysctl_sched_min_granularity);
-
- /*
- * If we executed more than what the latency constraint suggests,
- * reduce the rescheduling granularity. This way the total latency
- * of how much a task is not scheduled converges to
- * sysctl_sched_latency:
- */
+ ideal_runtime = sched_slice(cfs_rq, curr);
delta_exec = curr->sum_exec_runtime - curr->prev_sum_exec_runtime;
if (delta_exec > ideal_runtime)
- granularity = 0;
-
- /*
- * Take scheduling granularity into account - do not
- * preempt the current task unless the best task has
- * a larger than sched_granularity fairness advantage:
- *
- * scale granularity as key space is in fair_clock.
- */
- if (__delta > niced_granularity(curr, granularity))
resched_task(rq_of(cfs_rq)->curr);
}
static void entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr)
{
- struct sched_entity *next;
-
/*
* Dequeue and enqueue the task to update its
* position within the tree:
dequeue_entity(cfs_rq, curr, 0);
enqueue_entity(cfs_rq, curr, 0);
- /*
- * Reschedule if another task tops the current one.
- */
- next = __pick_next_entity(cfs_rq);
- if (next == curr)
- return;
-
- __check_preempt_curr_fair(cfs_rq, next, curr,
- sched_granularity(cfs_rq));
+ if (cfs_rq->nr_running > 1)
+ check_preempt_tick(cfs_rq, curr);
}
/**************************************************
/*
* Preempt the current task with a newly woken task if needed:
*/
-static void check_preempt_curr_fair(struct rq *rq, struct task_struct *p)
+static void check_preempt_wakeup(struct rq *rq, struct task_struct *p)
{
struct task_struct *curr = rq->curr;
struct cfs_rq *cfs_rq = task_cfs_rq(curr);
- unsigned long gran;
if (unlikely(rt_prio(p->prio))) {
update_rq_clock(rq);
resched_task(curr);
return;
}
+ if (is_same_group(curr, p)) {
+ s64 delta = curr->se.vruntime - p->se.vruntime;
- gran = sysctl_sched_wakeup_granularity;
- /*
- * Batch tasks prefer throughput over latency:
- */
- if (unlikely(p->policy == SCHED_BATCH))
- gran = sysctl_sched_batch_wakeup_granularity;
-
- if (is_same_group(curr, p))
- __check_preempt_curr_fair(cfs_rq, &p->se, &curr->se, gran);
+ if (delta > (s64)sysctl_sched_wakeup_granularity)
+ resched_task(curr);
+ }
}
static struct task_struct *pick_next_task_fair(struct rq *rq)
}
}
+#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
* total amount of pressure for CPU stays equal - new tasks
sched_info_queued(p);
update_curr(cfs_rq);
+ se->vruntime = cfs_rq->min_vruntime;
update_stats_enqueue(cfs_rq, se);
- /*
- * Child runs first: we let it run before the parent
- * until it reschedules once. We set up the key so that
- * it will preempt the parent:
- */
- se->fair_key = curr->fair_key -
- niced_granularity(curr, sched_granularity(cfs_rq)) - 1;
+
/*
* The first wait is dominated by the child-runs-first logic,
* so do not credit it with that waiting time yet:
* -granularity/2, so initialize the task with that:
*/
if (sched_feat(START_DEBIT))
- se->wait_runtime = -(sched_granularity(cfs_rq) / 2);
+ se->wait_runtime = -(__sched_period(cfs_rq->nr_running+1) / 2);
+
+ if (sysctl_sched_child_runs_first &&
+ curr->vruntime < se->vruntime) {
+
+ dequeue_entity(cfs_rq, curr, 0);
+ swap(curr->vruntime, se->vruntime);
+ enqueue_entity(cfs_rq, curr, 0);
+ }
- se->vruntime = cfs_rq->min_vruntime;
update_stats_enqueue(cfs_rq, se);
__enqueue_entity(cfs_rq, se);
resched_task(rq->curr);
.dequeue_task = dequeue_task_fair,
.yield_task = yield_task_fair,
- .check_preempt_curr = check_preempt_curr_fair,
+ .check_preempt_curr = check_preempt_wakeup,
.pick_next_task = pick_next_task_fair,
.put_prev_task = put_prev_task_fair,