sched/fair: Optimize cgroup pick_next_task_fair()
authorPeter Zijlstra <peterz@infradead.org>
Sat, 11 Feb 2012 05:05:00 +0000 (06:05 +0100)
committerIngo Molnar <mingo@kernel.org>
Mon, 10 Feb 2014 15:17:19 +0000 (16:17 +0100)
Since commit 2f36825b1 ("sched: Next buddy hint on sleep and preempt
path") it is likely we pick a new task from the same cgroup, doing a put
and then set on all intermediate entities is a waste of time, so try to
avoid this.

Measured using:

  mount nodev /cgroup -t cgroup -o cpu
  cd /cgroup
  mkdir a; cd a
  mkdir b; cd b
  mkdir c; cd c
  echo $$ > tasks
  perf stat --repeat 10 -- taskset 1 perf bench sched pipe

PRE :      4.542422684 seconds time elapsed   ( +-  0.33% )
POST:      4.389409991 seconds time elapsed   ( +-  0.32% )

Which shows a significant improvement of ~3.5%

Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/1328936700.2476.17.camel@laptop
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/sched/fair.c

index 846172107ba5939153d7c129dd6a4683412343d5..a81b241ff70fc42bd164a7070bc3cf1b19ecd318 100644 (file)
@@ -2907,17 +2907,36 @@ wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se);
  * 3) pick the "last" process, for cache locality
  * 4) do not run the "skip" process, if something else is available
  */
-static struct sched_entity *pick_next_entity(struct cfs_rq *cfs_rq)
+static struct sched_entity *
+pick_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *curr)
 {
-       struct sched_entity *se = __pick_first_entity(cfs_rq);
-       struct sched_entity *left = se;
+       struct sched_entity *left = __pick_first_entity(cfs_rq);
+       struct sched_entity *se;
+
+       /*
+        * If curr is set we have to see if its left of the leftmost entity
+        * still in the tree, provided there was anything in the tree at all.
+        */
+       if (!left || (curr && entity_before(curr, left)))
+               left = curr;
+
+       se = left; /* ideally we run the leftmost entity */
 
        /*
         * Avoid running the skip buddy, if running something else can
         * be done without getting too unfair.
         */
        if (cfs_rq->skip == se) {
-               struct sched_entity *second = __pick_next_entity(se);
+               struct sched_entity *second;
+
+               if (se == curr) {
+                       second = __pick_first_entity(cfs_rq);
+               } else {
+                       second = __pick_next_entity(se);
+                       if (!second || (curr && entity_before(curr, second)))
+                               second = curr;
+               }
+
                if (second && wakeup_preempt_entity(second, left) < 1)
                        se = second;
        }
@@ -2939,7 +2958,7 @@ static struct sched_entity *pick_next_entity(struct cfs_rq *cfs_rq)
        return se;
 }
 
-static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq);
+static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq);
 
 static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev)
 {
@@ -3594,22 +3613,23 @@ static void check_enqueue_throttle(struct cfs_rq *cfs_rq)
 }
 
 /* conditionally throttle active cfs_rq's from put_prev_entity() */
-static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq)
+static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq)
 {
        if (!cfs_bandwidth_used())
-               return;
+               return false;
 
        if (likely(!cfs_rq->runtime_enabled || cfs_rq->runtime_remaining > 0))
-               return;
+               return false;
 
        /*
         * it's possible for a throttled entity to be forced into a running
         * state (e.g. set_curr_task), in this case we're finished.
         */
        if (cfs_rq_throttled(cfs_rq))
-               return;
+               return true;
 
        throttle_cfs_rq(cfs_rq);
+       return true;
 }
 
 static enum hrtimer_restart sched_cfs_slack_timer(struct hrtimer *timer)
@@ -3719,7 +3739,7 @@ static inline u64 cfs_rq_clock_task(struct cfs_rq *cfs_rq)
 }
 
 static void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) {}
-static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq) {}
+static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq) { return false; }
 static void check_enqueue_throttle(struct cfs_rq *cfs_rq) {}
 static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) {}
 
@@ -4658,9 +4678,86 @@ preempt:
 static struct task_struct *
 pick_next_task_fair(struct rq *rq, struct task_struct *prev)
 {
-       struct task_struct *p;
        struct cfs_rq *cfs_rq = &rq->cfs;
        struct sched_entity *se;
+       struct task_struct *p;
+
+#ifdef CONFIG_FAIR_GROUP_SCHED
+       if (!cfs_rq->nr_running)
+               return NULL;
+
+       if (!prev || prev->sched_class != &fair_sched_class)
+               goto simple;
+
+       /*
+        * Because of the set_next_buddy() in dequeue_task_fair() it is rather
+        * likely that a next task is from the same cgroup as the current.
+        *
+        * Therefore attempt to avoid putting and setting the entire cgroup
+        * hierarchy, only change the part that actually changes.
+        */
+
+       do {
+               struct sched_entity *curr = cfs_rq->curr;
+
+               /*
+                * Since we got here without doing put_prev_entity() we also
+                * have to consider cfs_rq->curr. If it is still a runnable
+                * entity, update_curr() will update its vruntime, otherwise
+                * forget we've ever seen it.
+                */
+               if (curr && curr->on_rq)
+                       update_curr(cfs_rq);
+               else
+                       curr = NULL;
+
+               /*
+                * This call to check_cfs_rq_runtime() will do the throttle and
+                * dequeue its entity in the parent(s). Therefore the 'simple'
+                * nr_running test will indeed be correct.
+                */
+               if (unlikely(check_cfs_rq_runtime(cfs_rq)))
+                       goto simple;
+
+               se = pick_next_entity(cfs_rq, curr);
+               cfs_rq = group_cfs_rq(se);
+       } while (cfs_rq);
+
+       p = task_of(se);
+
+       /*
+        * Since we haven't yet done put_prev_entity and if the selected task
+        * is a different task than we started out with, try and touch the
+        * least amount of cfs_rqs.
+        */
+       if (prev != p) {
+               struct sched_entity *pse = &prev->se;
+
+               while (!(cfs_rq = is_same_group(se, pse))) {
+                       int se_depth = se->depth;
+                       int pse_depth = pse->depth;
+
+                       if (se_depth <= pse_depth) {
+                               put_prev_entity(cfs_rq_of(pse), pse);
+                               pse = parent_entity(pse);
+                       }
+                       if (se_depth >= pse_depth) {
+                               set_next_entity(cfs_rq_of(se), se);
+                               se = parent_entity(se);
+                       }
+               }
+
+               put_prev_entity(cfs_rq, pse);
+               set_next_entity(cfs_rq, se);
+       }
+
+       if (hrtick_enabled(rq))
+               hrtick_start_fair(rq, p);
+
+       return p;
+simple:
+       cfs_rq = &rq->cfs;
+#endif
 
        if (!cfs_rq->nr_running)
                return NULL;
@@ -4669,12 +4766,13 @@ pick_next_task_fair(struct rq *rq, struct task_struct *prev)
                prev->sched_class->put_prev_task(rq, prev);
 
        do {
-               se = pick_next_entity(cfs_rq);
+               se = pick_next_entity(cfs_rq, NULL);
                set_next_entity(cfs_rq, se);
                cfs_rq = group_cfs_rq(se);
        } while (cfs_rq);
 
        p = task_of(se);
+
        if (hrtick_enabled(rq))
                hrtick_start_fair(rq, p);