sched/fair: Provide runnable_load_avg back to cfs_rq

The cfs_rq's load_avg is composed of runnable_load_avg and blocked_load_avg.
Before this series, sometimes the runnable_load_avg is used, and sometimes
the load_avg is used. Completely replacing all uses of runnable_load_avg
with load_avg may be too big a leap, i.e., the blocked_load_avg is concerned
to result in overrated load. Therefore, we get runnable_load_avg back.

The new cfs_rq's runnable_load_avg is improved to be updated with all of the
runnable sched_eneities at the same time, so the one sched_entity updated and
the others stale problem is solved.

Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: arjan@linux.intel.com
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: fengguang.wu@intel.com
Cc: len.brown@intel.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: rafael.j.wysocki@intel.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1436918682-4971-7-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

kernel/sched/debug.c

@@ -207,6 +207,8 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)
 #ifdef CONFIG_SMP
 	SEQ_printf(m, "  .%-30s: %lu\n", "load_avg",
 			cfs_rq->avg.load_avg);
+	SEQ_printf(m, "  .%-30s: %lu\n", "runnable_load_avg",
+			cfs_rq->runnable_load_avg);
 	SEQ_printf(m, "  .%-30s: %lu\n", "util_avg",
 			cfs_rq->avg.util_avg);
 	SEQ_printf(m, "  .%-30s: %ld\n", "removed_load_avg",

kernel/sched/fair.c

@@ -2542,7 +2542,7 @@ static u32 __compute_runnable_contrib(u64 n)
  */
 static __always_inline int
 __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
-		  unsigned long weight, int running)
+		  unsigned long weight, int running, struct cfs_rq *cfs_rq)
 {
 	u64 delta, periods;
 	u32 contrib;
@@ -2582,8 +2582,11 @@ __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
 		 * period and accrue it.
 		 */
 		delta_w = 1024 - delta_w;
-		if (weight)
+		if (weight) {
 			sa->load_sum += weight * delta_w;
+			if (cfs_rq)
+				cfs_rq->runnable_load_sum += weight * delta_w;
+		}
 		if (running)
 			sa->util_sum += delta_w * scale_freq >> SCHED_CAPACITY_SHIFT;
 
@@ -2594,19 +2597,29 @@ __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
 		delta %= 1024;
 
 		sa->load_sum = decay_load(sa->load_sum, periods + 1);
+		if (cfs_rq) {
+			cfs_rq->runnable_load_sum =
+				decay_load(cfs_rq->runnable_load_sum, periods + 1);
+		}
 		sa->util_sum = decay_load((u64)(sa->util_sum), periods + 1);
 
 		/* Efficiently calculate \sum (1..n_period) 1024*y^i */
 		contrib = __compute_runnable_contrib(periods);
-		if (weight)
+		if (weight) {
 			sa->load_sum += weight * contrib;
+			if (cfs_rq)
+				cfs_rq->runnable_load_sum += weight * contrib;
+		}
 		if (running)
 			sa->util_sum += contrib * scale_freq >> SCHED_CAPACITY_SHIFT;
 	}
 
 	/* Remainder of delta accrued against u_0` */
-	if (weight)
+	if (weight) {
 		sa->load_sum += weight * delta;
+		if (cfs_rq)
+			cfs_rq->runnable_load_sum += weight * delta;
+	}
 	if (running)
 		sa->util_sum += delta * scale_freq >> SCHED_CAPACITY_SHIFT;
 
@@ -2614,6 +2627,10 @@ __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
 
 	if (decayed) {
 		sa->load_avg = div_u64(sa->load_sum, LOAD_AVG_MAX);
+		if (cfs_rq) {
+			cfs_rq->runnable_load_avg =
+				div_u64(cfs_rq->runnable_load_sum, LOAD_AVG_MAX);
+		}
 		sa->util_avg = (sa->util_sum << SCHED_LOAD_SHIFT) / LOAD_AVG_MAX;
 	}
 
@@ -2661,7 +2678,7 @@ static inline int update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq)
 	}
 
 	decayed = __update_load_avg(now, cpu_of(rq_of(cfs_rq)), sa,
-		scale_load_down(cfs_rq->load.weight), cfs_rq->curr != NULL);
+		scale_load_down(cfs_rq->load.weight), cfs_rq->curr != NULL, cfs_rq);
 
 #ifndef CONFIG_64BIT
 	smp_wmb();
@@ -2683,7 +2700,7 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg)
 	 * track group sched_entity load average for task_h_load calc in migration
 	 */
 	__update_load_avg(now, cpu, &se->avg,
-		se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
+		se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se, NULL);
 
 	if (update_cfs_rq_load_avg(now, cfs_rq) && update_tg)
 		update_tg_load_avg(cfs_rq, 0);
@@ -2703,11 +2720,15 @@ enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se)
 	}
 	else {
 		__update_load_avg(now, cpu_of(rq_of(cfs_rq)), sa,
-			se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
+			se->on_rq * scale_load_down(se->load.weight),
+			cfs_rq->curr == se, NULL);
 	}
 
 	decayed = update_cfs_rq_load_avg(now, cfs_rq);
 
+	cfs_rq->runnable_load_avg += sa->load_avg;
+	cfs_rq->runnable_load_sum += sa->load_sum;
+
 	if (migrated) {
 		cfs_rq->avg.load_avg += sa->load_avg;
 		cfs_rq->avg.load_sum += sa->load_sum;
@@ -2719,6 +2740,18 @@ enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se)
 		update_tg_load_avg(cfs_rq, 0);
 }
 
+/* Remove the runnable load generated by se from cfs_rq's runnable load average */
+static inline void
+dequeue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se)
+{
+	update_load_avg(se, 1);
+
+	cfs_rq->runnable_load_avg =
+		max_t(long, cfs_rq->runnable_load_avg - se->avg.load_avg, 0);
+	cfs_rq->runnable_load_sum =
+		max_t(s64, cfs_rq->runnable_load_sum - se->avg.load_sum, 0);
+}
+
 /*
  * Task first catches up with cfs_rq, and then subtract
  * itself from the cfs_rq (task must be off the queue now).
@@ -2740,7 +2773,7 @@ void remove_entity_load_avg(struct sched_entity *se)
 	last_update_time = cfs_rq->avg.last_update_time;
 #endif
 
-	__update_load_avg(last_update_time, cpu_of(rq_of(cfs_rq)), &se->avg, 0, 0);
+	__update_load_avg(last_update_time, cpu_of(rq_of(cfs_rq)), &se->avg, 0, 0, NULL);
 	atomic_long_add(se->avg.load_avg, &cfs_rq->removed_load_avg);
 	atomic_long_add(se->avg.util_avg, &cfs_rq->removed_util_avg);
 }
@@ -2770,6 +2803,8 @@ static int idle_balance(struct rq *this_rq);
 static inline void update_load_avg(struct sched_entity *se, int update_tg) {}
 static inline void
 enqueue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) {}
+static inline void
+dequeue_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) {}
 static inline void remove_entity_load_avg(struct sched_entity *se) {}
 
 static inline int idle_balance(struct rq *rq)
@@ -2977,7 +3012,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
 	 * Update run-time statistics of the 'current'.
 	 */
 	update_curr(cfs_rq);
-	update_load_avg(se, 1);
+	dequeue_entity_load_avg(cfs_rq, se);
 
 	update_stats_dequeue(cfs_rq, se);
 	if (flags & DEQUEUE_SLEEP) {
@@ -7863,7 +7898,7 @@ static void switched_from_fair(struct rq *rq, struct task_struct *p)
 #ifdef CONFIG_SMP
 	/* Catch up with the cfs_rq and remove our load when we leave */
 	__update_load_avg(cfs_rq->avg.last_update_time, cpu_of(rq), &se->avg,
-		se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se);
+		se->on_rq * scale_load_down(se->load.weight), cfs_rq->curr == se, NULL);
 
 	cfs_rq->avg.load_avg =
 		max_t(long, cfs_rq->avg.load_avg - se->avg.load_avg, 0);

kernel/sched/sched.h

@@ -368,6 +368,8 @@ struct cfs_rq {
 	 * CFS load tracking
 	 */
 	struct sched_avg avg;
+	u64 runnable_load_sum;
+	unsigned long runnable_load_avg;
 #ifdef CONFIG_FAIR_GROUP_SCHED
 	unsigned long tg_load_avg_contrib;
 #endif