[PATCH 3/3] sched: Kill sched_class::task_waking

[Peter Zijlstra]

With sched_class::task_waking being called only when we do
set_task_cpu(), we can make sched_class::migrate_task_rq() do the work
and eliminate sched_class::task_waking entirely.

Cc: Pavan Kondeti <pkondeti@xxxxxxxxxxxxxx>
Cc: Ben Segall <bsegall@xxxxxxxxxx>
Cc: Matt Fleming <matt@xxxxxxxxxxxxxxxxxxx>
Cc: Mike Galbraith <umgwanakikbuti@xxxxxxxxx>
Cc: Morten Rasmussen <morten.rasmussen@xxxxxxx>
Cc: Paul Turner <pjt@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: byungchul.park@xxxxxxx
Cc: Andrew Hunter <ahh@xxxxxxxxxx>
Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
---
 kernel/sched/core.c  |    9 -------
 kernel/sched/fair.c  |   58 ++++++++++++++++++++++++---------------------------
 kernel/sched/sched.h |    7 ++----
 3 files changed, 32 insertions(+), 42 deletions(-)

--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -1717,11 +1717,8 @@ ttwu_do_activate(struct rq *rq, struct t
 	if (p->sched_contributes_to_load)
 		rq->nr_uninterruptible--;
 
-	/*
-	 * If we migrated; we must have called sched_class::task_waking().
-	 */
 	if (wake_flags & WF_MIGRATED)
-		en_flags |= ENQUEUE_WAKING;
+		en_flags |= ENQUEUE_MIGRATED;
 #endif
 
 	ttwu_activate(rq, p, en_flags);
@@ -2049,10 +2046,6 @@ try_to_wake_up(struct task_struct *p, un
 	cpu = select_task_rq(p, p->wake_cpu, SD_BALANCE_WAKE, wake_flags);
 	if (task_cpu(p) != cpu) {
 		wake_flags |= WF_MIGRATED;
-
-		if (p->sched_class->task_waking)
-			p->sched_class->task_waking(p);
-
 		set_task_cpu(p, cpu);
 	}
 #endif /* CONFIG_SMP */
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -3282,7 +3282,7 @@ static inline void check_schedstat_requi
  *
  * WAKEUP (remote)
  *
- *	->task_waking_fair()
+ *	->migrate_task_rq_fair() (p->state == TASK_WAKING)
  *	  vruntime -= min_vruntime
  *
  *	enqueue
@@ -3297,7 +3297,7 @@ static inline void check_schedstat_requi
 static void
 enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
 {
-	bool renorm = !(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_WAKING);
+	bool renorm = !(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_MIGRATED);
 	bool curr = cfs_rq->curr == se;
 
 	/*
@@ -4865,33 +4865,6 @@ static unsigned long cpu_avg_load_per_ta
 	return 0;
 }
 
-/*
- * Called to migrate a waking task; as blocked tasks retain absolute vruntime
- * the migration needs to deal with this by subtracting the old and adding the
- * new min_vruntime -- the latter is done by enqueue_entity() when placing
- * the task on the new runqueue.
- */
-static void task_waking_fair(struct task_struct *p)
-{
-	struct sched_entity *se = &p->se;
-	struct cfs_rq *cfs_rq = cfs_rq_of(se);
-	u64 min_vruntime;
-
-#ifndef CONFIG_64BIT
-	u64 min_vruntime_copy;
-
-	do {
-		min_vruntime_copy = cfs_rq->min_vruntime_copy;
-		smp_rmb();
-		min_vruntime = cfs_rq->min_vruntime;
-	} while (min_vruntime != min_vruntime_copy);
-#else
-	min_vruntime = cfs_rq->min_vruntime;
-#endif
-
-	se->vruntime -= min_vruntime;
-}
-
 #ifdef CONFIG_FAIR_GROUP_SCHED
 /*
  * effective_load() calculates the load change as seen from the root_task_group
@@ -5648,6 +5621,32 @@ select_task_rq_fair(struct task_struct *
 static void migrate_task_rq_fair(struct task_struct *p)
 {
 	/*
+	 * As blocked tasks retain absolute vruntime the migration needs to
+	 * deal with this by subtracting the old and adding the new
+	 * min_vruntime -- the latter is done by enqueue_entity() when placing
+	 * the task on the new runqueue.
+	 */
+	if (p->state == TASK_WAKING) {
+		struct sched_entity *se = &p->se;
+		struct cfs_rq *cfs_rq = cfs_rq_of(se);
+		u64 min_vruntime;
+
+#ifndef CONFIG_64BIT
+		u64 min_vruntime_copy;
+
+		do {
+			min_vruntime_copy = cfs_rq->min_vruntime_copy;
+			smp_rmb();
+			min_vruntime = cfs_rq->min_vruntime;
+		} while (min_vruntime != min_vruntime_copy);
+#else
+		min_vruntime = cfs_rq->min_vruntime;
+#endif
+
+		se->vruntime -= min_vruntime;
+	}
+
+	/*
 	 * We are supposed to update the task to "current" time, then its up to date
 	 * and ready to go to new CPU/cfs_rq. But we have difficulty in getting
 	 * what current time is, so simply throw away the out-of-date time. This
@@ -8916,7 +8915,6 @@ const struct sched_class fair_sched_clas
 	.rq_online		= rq_online_fair,
 	.rq_offline		= rq_offline_fair,
 
-	.task_waking		= task_waking_fair,
 	.task_dead		= task_dead_fair,
 	.set_cpus_allowed	= set_cpus_allowed_common,
 #endif
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1164,7 +1164,7 @@ extern const u32 sched_prio_to_wmult[40]
  *
  * ENQUEUE_HEAD      - place at front of runqueue (tail if not specified)
  * ENQUEUE_REPLENISH - CBS (replenish runtime and postpone deadline)
- * ENQUEUE_WAKING    - sched_class::task_waking was called
+ * ENQUEUE_MIGRATED  - the task was migrated during wakeup
  *
  */
 
@@ -1179,9 +1179,9 @@ extern const u32 sched_prio_to_wmult[40]
 #define ENQUEUE_HEAD		0x08
 #define ENQUEUE_REPLENISH	0x10
 #ifdef CONFIG_SMP
-#define ENQUEUE_WAKING		0x20
+#define ENQUEUE_MIGRATED	0x20
 #else
-#define ENQUEUE_WAKING		0x00
+#define ENQUEUE_MIGRATED	0x00
 #endif
 
 #define RETRY_TASK		((void *)-1UL)
@@ -1213,7 +1213,6 @@ struct sched_class {
 	int  (*select_task_rq)(struct task_struct *p, int task_cpu, int sd_flag, int flags);
 	void (*migrate_task_rq)(struct task_struct *p);
 
-	void (*task_waking) (struct task_struct *task);
 	void (*task_woken) (struct rq *this_rq, struct task_struct *task);
 
 	void (*set_cpus_allowed)(struct task_struct *p,