[PATCH 2/3] sched/fair: Move task_numa_work() init to init_numa_balancing()

[Valentin Schneider]

We only need to set the callback_head worker function once, do it
during sched_fork().

While at it, move the comment regarding double task_work addition to
init_numa_balancing(), since the double add sentinel is first set there.

Suggested-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Signed-off-by: Valentin Schneider <valentin.schneider@xxxxxxx>
---
 kernel/sched/fair.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 476b0201a8fb..74faa55bc52a 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -2441,7 +2441,7 @@ void task_numa_work(struct callback_head *work)
 
 	SCHED_WARN_ON(p != container_of(work, struct task_struct, numa_work));
 
-	work->next = work; /* protect against double add */
+	work->next = work;
 	/*
 	 * Who cares about NUMA placement when they're dying.
 	 *
@@ -2585,12 +2585,15 @@ void init_numa_balancing(unsigned long clone_flags, struct task_struct *p)
 	p->node_stamp			= 0;
 	p->numa_scan_seq		= mm ? mm->numa_scan_seq : 0;
 	p->numa_scan_period		= sysctl_numa_balancing_scan_delay;
+	/* Protect against double add, see task_tick_numa and task_numa_work */
 	p->numa_work.next		= &p->numa_work;
 	p->numa_faults			= NULL;
 	p->numa_group			= NULL;
 	p->last_task_numa_placement	= 0;
 	p->last_sum_exec_runtime	= 0;
 
+	init_task_work(&p->numa_work, task_numa_work);
+
 	/* New address space, reset the preferred nid */
 	if (!(clone_flags & CLONE_VM)) {
 		p->numa_preferred_nid = NUMA_NO_NODE;
@@ -2639,10 +2642,8 @@ static void task_tick_numa(struct rq *rq, struct task_struct *curr)
 			curr->numa_scan_period = task_scan_start(curr);
 		curr->node_stamp += period;
 
-		if (!time_before(jiffies, curr->mm->numa_next_scan)) {
-			init_task_work(work, task_numa_work); /* TODO: move this into sched_fork() */
+		if (!time_before(jiffies, curr->mm->numa_next_scan))
 			task_work_add(curr, work, true);
-		}
 	}
 }
 
--
2.22.0