Re: [PATCH 2/2] sched/fair: Cleanup in migrate_degrades_locality() to improve readability

[Peter Zijlstra]

On Wed, Jun 14, 2023 at 10:22:24AM +0000, Swapnil Sapkal wrote:
> The migrate_degrades_locality() returns tristate value whether
> the migration will improve locality, degrades locality or no
> impact. Handle this return values with enum to improve the
> readability.

I can see how you ended up there, that tristate is weird, but perhaps
don't make it more complicated than it should be?

---
 kernel/sched/fair.c | 39 ++++++++++++++++++++-------------------
 1 file changed, 20 insertions(+), 19 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 373ff5f55884..a8449f594348 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -8446,42 +8446,42 @@ static int task_hot(struct task_struct *p, struct lb_env *env)
 #ifdef CONFIG_NUMA_BALANCING
 /*
  * Returns 1, if task migration degrades locality
- * Returns 0, if task migration improves locality i.e migration preferred.
- * Returns -1, if task migration is not affected by locality.
+ * Returns 0, if task migration is not affected by locality.
+ * Returns -1, if task migration improves locality i.e migration preferred.
  */
-static int migrate_degrades_locality(struct task_struct *p, struct lb_env *env)
+static long migrate_degrades_locality(struct task_struct *p, struct lb_env *env)
 {
 	struct numa_group *numa_group = rcu_dereference(p->numa_group);
 	unsigned long src_weight, dst_weight;
 	int src_nid, dst_nid, dist;
 
 	if (!static_branch_likely(&sched_numa_balancing))
-		return -1;
+		return 0;
 
 	if (!p->numa_faults || !(env->sd->flags & SD_NUMA))
-		return -1;
+		return 0;
 
 	src_nid = cpu_to_node(env->src_cpu);
 	dst_nid = cpu_to_node(env->dst_cpu);
 
 	if (src_nid == dst_nid)
-		return -1;
+		return 0;
 
 	/* Migrating away from the preferred node is always bad. */
 	if (src_nid == p->numa_preferred_nid) {
 		if (env->src_rq->nr_running > env->src_rq->nr_preferred_running)
 			return 1;
 		else
-			return -1;
+			return 0;
 	}
 
 	/* Encourage migration to the preferred node. */
 	if (dst_nid == p->numa_preferred_nid)
-		return 0;
+		return -1;
 
 	/* Leaving a core idle is often worse than degrading locality. */
 	if (env->idle == CPU_IDLE)
-		return -1;
+		return 0;
 
 	dist = node_distance(src_nid, dst_nid);
 	if (numa_group) {
@@ -8492,14 +8492,14 @@ static int migrate_degrades_locality(struct task_struct *p, struct lb_env *env)
 		dst_weight = task_weight(p, dst_nid, dist);
 	}
 
-	return dst_weight < src_weight;
+	return src_weight - dst_weight;
 }
 
 #else
-static inline int migrate_degrades_locality(struct task_struct *p,
+static inline long migrate_degrades_locality(struct task_struct *p,
 					     struct lb_env *env)
 {
-	return -1;
+	return 0;
 }
 #endif
 
@@ -8509,7 +8509,7 @@ static inline int migrate_degrades_locality(struct task_struct *p,
 static
 int can_migrate_task(struct task_struct *p, struct lb_env *env)
 {
-	int tsk_cache_hot;
+	long degrades, hot;
 
 	lockdep_assert_rq_held(env->src_rq);
 
@@ -8578,13 +8578,14 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env)
 	if (env->flags & LBF_ACTIVE_LB)
 		return 1;
 
-	tsk_cache_hot = migrate_degrades_locality(p, env);
-	if (tsk_cache_hot == -1)
-		tsk_cache_hot = task_hot(p, env);
+	degrades = migrate_degrades_locality(p, env);
+	if (!degrades)
+		hot = task_hot(p, env);
+	else
+		hot = degrades > 0;
 
-	if (tsk_cache_hot <= 0 ||
-	    env->sd->nr_balance_failed > env->sd->cache_nice_tries) {
-		if (tsk_cache_hot == 1) {
+	if (env->sd->nr_balance_failed > env->sd->cache_nice_tries) {
+		if (hot) {
 			schedstat_inc(env->sd->lb_hot_gained[env->idle]);
 			schedstat_inc(p->stats.nr_forced_migrations);
 		}