sched/fair: Fall back to sched-idle CPU if idle CPU isn't found

mainline inclusion
from mainline-v5.4-rc1
commit 3c29e651
category: feature
bugzilla: 38260, https://bugzilla.openeuler.org/show_bug.cgi?id=22
CVE: NA
---------------------------

We try to find an idle CPU to run the next task, but in case we don't
find an idle CPU it is better to pick a CPU which will run the task the
soonest, for performance reason.

A CPU which isn't idle but has only SCHED_IDLE activity queued on it
should be a good target based on this criteria as any normal fair task
will most likely preempt the currently running SCHED_IDLE task
immediately. In fact, choosing a SCHED_IDLE CPU over a fully idle one
shall give better results as it should be able to run the task sooner
than an idle CPU (which requires to be woken up from an idle state).

This patch updates both fast and slow paths with this optimization.
Signed-off-by: NViresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: NPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: chris.redpath@arm.com
Cc: quentin.perret@linaro.org
Cc: songliubraving@fb.com
Cc: steven.sistare@oracle.com
Cc: subhra.mazumdar@oracle.com
Cc: tkjos@google.com
Link: https://lkml.kernel.org/r/eeafa25fdeb6f6edd5b2da716bc8f0ba7708cbcf.1561523542.git.viresh.kumar@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>
Signed-off-by: NCheng Jian <cj.chengjian@huawei.com>
Reviewed-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>

kernel/sched/fair.c

@@ -5379,6 +5379,15 @@ static struct {
 
 #endif /* CONFIG_NO_HZ_COMMON */
 
+/* CPU only has SCHED_IDLE tasks enqueued */
+static int sched_idle_cpu(int cpu)
+{
+	struct rq *rq = cpu_rq(cpu);
+
+	return unlikely(rq->nr_running == rq->cfs.idle_h_nr_running &&
+			rq->nr_running);
+}
+
 /**
  * __cpu_load_update - update the rq->cpu_load[] statistics
  * @this_rq: The rq to update statistics for
@@ -5938,7 +5947,7 @@ find_idlest_group_cpu(struct sched_group *group, struct task_struct *p, int this
 	unsigned int min_exit_latency = UINT_MAX;
 	u64 latest_idle_timestamp = 0;
 	int least_loaded_cpu = this_cpu;
-	int shallowest_idle_cpu = -1;
+	int shallowest_idle_cpu = -1, si_cpu = -1;
 	int i;
 
 	/* Check if we have any choice: */
@@ -5969,7 +5978,12 @@ find_idlest_group_cpu(struct sched_group *group, struct task_struct *p, int this
 				latest_idle_timestamp = rq->idle_stamp;
 				shallowest_idle_cpu = i;
 			}
-		} else if (shallowest_idle_cpu == -1) {
+		} else if (shallowest_idle_cpu == -1 && si_cpu == -1) {
+			if (sched_idle_cpu(i)) {
+				si_cpu = i;
+				continue;
+			}
+
 			load = weighted_cpuload(cpu_rq(i));
 			if (load < min_load) {
 				min_load = load;
@@ -5978,7 +5992,11 @@ find_idlest_group_cpu(struct sched_group *group, struct task_struct *p, int this
 		}
 	}
 
-	return shallowest_idle_cpu != -1 ? shallowest_idle_cpu : least_loaded_cpu;
+	if (shallowest_idle_cpu != -1)
+		return shallowest_idle_cpu;
+	if (si_cpu != -1)
+		return si_cpu;
+	return least_loaded_cpu;
 }
 
 static inline int find_idlest_cpu(struct sched_domain *sd, struct task_struct *p,
@@ -6131,7 +6149,7 @@ static int select_idle_core(struct task_struct *p, struct sched_domain *sd, int
  */
 static int select_idle_smt(struct task_struct *p, struct sched_domain *sd, int target)
 {
-	int cpu;
+	int cpu, si_cpu = -1;
 
 	if (!static_branch_likely(&sched_smt_present))
 		return -1;
@@ -6141,9 +6159,11 @@ static int select_idle_smt(struct task_struct *p, struct sched_domain *sd, int t
 			continue;
 		if (available_idle_cpu(cpu))
 			return cpu;
+		if (si_cpu == -1 && sched_idle_cpu(cpu))
+			si_cpu = cpu;
 	}
 
-	return -1;
+	return si_cpu;
 }
 
 #else /* CONFIG_SCHED_SMT */
@@ -6171,11 +6191,11 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
 	u64 avg_cost, avg_idle;
 	u64 time, cost;
 	s64 delta;
-	int cpu, nr = INT_MAX;
+	int cpu, nr = INT_MAX, si_cpu = -1;
 
 	this_sd = rcu_dereference(*this_cpu_ptr(&sd_llc));
 	if (!this_sd)
-		return -1;
+		return si_cpu;
 
 	/*
 	 * Due to large variance we need a large fuzz factor; hackbench in
@@ -6185,7 +6205,7 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
 	avg_cost = this_sd->avg_scan_cost + 1;
 
 	if (sched_feat(SIS_AVG_CPU) && avg_idle < avg_cost)
-		return -1;
+		return si_cpu;
 
 	if (sched_feat(SIS_PROP)) {
 		u64 span_avg = sd->span_weight * avg_idle;
@@ -6199,11 +6219,13 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t
 
 	for_each_cpu_wrap(cpu, sched_domain_span(sd), target) {
 		if (!--nr)
-			return -1;
+			return si_cpu;
 		if (!cpumask_test_cpu(cpu, &p->cpus_allowed))
 			continue;
 		if (available_idle_cpu(cpu))
 			break;
+		if (si_cpu == -1 && sched_idle_cpu(cpu))
+			si_cpu = cpu;
 	}
 
 	time = local_clock() - time;
@@ -6222,13 +6244,14 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target)
 	struct sched_domain *sd;
 	int i, recent_used_cpu;
 
-	if (available_idle_cpu(target))
+	if (available_idle_cpu(target) || sched_idle_cpu(target))
 		return target;
 
 	/*
 	 * If the previous CPU is cache affine and idle, don't be stupid:
 	 */
-	if (prev != target && cpus_share_cache(prev, target) && available_idle_cpu(prev))
+	if (prev != target && cpus_share_cache(prev, target) &&
+	    (available_idle_cpu(prev) || sched_idle_cpu(prev)))
 		return prev;
 
 	/* Check a recently used CPU as a potential idle candidate: */
@@ -6236,7 +6259,7 @@ static int select_idle_sibling(struct task_struct *p, int prev, int target)
 	if (recent_used_cpu != prev &&
 	    recent_used_cpu != target &&
 	    cpus_share_cache(recent_used_cpu, target) &&
-	    available_idle_cpu(recent_used_cpu) &&
+	    (available_idle_cpu(recent_used_cpu) || sched_idle_cpu(recent_used_cpu)) &&
 	    cpumask_test_cpu(p->recent_used_cpu, &p->cpus_allowed)) {
 		/*
 		 * Replace recent_used_cpu with prev as it is a potential