While poking at this code recently I noted we do a pointless
unlock+lock cycle in sched_balance_newidle(). We drop the rq->lock (so
we can balance) but then instantly grab the same rq->lock again in
sched_balance_update_blocked_averages().

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 kernel/sched/fair.c |   27 ++++++++++++++++++---------
 1 file changed, 18 insertions(+), 9 deletions(-)

--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -9902,15 +9902,11 @@ static unsigned long task_h_load(struct
 }
 #endif /* !CONFIG_FAIR_GROUP_SCHED */
 
-static void sched_balance_update_blocked_averages(int cpu)
+static void __sched_balance_update_blocked_averages(struct rq *rq)
 {
 	bool decayed = false, done = true;
-	struct rq *rq = cpu_rq(cpu);
-	struct rq_flags rf;
 
-	rq_lock_irqsave(rq, &rf);
 	update_blocked_load_tick(rq);
-	update_rq_clock(rq);
 
 	decayed |= __update_blocked_others(rq, &done);
 	decayed |= __update_blocked_fair(rq, &done);
@@ -9918,7 +9914,15 @@ static void sched_balance_update_blocked
 	update_blocked_load_status(rq, !done);
 	if (decayed)
 		cpufreq_update_util(rq, 0);
-	rq_unlock_irqrestore(rq, &rf);
+}
+
+static void sched_balance_update_blocked_averages(int cpu)
+{
+	struct rq *rq = cpu_rq(cpu);
+
+	guard(rq_lock_irqsave)(rq);
+	update_rq_clock(rq);
+	__sched_balance_update_blocked_averages(rq);
 }
 
 /********** Helpers for sched_balance_find_src_group ************************/
@@ -12865,12 +12869,17 @@ static int sched_balance_newidle(struct
 	}
 	rcu_read_unlock();
 
+	/*
+	 * Include sched_balance_update_blocked_averages() in the cost
+	 * calculation because it can be quite costly -- this ensures we skip
+	 * it when avg_idle gets to be very low.
+	 */
+	t0 = sched_clock_cpu(this_cpu);
+	__sched_balance_update_blocked_averages(this_rq);
+
 	rq_modified_clear(this_rq);
 	raw_spin_rq_unlock(this_rq);
 
-	t0 = sched_clock_cpu(this_cpu);
-	sched_balance_update_blocked_averages(this_cpu);
-
 	rcu_read_lock();
 	for_each_domain(this_cpu, sd) {
 		u64 domain_cost;

On 11/27/25 9:09 PM, Peter Zijlstra wrote:
> While poking at this code recently I noted we do a pointless
> unlock+lock cycle in sched_balance_newidle(). We drop the rq->lock (so
> we can balance) but then instantly grab the same rq->lock again in
> sched_balance_update_blocked_averages().
> 
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> ---
>   kernel/sched/fair.c |   27 ++++++++++++++++++---------
>   1 file changed, 18 insertions(+), 9 deletions(-)
> 
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -9902,15 +9902,11 @@ static unsigned long task_h_load(struct
>   }
>   #endif /* !CONFIG_FAIR_GROUP_SCHED */
>   
> -static void sched_balance_update_blocked_averages(int cpu)
> +static void __sched_balance_update_blocked_averages(struct rq *rq)
>   {
>   	bool decayed = false, done = true;
> -	struct rq *rq = cpu_rq(cpu);
> -	struct rq_flags rf;
>   
> -	rq_lock_irqsave(rq, &rf);
>   	update_blocked_load_tick(rq);
> -	update_rq_clock(rq);
>   
>   	decayed |= __update_blocked_others(rq, &done);
>   	decayed |= __update_blocked_fair(rq, &done);
> @@ -9918,7 +9914,15 @@ static void sched_balance_update_blocked
>   	update_blocked_load_status(rq, !done);
>   	if (decayed)
>   		cpufreq_update_util(rq, 0);
> -	rq_unlock_irqrestore(rq, &rf);
> +}
> +
> +static void sched_balance_update_blocked_averages(int cpu)
> +{
> +	struct rq *rq = cpu_rq(cpu);
> +
> +	guard(rq_lock_irqsave)(rq);
> +	update_rq_clock(rq);
> +	__sched_balance_update_blocked_averages(rq);
>   }
>   
>   /********** Helpers for sched_balance_find_src_group ************************/
> @@ -12865,12 +12869,17 @@ static int sched_balance_newidle(struct
>   	}
>   	rcu_read_unlock();
>   
> +	/*
> +	 * Include sched_balance_update_blocked_averages() in the cost
> +	 * calculation because it can be quite costly -- this ensures we skip
> +	 * it when avg_idle gets to be very low.
> +	 */
> +	t0 = sched_clock_cpu(this_cpu);
> +	__sched_balance_update_blocked_averages(this_rq);
> +

I think we do update_rq_clock earlier as early as __schedule.
no warnings seen.

Reviewed-by: Shrikanth Hegde <sshegde@linux.ibm.com>