On 01/28/21 18:31, Valentin Schneider wrote:
> check_misfit_status() checks for both capacity pressure & available CPUs
> with higher capacity. Now that we have a sane(ish) capacity comparison
> margin which is used throughout load-balance, this can be condensed into a
> single check:
>
> capacity_greater(<root_domain max capacity>, <misfit task CPU's capacity>);
>
> This has the added benefit of returning false if the misfit task CPU's is
> heavily pressured, but there are no better candidates for migration.
>
> Signed-off-by: Valentin Schneider <valentin.schnei...@arm.com>
> ---
check_cpu_capacity() call looks redundant now, yes.
Reviewed-by: Qais Yousef <qais.you...@arm.com>
Thanks
--
Qais Yousef
> kernel/sched/fair.c | 10 ++++------
> 1 file changed, 4 insertions(+), 6 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 0959a770ecc0..ef44474b8fbf 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -8254,14 +8254,12 @@ check_cpu_capacity(struct rq *rq, struct sched_domain
> *sd)
>
> /*
> * Check whether a rq has a misfit task and if it looks like we can actually
> - * help that task: we can migrate the task to a CPU of higher capacity, or
> - * the task's current CPU is heavily pressured.
> + * help that task: we can migrate the task to a CPU of higher capacity.
> */
> -static inline int check_misfit_status(struct rq *rq, struct sched_domain *sd)
> +static inline int check_misfit_status(struct rq *rq)
> {
> return rq->misfit_task_load &&
> - (capacity_greater(rq->rd->max_cpu_capacity,
> rq->cpu_capacity_orig) ||
> - check_cpu_capacity(rq, sd));
> + capacity_greater(rq->rd->max_cpu_capacity, rq->cpu_capacity);
> }
>
> /*
> @@ -10238,7 +10236,7 @@ static void nohz_balancer_kick(struct rq *rq)
> * When ASYM_CPUCAPACITY; see if there's a higher capacity CPU
> * to run the misfit task on.
> */
> - if (check_misfit_status(rq, sd)) {
> + if (check_misfit_status(rq)) {
> flags = NOHZ_KICK_MASK;
> goto unlock;
> }
> --
> 2.27.0
>
