Re: [Patch v2 1/6] sched/fair: Determine active load balance for SMT sched groups

[Peter Zijlstra]

On Thu, Jun 08, 2023 at 03:32:27PM -0700, Tim Chen wrote:

> +/* One group has more than one SMT CPU while the other group does not */
> +static inline bool smt_vs_nonsmt_groups(struct sched_group *sg1,
> +				    struct sched_group *sg2)
> +{
> +	if (!sg1 || !sg2)
> +		return false;
> +
> +	return (sg1->flags & SD_SHARE_CPUCAPACITY) !=
> +		(sg2->flags & SD_SHARE_CPUCAPACITY);
> +}
> +
> +static inline bool smt_balance(struct lb_env *env, struct sg_lb_stats *sgs,
> +			       struct sched_group *group)
> +{
> +	if (env->idle == CPU_NOT_IDLE)
> +		return false;
> +
> +	/*
> +	 * For SMT source group, it is better to move a task
> +	 * to a CPU that doesn't have multiple tasks sharing its CPU capacity.
> +	 * Note that if a group has a single SMT, SD_SHARE_CPUCAPCITY
> +	 * will not be on.
> +	 */
> +	if (group->flags & SD_SHARE_CPUCAPACITY &&
> +	    sgs->sum_h_nr_running > 1)
> +		return true;

AFAICT this does the right thing for SMT>2

> +
> +	return false;
> +}
> +
>  static inline bool
>  sched_reduced_capacity(struct rq *rq, struct sched_domain *sd)
>  {

> @@ -9537,6 +9581,18 @@ static bool update_sd_pick_busiest(struct lb_env *env,
>  		break;
>  
>  	case group_has_spare:
> +		/*
> +		 * Do not pick sg with SMT CPUs over sg with pure CPUs,
> +		 * as we do not want to pull task off half empty SMT core
> +		 * and make the core idle.
> +		 */
> +		if (smt_vs_nonsmt_groups(sds->busiest, sg)) {
> +			if (sg->flags & SD_SHARE_CPUCAPACITY)
> +				return false;
> +			else
> +				return true;
> +		}

However, here I'm not at all sure. Consider SMT-4 with 2 active CPUs, we
still very much would like to pull one task off if we have an idle core
somewhere, no?