Re: [PATCH] sched, fair: Allow a small degree of load imbalance between SD_NUMA domains

[Rik van Riel]

On Wed, 2019-12-18 at 15:44 +0000, Mel Gorman wrote:

> +			/*
> +			 * Ignore imbalance unless busiest sd is close
> to 50%
> +			 * utilisation. At that point balancing for
> memory
> +			 * bandwidth and potentially avoiding
> unnecessary use
> +			 * of HT siblings is as relevant as memory
> locality.
> +			 */
> +			imbalance_max = (busiest->group_weight >> 1) -
> imbalance_adj;
> +			if (env->imbalance <= imbalance_adj &&
> +			    busiest->sum_nr_running < imbalance_max) {
> +				env->imbalance = 0;
> +			}
> +		}
>  		return;
>  	}

I can see how the 50% point is often great for HT,
but I wonder if that is also the case for SMT4 and
SMT8 systems...

-- 
All Rights Reversed.

Attachment: signature.asc
Description: This is a digitally signed message part