Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 12 Jun 2023 14:05:28 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Tim Chen <tim.c.chen@linux.intel.com>
Cc:     Juri Lelli <juri.lelli@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>,
        Ricardo Neri <ricardo.neri@intel.com>,
        "Ravi V . Shankar" <ravi.v.shankar@intel.com>,
        Ben Segall <bsegall@google.com>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Len Brown <len.brown@intel.com>, Mel Gorman <mgorman@suse.de>,
        "Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
        Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Valentin Schneider <vschneid@redhat.com>,
        Ionela Voinescu <ionela.voinescu@arm.com>, x86@kernel.org,
        linux-kernel@vger.kernel.org,
        Shrikanth Hegde <sshegde@linux.vnet.ibm.com>,
        Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
        naveen.n.rao@linux.vnet.ibm.com,
        Yicong Yang <yangyicong@hisilicon.com>,
        Barry Song <v-songbaohua@oppo.com>,
        Chen Yu <yu.c.chen@intel.com>, Hillf Danton <hdanton@sina.com>
Subject: Re: [Patch v2 3/6] sched/fair: Implement prefer sibling imbalance
 calculation between asymmetric groups
Message-ID: <20230612120528.GL4253@hirez.programming.kicks-ass.net>
References: <cover.1686263351.git.tim.c.chen@linux.intel.com>
 <ef2400ea29f0e00aed1e2d75c1a352e63349059e.1686263351.git.tim.c.chen@linux.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ef2400ea29f0e00aed1e2d75c1a352e63349059e.1686263351.git.tim.c.chen@linux.intel.com>
Precedence: bulk

On Thu, Jun 08, 2023 at 03:32:29PM -0700, Tim Chen wrote:

> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 03573362274f..0b0904263d51 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -9372,6 +9372,65 @@ static inline bool smt_balance(struct lb_env *env, struct sg_lb_stats *sgs,
>  	return false;
>  }
>  
> +static inline long sibling_imbalance(struct lb_env *env,
> +				    struct sd_lb_stats *sds,
> +				    struct sg_lb_stats *busiest,
> +				    struct sg_lb_stats *local)
> +{
> +	int ncores_busiest, ncores_local;
> +	long imbalance;
> +
> +	if (env->idle == CPU_NOT_IDLE)
> +		return 0;
> +
> +	ncores_busiest = sds->busiest->cores;
> +	ncores_local = sds->local->cores;
> +
> +	if (ncores_busiest == ncores_local &&
> +	    (!(env->sd->flags & SD_ASYM_PACKING) ||
> +	      sched_asym_equal(env->dst_cpu,
> +			      sds->busiest->asym_prefer_cpu))) {
> +		imbalance = busiest->sum_nr_running;
> +		lsub_positive(&imbalance, local->sum_nr_running);
> +		return imbalance;
> +	}
> +
> +	/* Balance such that nr_running/ncores ratio are same on both groups */
> +	imbalance = ncores_local * busiest->sum_nr_running;
> +	lsub_positive(&imbalance, ncores_busiest * local->sum_nr_running);
> +	/* Normalize imbalance to become tasks to be moved to restore balance */
> +	imbalance /= ncores_local + ncores_busiest;
> +
> +	if (env->sd->flags & SD_ASYM_PACKING) {
> +		int limit;
> +
> +		if (!busiest->sum_nr_running)
> +			goto out;

This seems out-of-place, shouldn't we have terminate sooner if busiest
is empty?

> +
> +		if (sched_asym_prefer(env->dst_cpu, sds->busiest->asym_prefer_cpu)) {
> +			/* Don't leave preferred core idle */
> +			if (imbalance == 0 && local->sum_nr_running < ncores_local)
> +				imbalance = 1;
> +			goto out;
> +		}
> +
> +		/* Limit tasks moved from preferred group, don't leave cores idle */
> +		limit = busiest->sum_nr_running;
> +		lsub_positive(&limit, ncores_busiest);
> +		if (imbalance > limit)
> +			imbalance = limit;

How does this affect the server parts that have larger than single core
turbo domains?

> +
> +		goto out;
> +	}
> +
> +	/* Take advantage of resource in an empty sched group */
> +	if (imbalance == 0 && local->sum_nr_running == 0 &&
> +	    busiest->sum_nr_running > 1)
> +		imbalance = 1;
> +out:
> +	return imbalance << 1;
> +}


But basically you have:

        LcBn - BcLn
  imb = -----------
           LcBc

Which makes sense, except you then return:

  imb * 2

which then made me wonder about rounding.

Do we want to to add (LcBc -1) or (LcBc/2) to resp. ceil() or round()
the thing before division? Because currently it uses floor().

If you evaludate it like:


        2 * (LcBn - BcLn)
  imb = -----------------
              LcBc

The result is different from what you have now.

What actual behaviour is desired in these low imbalance cases? and can
you write a comment as to why we do as we do etc..?