linux-kernel - Re: [PATCH v4 2/3] sched: Consider max cost of idle balance per sched domain

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 2 Sep 2013 12:24:46 +0530
From:	Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
To:	Jason Low <jason.low2@...com>
Cc:	mingo@...hat.com, peterz@...radead.org,
	linux-kernel@...r.kernel.org, efault@....de, pjt@...gle.com,
	preeti@...ux.vnet.ibm.com, akpm@...ux-foundation.org,
	mgorman@...e.de, riel@...hat.com, aswin@...com, scott.norton@...com
Subject: Re: [PATCH v4 2/3] sched: Consider max cost of idle balance per
 sched domain

* Jason Low <jason.low2@...com> [2013-08-29 13:05:35]:

> +	u64 curr_cost = 0;
> 
>  	this_rq->idle_stamp = rq_clock(this_rq);
> 
> -	if (this_rq->avg_idle < sysctl_sched_migration_cost)
> +	if (this_rq->avg_idle < this_rq->max_idle_balance_cost)
>  		return;
> 

Since max_idle_balance_cost includes the cost of balances across all
domains. Can the cost of balance at a higher domain being higher result
in not doing load balance at a lower level?

Shouldnt the check below for sd->max_newidle_lb_cost mean that we can
actually do away with this check.

>  	/*
> @@ -5299,14 +5300,29 @@ void idle_balance(int this_cpu, struct rq *this_rq)
>  	for_each_domain(this_cpu, sd) {
>  		unsigned long interval;
>  		int balance = 1;
> +		u64 t0, domain_cost, max = 5*sysctl_sched_migration_cost;
> 
>  		if (!(sd->flags & SD_LOAD_BALANCE))
>  			continue;
> 
> +		if (this_rq->avg_idle < curr_cost + sd->max_newidle_lb_cost)
> +			break;

I am referring to this check in my above comment.

> +
>  		if (sd->flags & SD_BALANCE_NEWIDLE) {
> +			t0 = sched_clock_cpu(smp_processor_id());
> +
>  			/* If we've pulled tasks over stop searching: */
>  			pulled_task = load_balance(this_cpu, this_rq,
>  						   sd, CPU_NEWLY_IDLE, &balance);
> +
> +			domain_cost = sched_clock_cpu(smp_processor_id()) - t0;
> +			if (domain_cost > max)
> +				domain_cost = max;
> +
> +			if (domain_cost > sd->max_newidle_lb_cost)
> +				sd->max_newidle_lb_cost = domain_cost;

If we face a runq lock contention, then domain_cost can go up.
The runq lock contention could be temporary, but we carry the domain
cost forever (i.e till the next reboot).  How about averaging the cost +
penalty for unsuccessful balance.

Something like 
			domain_cost = sched_clock_cpu(smp_processor_id()) - t0;
			if (!pulled_task)
				domain_cost *= 2;
		
			sd->max_newidle_lb_cost += domain_cost;
			sd->max_newidle_lb_cost /= 2;
				
				
Maybe the name could then change to avg_newidle_lb_cost.

> +
> +			curr_cost += domain_cost;
>  		}
> 
-- 
Thanks and Regards
Srikar Dronamraju

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/