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;
Message-ID: <b85e608f-7293-9588-5e85-36b62f1fdbeb@arm.com>
Date:   Mon, 12 Sep 2022 10:44:15 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.9.1
Subject: Re: [PATCH 1/4] sched/fair: make sure to try to detach at least one
 movable task
Content-Language: en-US
To:     Vincent Guittot <vincent.guittot@linaro.org>, mingo@redhat.com,
        peterz@infradead.org, juri.lelli@redhat.com, rostedt@goodmis.org,
        bsegall@google.com, mgorman@suse.de, bristot@redhat.com,
        vschneid@redhat.com, linux-kernel@vger.kernel.org
Cc:     zhangqiao22@huawei.com
References: <20220825122726.20819-1-vincent.guittot@linaro.org>
 <20220825122726.20819-2-vincent.guittot@linaro.org>
From:   Dietmar Eggemann <dietmar.eggemann@arm.com>
In-Reply-To: <20220825122726.20819-2-vincent.guittot@linaro.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 25/08/2022 14:27, Vincent Guittot wrote:

s/sched/fair: make/sched/fair: Make

> During load balance, we try at most env->loop_max time to move a task.
> But it can happen that the loop_max LRU tasks (ie tail of
> the cfs_tasks list) can't be moved to dst_cpu because of affinity.
> In this case, loop in the list until we found at least one.
> 
> The maximum of detached tasks remained the same as before.

Not sure how this relates to the patch? Isn't this given by the
`env->imbalance <= 0` check at the end of detach_tasks()?

> 
> Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
> ---
>  kernel/sched/fair.c | 12 +++++++++---
>  1 file changed, 9 insertions(+), 3 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index da388657d5ac..02b7b808e186 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -8052,8 +8052,12 @@ static int detach_tasks(struct lb_env *env)
>  		p = list_last_entry(tasks, struct task_struct, se.group_node);
>  
>  		env->loop++;
> -		/* We've more or less seen every task there is, call it quits */
> -		if (env->loop > env->loop_max)
> +		/*
> +		 * We've more or less seen every task there is, call it quits

I never understood this `more or less`. Either we have seen all tasks or
not?

> +		 * unless we haven't found any movable task yet.
> +		 */
> +		if (env->loop > env->loop_max &&
> +		    !(env->flags & LBF_ALL_PINNED))
>  			break;
>  
>  		/* take a breather every nr_migrate tasks */
> @@ -10182,7 +10186,9 @@ static int load_balance(int this_cpu, struct rq *this_rq,
>  
>  		if (env.flags & LBF_NEED_BREAK) {
>  			env.flags &= ~LBF_NEED_BREAK;
> -			goto more_balance;
> +			/* Stop if we tried all running tasks */

Would say s/running/runnable but I see that we do use running/runnable
interchangeably.

> +			if (env.loop < busiest->nr_running)
> +				goto more_balance;
>  		}
>  
>  		/*

IMHO, there will be some interaction with the `All tasks on this
runqueue were pinned by CPU affinity` check at the end of load_balance()?