linux-kernel - Re: [PATCH] sched/uclamp: New sysctl to tweak boost for uclamp tasks

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]

Message-ID: <YgPIzXWpWeEpUJbi@FVFF7649Q05P>
Date:   Wed, 9 Feb 2022 13:59:41 +0000
From:   Vincent Donnefort <vincent.donnefort@....com>
To:     Chitti Babu Theegala <quic_ctheegal@...cinc.com>
Cc:     mingo@...hat.com, peterz@...radead.org, juri.lelli@...hat.com,
        vincent.guittot@...aro.org, dietmar.eggemann@....com,
        rostedt@...dmis.org, joel@...lfernandes.org,
        linux-arm-msm@...r.kernel.org, quic_lingutla@...cinc.com,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] sched/uclamp: New sysctl to tweak boost for uclamp tasks

[...]

> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 634d571e39d77..55019eb9c83c5 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -1269,6 +1269,14 @@ unsigned int sysctl_sched_uclamp_util_max = SCHED_CAPACITY_SCALE;
>   */
>  unsigned int sysctl_sched_uclamp_util_min_rt_default = SCHED_CAPACITY_SCALE;
>  
> +/* UCLAMP_MIN knob clamps the util_value within uclamp_min & max. By default,
> + * these uclamped tasks are also boosted to bigger cores which can impact power.
> + *
> + * This knob provides an option to turn-off 'boosting to gold cores' and
> + * just restrict to 'clamping util value' job.
> + */
> +unsigned int sysctl_sched_uclamp_util_min_to_boost = 1;
> +
>  /* All clamps are required to be less or equal than these values */
>  static struct uclamp_se uclamp_default[UCLAMP_CNT];
>  
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 7cd316793f4e3..44adcbfa9eb32 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -6829,7 +6829,7 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu, int sy
>  		goto unlock;
>  
>  	latency_sensitive = uclamp_latency_sensitive(p);

I don't think this is the right tree.


> -	boosted = uclamp_boosted(p);
> +	boosted = sysctl_sched_uclamp_util_min_to_boost && uclamp_boosted(p);
>  	target_cap = boosted ? 0 : ULONG_MAX;
>  
>  	for (; pd; pd = pd->next) {
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index 8f196ed2ec065..6463d60fed84e 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -1882,6 +1882,15 @@ static struct ctl_table kern_table[] = {
>  		.mode		= 0644,
>  		.proc_handler	= sysctl_sched_uclamp_handler,
>  	},
> +	{
> +		.procname	= "sched_util_clamp_min_to_boost",
> +		.data		= &sysctl_sched_uclamp_util_min_to_boost,
> +		.maxlen		= sizeof(unsigned int),
> +		.mode		= 0644,
> +		.proc_handler	= proc_dointvec_minmax,
> +		.extra1		= SYSCTL_ZERO,
> +		.extra2		= SYSCTL_ONE,
> +	},
>  #endif
>  #ifdef CONFIG_SCHED_AUTOGROUP
>  	{
> -- 
> 2.17.1
>