linux-kernel - Re: [RFC PATCH 3/3] sched/fair: Change @running of __update_load_avg() to @update

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:	Thu, 2 Jun 2016 10:25:39 +0100
From:	Juri Lelli <juri.lelli@....com>
To:	Dietmar Eggemann <dietmar.eggemann@....com>
Cc:	Peter Zijlstra <peterz@...radead.org>,
	linux-kernel@...r.kernel.org,
	Vincent Guittot <vincent.guittot@...aro.org>,
	Ben Segall <bsegall@...gle.com>,
	Morten Rasmussen <morten.rasmussen@....com>,
	Yuyang Du <yuyang.du@...el.com>
Subject: Re: [RFC PATCH 3/3] sched/fair: Change @running of
 __update_load_avg() to @update_util

Hi,

another minor comment below. :-)

On 01/06/16 20:39, Dietmar Eggemann wrote:
> The information whether a se/cfs_rq should get its load and
> utilization (se representing a task and root cfs_rq) or only its load
> (se representing a task group and cfs_rq owned by this se) updated can
> be passed into __update_load_avg() to avoid the additional if/else
> condition to set update_util.
> 
> @running is changed to @update_util which now carries the information if
> the utilization of the se/cfs_rq should be updated and if the se/cfs_rq
> is running or not.
> 
> Signed-off-by: Dietmar Eggemann <dietmar.eggemann@....com>
> ---
>  kernel/sched/fair.c | 42 +++++++++++++++++++++---------------------
>  1 file changed, 21 insertions(+), 21 deletions(-)
> 
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 3ae8e79fb687..a1c13975cf56 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -2669,6 +2669,10 @@ static u32 __compute_runnable_contrib(u64 n)
>  
>  #define cap_scale(v, s) ((v)*(s) >> SCHED_CAPACITY_SHIFT)
>  
> +#define upd_util_se(se, rng) ((entity_is_task(se) << 1) | (rng))
> +#define upd_util_cfs_rq(cfs_rq) \
> +		(((&rq_of(cfs_rq)->cfs == cfs_rq) << 1) | !!cfs_rq->curr)
> +
>  /*
>   * We can represent the historical contribution to runnable average as the
>   * coefficients of a geometric series.  To do this we sub-divide our runnable
> @@ -2699,13 +2703,12 @@ static u32 __compute_runnable_contrib(u64 n)
>   */
>  static __always_inline int
>  __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
> -		  unsigned long weight, int running, struct cfs_rq *cfs_rq)
> +		  unsigned long weight, int update_util, struct cfs_rq *cfs_rq)
>  {
>  	u64 delta, scaled_delta, periods;
>  	u32 contrib;
>  	unsigned int delta_w, scaled_delta_w, decayed = 0;
>  	unsigned long scale_freq, scale_cpu;
> -	int update_util = 0;
>  
>  	delta = now - sa->last_update_time;
>  	/*
> @@ -2726,12 +2729,6 @@ __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
>  		return 0;
>  	sa->last_update_time = now;
>  
> -	if (cfs_rq) {
> -		if (&rq_of(cfs_rq)->cfs == cfs_rq)
> -			update_util = 1;
> -	} else if (entity_is_task(container_of(sa, struct sched_entity, avg)))
> -			update_util = 1;
> -
>  	scale_freq = arch_scale_freq_capacity(NULL, cpu);
>  	scale_cpu = arch_scale_cpu_capacity(NULL, cpu);
>  
> @@ -2757,7 +2754,7 @@ __update_load_avg(u64 now, int cpu, struct sched_avg *sa,
>  						weight * scaled_delta_w;
>  			}
>  		}
> -		if (update_util && running)
> +		if (update_util == 0x3)

How about a define for these masks?

Best,

- Juri