public inbox for linux-pm@vger.kernel.org
 help / color / mirror / Atom feed
From: Lukasz Luba <lukasz.luba@arm.com>
To: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Linux PM <linux-pm@vger.kernel.org>,
	Daniel Lezcano <daniel.lezcano@linaro.org>,
	Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Subject: Re: [PATCH v1 13/16] thermal: gov_fair_share: Eliminate unnecessary integer divisions
Date: Fri, 19 Apr 2024 19:46:01 +0100	[thread overview]
Message-ID: <a8f40452-018b-40d6-9999-ad8b61834448@arm.com> (raw)
In-Reply-To: <2170379.OBFZWjSADL@kreacher>



On 4/10/24 18:00, Rafael J. Wysocki wrote:
> From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> 
> The computations carried out by fair_share_throttle() for each trip
> point include at least one redundant integer division which introduces
> superfluous rounding errors.  Also the multiplications by 100 in it are
> not really necessary and can be eliminated.
> 
> Rearrange fair_share_throttle() to carry out only one integer division per
> trip and only as many integer multiplications as necessary and rename one
> variable in it (while at it).
> 
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> ---
>   drivers/thermal/gov_fair_share.c |   32 +++++++++++++++-----------------
>   1 file changed, 15 insertions(+), 17 deletions(-)
> 
> Index: linux-pm/drivers/thermal/gov_fair_share.c
> ===================================================================
> --- linux-pm.orig/drivers/thermal/gov_fair_share.c
> +++ linux-pm/drivers/thermal/gov_fair_share.c
> @@ -41,12 +41,6 @@ static int get_trip_level(struct thermal
>   	return trip_level;
>   }
>   
> -static long get_target_state(struct thermal_zone_device *tz,
> -		struct thermal_cooling_device *cdev, int percentage, int level)
> -{
> -	return (long)(percentage * level * cdev->max_state) / (100 * tz->num_trips);
> -}
> -
>   /**
>    * fair_share_throttle - throttles devices associated with the given zone
>    * @tz: thermal_zone_device
> @@ -58,7 +52,7 @@ static long get_target_state(struct ther
>    *
>    * Parameters used for Throttling:
>    * P1. max_state: Maximum throttle state exposed by the cooling device.
> - * P2. percentage[i]/100:
> + * P2. weight[i]/total_weight:
>    *	How 'effective' the 'i'th device is, in cooling the given zone.
>    * P3. trip_level/max_no_of_trips:
>    *	This describes the extent to which the devices should be throttled.
> @@ -72,30 +66,34 @@ static void fair_share_throttle(struct t
>   {
>   	struct thermal_instance *instance;
>   	int total_weight = 0;
> -	int total_instance = 0;
> +	int nr_instances = 0;
>   
>   	list_for_each_entry(instance, &tz->thermal_instances, tz_node) {
>   		if (instance->trip != trip)
>   			continue;
>   
>   		total_weight += instance->weight;
> -		total_instance++;
> +		nr_instances++;
>   	}
>   
>   	list_for_each_entry(instance, &tz->thermal_instances, tz_node) {
> -		int percentage;
>   		struct thermal_cooling_device *cdev = instance->cdev;
> +		u64 dividend;
> +		u32 divisor;
>   
>   		if (instance->trip != trip)
>   			continue;
>   
> -		if (!total_weight)
> -			percentage = 100 / total_instance;
> -		else
> -			percentage = (instance->weight * 100) / total_weight;
> -
> -		instance->target = get_target_state(tz, cdev, percentage,
> -						    trip_level);
> +		dividend = trip_level;
> +		dividend *= cdev->max_state;
> +		divisor = tz->num_trips;
> +		if (total_weight) {
> +			dividend *= instance->weight;
> +			divisor *= total_weight;
> +		} else {
> +			divisor *= nr_instances;
> +		}
> +		instance->target = div_u64(dividend, divisor);
>   
>   		mutex_lock(&cdev->lock);
>   		__thermal_cdev_update(cdev);
> 
> 
> 


Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>

  reply	other threads:[~2024-04-19 18:45 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-10 15:41 [PATCH v1 00/16] thermal: core: Redesign the governor interface Rafael J. Wysocki
2024-04-10 16:04 ` [PATCH v1 02/16] thermal: gov_bang_bang: Use .trip_crossed() instead of .throttle() Rafael J. Wysocki
2024-04-19  9:34   ` Lukasz Luba
2024-04-23 17:04   ` Daniel Lezcano
2024-04-10 16:05 ` [PATCH v1 03/16] thermal: gov_bang_bang: Clean up thermal_zone_trip_update() Rafael J. Wysocki
2024-04-19  9:41   ` Lukasz Luba
2024-04-23 17:05   ` Daniel Lezcano
2024-04-10 16:06 ` [PATCH v1 04/16] thermal: gov_bang_bang: Fold thermal_zone_trip_update() into its caller Rafael J. Wysocki
2024-04-19  9:42   ` Lukasz Luba
2024-04-23 17:06   ` Daniel Lezcano
2024-04-10 16:08 ` [PATCH v1 05/16] thermal: core: Introduce .manage() callback for thermal governors Rafael J. Wysocki
2024-04-19  9:44   ` Lukasz Luba
2024-04-23 17:07   ` Daniel Lezcano
2024-04-10 16:10 ` [PATCH v1 06/16] thermal: gov_power_allocator: Use .manage() callback instead of .throttle() Rafael J. Wysocki
2024-04-19  9:50   ` Lukasz Luba
2024-04-10 16:10 ` [PATCH v1 01/16] thermal: core: Introduce .trip_crossed() callback for thermal governors Rafael J. Wysocki
2024-04-19  8:47   ` Lukasz Luba
2024-04-23 17:14   ` Daniel Lezcano
2024-04-23 17:25     ` Rafael J. Wysocki
2024-04-23 17:58       ` Daniel Lezcano
2024-04-10 16:12 ` [PATCH v1 07/16] thermal: gov_power_allocator: Eliminate a redundant variable Rafael J. Wysocki
2024-04-19  9:56   ` Lukasz Luba
2024-04-23 17:35   ` Daniel Lezcano
2024-04-23 17:54     ` Rafael J. Wysocki
2024-04-23 18:00       ` Daniel Lezcano
2024-04-23 18:05         ` Rafael J. Wysocki
2024-04-23 18:09           ` Daniel Lezcano
2024-04-10 16:13 ` [PATCH v1 08/16] thermal: gov_step_wise: Use .manage() callback instead of .throttle() Rafael J. Wysocki
2024-04-19 17:20   ` Lukasz Luba
2024-04-24  7:54   ` Daniel Lezcano
2024-04-10 16:43 ` [PATCH v1 09/16] thermal: gov_step_wise: Use trip thresholds instead of trip temperatures Rafael J. Wysocki
2024-04-19 19:11   ` Lukasz Luba
2024-04-24  7:03   ` Daniel Lezcano
2024-04-10 16:44 ` [PATCH v1 10/16] thermal: gov_step_wise: Clean up thermal_zone_trip_update() Rafael J. Wysocki
2024-04-19 18:21   ` Lukasz Luba
2024-04-10 16:57 ` [PATCH v1 11/16] thermal: gov_fair_share: Use .manage() callback instead of .throttle() Rafael J. Wysocki
2024-04-19 18:28   ` Lukasz Luba
2024-04-10 16:58 ` [PATCH v1 12/16] thermal: gov_fair_share: Use trip thresholds instead of trip temperatures Rafael J. Wysocki
2024-04-19 19:18   ` Lukasz Luba
2024-04-10 17:00 ` [PATCH v1 13/16] thermal: gov_fair_share: Eliminate unnecessary integer divisions Rafael J. Wysocki
2024-04-19 18:46   ` Lukasz Luba [this message]
2024-04-10 17:03 ` [PATCH v1 14/16] thermal: gov_user_space: Use .trip_crossed() instead of .throttle() Rafael J. Wysocki
2024-04-19 18:16   ` Lukasz Luba
2024-04-24  9:14   ` Daniel Lezcano
2024-04-24 11:32     ` Srinivas Pandruvada
2024-04-24 11:34       ` Daniel Lezcano
2024-04-24 11:44         ` Srinivas Pandruvada
2024-04-10 17:42 ` [PATCH v1 15/16] thermal: core: Drop the .throttle() governor callback Rafael J. Wysocki
2024-04-19 18:50   ` Lukasz Luba
2024-04-24  9:11   ` Daniel Lezcano
2024-04-10 17:44 ` [PATCH v1 16/16] thermal: core: Relocate critical and hot trip handling Rafael J. Wysocki
2024-04-19 18:52   ` Lukasz Luba
2024-04-24  9:17   ` Daniel Lezcano

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=a8f40452-018b-40d6-9999-ad8b61834448@arm.com \
    --to=lukasz.luba@arm.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=rjw@rjwysocki.net \
    --cc=srinivas.pandruvada@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox