mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Zhang, Rui" <rui.zhang@intel.com>
To: "linux-pm@vger.kernel.org" <linux-pm@vger.kernel.org>,
	"rjw@rjwysocki.net" <rjw@rjwysocki.net>
Cc: "lukasz.luba@arm.com" <lukasz.luba@arm.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"daniel.lezcano@linaro.org" <daniel.lezcano@linaro.org>,
	"peter@piie.net" <peter@piie.net>
Subject: Re: [PATCH v1 4/4] thermal: gov_bang_bang: Use governor_data to reduce overhead
Date: Wed, 14 Aug 2024 06:09:12 +0000	[thread overview]
Message-ID: <2f16fd5a59d6655ec9339473d516ac49c89e43f8.camel@intel.com> (raw)
In-Reply-To: <2285575.iZASKD2KPV@rjwysocki.net>

On Tue, 2024-08-13 at 16:29 +0200, Rafael J. Wysocki wrote:
> From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> 
> After running once, the for_each_trip_desc() loop in
> bang_bang_manage() is pure needless overhead because it is not going
> to
> make any changes unless a new cooling device has been bound to one of
> the trips in the thermal zone or the system is resuming from sleep.
> 
> For this reason, make bang_bang_manage() set governor_data for the
> thermal zone and check it upfront to decide whether or not it needs
> to
> do anything.
> 
> However, governor_data needs to be reset in some cases to let
> bang_bang_manage() know that it should walk the trips again, so add
> an
> .update_tz() callback to the governor and make the core additionally
> invoke it during system resume.
> 
> To avoid affecting the other users of that callback unnecessarily,
> add
> a special notification reason for system resume, THERMAL_TZ_RESUME,
> and
> also pass it to __thermal_zone_device_update() called during system
> resume for consistency.
> 
> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> ---
>  drivers/thermal/gov_bang_bang.c |   18 ++++++++++++++++++
>  drivers/thermal/thermal_core.c  |    3 ++-
>  include/linux/thermal.h         |    1 +
>  3 files changed, 21 insertions(+), 1 deletion(-)
> 
> Index: linux-pm/drivers/thermal/gov_bang_bang.c
> ===================================================================
> --- linux-pm.orig/drivers/thermal/gov_bang_bang.c
> +++ linux-pm/drivers/thermal/gov_bang_bang.c
> @@ -86,6 +86,10 @@ static void bang_bang_manage(struct ther
>         const struct thermal_trip_desc *td;
>         struct thermal_instance *instance;
>  
> +       /* If the code below has run already, nothing needs to be
> done. */
> +       if (tz->governor_data)
> +               return;
> +
>         for_each_trip_desc(tz, td) {
>                 const struct thermal_trip *trip = &td->trip;
>  
> @@ -107,11 +111,25 @@ static void bang_bang_manage(struct ther
>                                 bang_bang_set_instance_target(instanc
> e, 0);
>                 }
>         }
> +
> +       tz->governor_data = (void *)true;
> +}
> +
> +static void bang_bang_update_tz(struct thermal_zone_device *tz,
> +                               enum thermal_notify_event reason)
> +{
> +       /*
> +        * Let bang_bang_manage() know that it needs to walk trips
> after binding
> +        * a new cdev and after system resume.
> +        */
> +       if (reason == THERMAL_TZ_BIND_CDEV || reason ==
> THERMAL_TZ_RESUME)
> +               tz->governor_data = NULL;
>  }

can we do the cdev initialization for BIND_CDEV and RESUME notification
in .update_tz() directly?

Then we don't need .manage() callback. This makes more sense to me
because bang_bang governor cares about trip point crossing only.

thanks,
rui
>  
>  static struct thermal_governor thermal_gov_bang_bang = {
>         .name           = "bang_bang",
>         .trip_crossed   = bang_bang_control,
>         .manage         = bang_bang_manage,
> +       .update_tz      = bang_bang_update_tz,
>  };
>  THERMAL_GOVERNOR_DECLARE(thermal_gov_bang_bang);
> Index: linux-pm/drivers/thermal/thermal_core.c
> ===================================================================
> --- linux-pm.orig/drivers/thermal/thermal_core.c
> +++ linux-pm/drivers/thermal/thermal_core.c
> @@ -1692,7 +1692,8 @@ static void thermal_zone_device_resume(s
>  
>         thermal_debug_tz_resume(tz);
>         thermal_zone_device_init(tz);
> -       __thermal_zone_device_update(tz, THERMAL_EVENT_UNSPECIFIED);
> +       thermal_governor_update_tz(tz, THERMAL_TZ_RESUME);
> +       __thermal_zone_device_update(tz, THERMAL_TZ_RESUME);
>  
>         complete(&tz->resume);
>         tz->resuming = false;
> Index: linux-pm/include/linux/thermal.h
> ===================================================================
> --- linux-pm.orig/include/linux/thermal.h
> +++ linux-pm/include/linux/thermal.h
> @@ -55,6 +55,7 @@ enum thermal_notify_event {
>         THERMAL_TZ_BIND_CDEV, /* Cooling dev is bind to the thermal
> zone */
>         THERMAL_TZ_UNBIND_CDEV, /* Cooling dev is unbind from the
> thermal zone */
>         THERMAL_INSTANCE_WEIGHT_CHANGED, /* Thermal instance weight
> changed */
> +       THERMAL_TZ_RESUME, /* Thermal zone is resuming after system
> sleep */
>  };
>  
>  /**
> 
> 
> 


  parent reply	other threads:[~2024-08-14  6:09 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-13 14:23 [PATCH v1 0/4] thermal: gov_bang_bang: Prevent cooling devices from getting stuck in the "on" state Rafael J. Wysocki
2024-08-13 14:25 ` [PATCH v1 1/4] thermal: gov_bang_bang: Call __thermal_cdev_update() directly Rafael J. Wysocki
2024-08-13 21:03   ` Peter Kästle
2024-08-14  6:18   ` Zhang, Rui
2024-08-14 17:34     ` Rafael J. Wysocki
2024-08-16  3:00   ` Zhang, Rui
2024-08-13 14:26 ` [PATCH v1 2/4] thermal: gov_bang_bang: Split bang_bang_control() Rafael J. Wysocki
2024-08-13 21:03   ` Peter Kästle
2024-08-16  3:00   ` Zhang, Rui
2024-08-13 14:27 ` [PATCH v1 3/4] thermal: gov_bang_bang: Add .manage() callback Rafael J. Wysocki
2024-08-13 21:04   ` Peter Kästle
2024-08-13 21:07   ` Peter Kästle
2024-08-14 17:18     ` Rafael J. Wysocki
2024-08-14 20:55       ` Peter Kästle
2024-08-16  3:00   ` Zhang, Rui
2024-08-13 14:29 ` [PATCH v1 4/4] thermal: gov_bang_bang: Use governor_data to reduce overhead Rafael J. Wysocki
2024-08-13 21:08   ` Peter Kästle
2024-08-14  6:09   ` Zhang, Rui [this message]
2024-08-14 17:26     ` Rafael J. Wysocki
2024-08-15  3:26       ` Zhang, Rui
2024-08-15 12:35         ` Rafael J. Wysocki
2024-08-16  2:59           ` Zhang, Rui

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=2f16fd5a59d6655ec9339473d516ac49c89e43f8.camel@intel.com \
    --to=rui.zhang@intel.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=lukasz.luba@arm.com \
    --cc=peter@piie.net \
    --cc=rjw@rjwysocki.net \
    /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

all inboxes | Powered by JetHome®