linux-kernel - Re: [PATCH 1/2] thermal: core: Move thermal_cdev

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] [day] [month] [year] [list]

Message-ID: <a91cf79e0ee7da4e0ac108168eceb5c63c78f8e1.camel@intel.com>
Date:   Sat, 11 Apr 2020 10:55:42 +0800
From:   Zhang Rui <rui.zhang@...el.com>
To:     Daniel Lezcano <daniel.lezcano@...aro.org>
Cc:     linux-kernel@...r.kernel.org, amit.kucheria@...durent.com,
        "open list:THERMAL" <linux-pm@...r.kernel.org>
Subject: Re: [PATCH 1/2] thermal: core: Move thermal_cdev_update next to
 updated=false

On Fri, 2020-04-10 at 13:26 +0200, Daniel Lezcano wrote:
> On 10/04/2020 12:14, Zhang Rui wrote:
> > Hi, Daniel,
> > 
> > On Thu, 2020-04-09 at 17:15 +0200, Daniel Lezcano wrote:
> > > The call to the thermal_cdev_update() function is done after
> > > browsing
> > > the thermal instances which sets the updated flag by browsing
> > > them
> > > again.
> > > 
> > > Instead of doing this, let's move the call right after setting
> > > the
> > > cooling device 'updated' flag as it is done in the other
> > > governors.
> > 
> > The reason we do this in two steps is that we want to avoid
> > redundant
> > cooling device state changes.
> > 
> > Further more, I think it is better to move the thermal_cdev_update
> > out
> > of .throllte() callback, to thermal_zone_device_update(). So that
> > we do
> > not need to update the cooling device for each trip point.
> > 
> > is there any specific reason we need to do thermal_cdev_update()
> > for
> > every potential change?
> 
> I agree we can go further and move the cooling device update in the
> thermal_zone_device_update() by letting the throttle callback let us
> know an update is needed with the return value.
> 
> Makes sense to provide more changes on top of those two patches ?

Hmmm, without the update flag, we can only updating all the cooling
devices blindly. And this is time consuming for some cooling devices.

thanks,
rui

> 
> > > Signed-off-by: Daniel Lezcano <daniel.lezcano@...aro.org>
> > > ---
> > >  drivers/thermal/gov_bang_bang.c | 10 +---------
> > >  drivers/thermal/step_wise.c     | 10 +---------
> > >  2 files changed, 2 insertions(+), 18 deletions(-)
> > > 
> > > diff --git a/drivers/thermal/gov_bang_bang.c
> > > b/drivers/thermal/gov_bang_bang.c
> > > index 991a1c54296d..c292a69845bb 100644
> > > --- a/drivers/thermal/gov_bang_bang.c
> > > +++ b/drivers/thermal/gov_bang_bang.c
> > > @@ -64,6 +64,7 @@ static void thermal_zone_trip_update(struct
> > > thermal_zone_device *tz, int trip)
> > >  		mutex_lock(&instance->cdev->lock);
> > >  		instance->cdev->updated = false; /* cdev needs update
> > > */
> > >  		mutex_unlock(&instance->cdev->lock);
> > > +		thermal_cdev_update(instance->cdev);
> > >  	}
> > >  
> > >  	mutex_unlock(&tz->lock);
> > > @@ -98,17 +99,8 @@ static void thermal_zone_trip_update(struct
> > > thermal_zone_device *tz, int trip)
> > >   */
> > >  static int bang_bang_control(struct thermal_zone_device *tz, int
> > > trip)
> > >  {
> > > -	struct thermal_instance *instance;
> > > -
> > >  	thermal_zone_trip_update(tz, trip);
> > >  
> > > -	mutex_lock(&tz->lock);
> > > -
> > > -	list_for_each_entry(instance, &tz->thermal_instances, tz_node)
> > > -		thermal_cdev_update(instance->cdev);
> > > -
> > > -	mutex_unlock(&tz->lock);
> > > -
> > >  	return 0;
> > >  }
> > >  
> > > diff --git a/drivers/thermal/step_wise.c
> > > b/drivers/thermal/step_wise.c
> > > index 2ae7198d3067..298eedac0293 100644
> > > --- a/drivers/thermal/step_wise.c
> > > +++ b/drivers/thermal/step_wise.c
> > > @@ -167,6 +167,7 @@ static void thermal_zone_trip_update(struct
> > > thermal_zone_device *tz, int trip)
> > >  		mutex_lock(&instance->cdev->lock);
> > >  		instance->cdev->updated = false; /* cdev needs update
> > > */
> > >  		mutex_unlock(&instance->cdev->lock);
> > > +		thermal_cdev_update(instance->cdev);
> > >  	}
> > >  
> > >  	mutex_unlock(&tz->lock);
> > > @@ -185,20 +186,11 @@ static void thermal_zone_trip_update(struct
> > > thermal_zone_device *tz, int trip)
> > >   */
> > >  static int step_wise_throttle(struct thermal_zone_device *tz,
> > > int
> > > trip)
> > >  {
> > > -	struct thermal_instance *instance;
> > > -
> > >  	thermal_zone_trip_update(tz, trip);
> > >  
> > >  	if (tz->forced_passive)
> > >  		thermal_zone_trip_update(tz, THERMAL_TRIPS_NONE);
> > >  
> > > -	mutex_lock(&tz->lock);
> > > -
> > > -	list_for_each_entry(instance, &tz->thermal_instances, tz_node)
> > > -		thermal_cdev_update(instance->cdev);
> > > -
> > > -	mutex_unlock(&tz->lock);
> > > -
> > >  	return 0;
> > >  }
> > >  
> 
>