Message-ID: <afcc7b08ebe2578d32e6595d258afeec3e73512e.camel@linux.intel.com>
Date:   Tue, 17 May 2022 12:02:13 -0700
From:   srinivas pandruvada <srinivas.pandruvada@...ux.intel.com>
To:     "Rafael J. Wysocki" <rafael@...nel.org>
Cc:     Daniel Lezcano <daniel.lezcano@...exp.org>,
        Daniel Lezcano <daniel.lezcano@...aro.org>,
        Kevin Hilman <khilman@...libre.com>,
        Alexandre Bailon <abailon@...libre.com>,
        Linux PM <linux-pm@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Amit Kucheria <amitk@...nel.org>,
        Zhang Rui <rui.zhang@...el.com>,
        Jonathan Corbet <corbet@....net>, Len Brown <lenb@...nel.org>,
        Raju Rangoju <rajur@...lsio.com>,
        "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Ido Schimmel <idosch@...dia.com>,
        Petr Machata <petrm@...dia.com>,
        Luca Coelho <luciano.coelho@...el.com>,
        Kalle Valo <kvalo@...nel.org>, Peter Kaestle <peter@...e.net>,
        Hans de Goede <hdegoede@...hat.com>,
        Mark Gross <markgross@...nel.org>,
        Sebastian Reichel <sre@...nel.org>,
        Miquel Raynal <miquel.raynal@...tlin.com>,
        Support Opensource <support.opensource@...semi.com>,
        Shawn Guo <shawnguo@...nel.org>,
        Sascha Hauer <s.hauer@...gutronix.de>,
        Pengutronix Kernel Team <kernel@...gutronix.de>,
        Fabio Estevam <festevam@...il.com>,
        NXP Linux Team <linux-imx@....com>,
        Niklas Söderlund 
        <niklas.soderlund@...natech.se>,
        Miri Korenblit <miriam.rachel.korenblit@...el.com>,
        Johannes Berg <johannes.berg@...el.com>,
        Sumeet Pawnikar <sumeet.r.pawnikar@...el.com>,
        Dan Carpenter <dan.carpenter@...cle.com>,
        Chuansheng Liu <chuansheng.liu@...el.com>,
        Jiasheng Jiang <jiasheng@...as.ac.cn>,
        Antoine Tenart <atenart@...nel.org>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        "open list:DOCUMENTATION" <linux-doc@...r.kernel.org>,
        "open list:ACPI THERMAL DRIVER" <linux-acpi@...r.kernel.org>,
        "open list:CXGB4 ETHERNET DRIVER (CXGB4)" <netdev@...r.kernel.org>,
        "open list:INTEL WIRELESS WIFI LINK (iwlwifi)" 
        <linux-wireless@...r.kernel.org>,
        "open list:ACER ASPIRE ONE TEMPERATURE AND FAN DRIVER" 
        <platform-driver-x86@...r.kernel.org>,
        "moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE" 
        <linux-arm-kernel@...ts.infradead.org>,
        "open list:RENESAS R-CAR THERMAL DRIVERS" 
        <linux-renesas-soc@...r.kernel.org>
Subject: Re: [PATCH v2 01/14] thermal/core: Change thermal_zone_ops to
 thermal_sensor_ops

On Tue, 2022-05-17 at 20:53 +0200, Rafael J. Wysocki wrote:
> On Tue, May 17, 2022 at 6:51 PM srinivas pandruvada
> <srinivas.pandruvada@...ux.intel.com> wrote:
> > 
> > On Tue, 2022-05-17 at 17:42 +0200, Rafael J. Wysocki wrote:
> > > On Sat, May 7, 2022 at 2:55 PM Daniel Lezcano
> > > <daniel.lezcano@...exp.org> wrote:
> > > > 
> > > > A thermal zone is software abstraction of a sensor associated
> > > > with
> > > > properties and cooling devices if any.
> > > > 
> > > > The fact that we have thermal_zone and thermal_zone_ops mixed
> > > > is
> > > > confusing and does not clearly identify the different
> > > > components
> > > > entering in the thermal management process. A thermal zone
> > > > appears
> > > > to
> > > > be a sensor while it is not.
> > > 
> > > Well, the majority of the operations in thermal_zone_ops don't
> > > apply
> > > to thermal sensors.  For example, ->set_trips(), -
> > > >get_trip_type(),
> > > ->get_trip_temp().
> > > 
> > In past we discussed adding thermal sensor sysfs with threshold to
> > notify temperature.
> > 
> > So sensor can have set/get_threshold() functions instead of the
> > set/get_trip for zones.
> > 
> > Like we have /sys/class/thermal_zone* we can have
> > /sys/class/thermal_sensor*.
> 
> Exactly, so renaming thermal_zone_ops as thermal_sensor_ops isn't
> quite helpful in this respect.
> 
> IMO there should be operations for sensors and there should be
> operations for thermal zones and those two sets of operations should
> be different.
> 
> > Thermal sensor(s) are bound to  thermal zones.
> 
> So I think that this binding should be analogous to the binding
> between thermal zones and cooling devices.
> 
> > This can also include multiple sensors in a zone and can create a
> > virtual sensor also.
> 
> It can.
> 
> However, what's the difference between a thermal zone with multiple
> sensors and a thermal zone with one virtual sensor being an aggregate
> of multiple physical sensors?
> 
Either way is fine. A thermal sensor can be aggregate of other sensors.

> Both involve some type of aggregation of temperature values measured
> by the physical sensors.
> 
> > > > In order to set the scene for multiple thermal sensors
> > > > aggregated
> > > > into
> > > > a single thermal zone. Rename the thermal_zone_ops to
> > > > thermal_sensor_ops, that will appear clearyl the thermal zone
> > > > is
> > > > not a
> > > > sensor but an abstraction of one [or multiple] sensor(s).
> > > 
> > > So I'm not convinced that the renaming mentioned above is
> > > particularly
> > > clean either.
> > > 
> > > IMV the way to go would be to split the thermal sensor
> > > operations,
> > > like ->get_temp(), out of thermal_zone_ops.
> > > 
> > > But then it is not clear what a thermal zone with multiple
> > > sensors in
> > > it really means.  I guess it would require an aggregation
> > > function to
> > > combine the thermal sensors in it that would produce an effective
> > > temperature to check against the trip points.
> > > 
> > > Honestly, I don't think that setting a separate set of trips for
> > > each
> > > sensor in a thermal zone would make a lot of sense.
> > 

Powered by blists - more mailing lists