Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:6 as permitted sender) client-ip=2620:137:e000::3:6;
MIME-Version: 1.0
References: <1957441.PYKUYFuaPT@kreacher> <1882755.CQOukoFCf9@kreacher>
 <3c9f7b7d-a9ca-79ae-4e64-367a82f953bc@linaro.org> <CAJZ5v0ifXSWd4QL+j-=8OBr1aQr0WAeMBGYtis_gkHJDBLeWMg@mail.gmail.com>
 <d4bdea97-1e92-6422-01c4-bf7caca4ea45@linaro.org>
In-Reply-To: <d4bdea97-1e92-6422-01c4-bf7caca4ea45@linaro.org>
From:   "Rafael J. Wysocki" <rafael@kernel.org>
Date:   Wed, 27 Sep 2023 18:09:44 +0200
Message-ID: <CAJZ5v0hOJYuaJ2VXPCrHZspOjj342+NkB=nHXPT6YKvy90XXBA@mail.gmail.com>
Subject: Re: [PATCH v1 06/13] thermal: gov_fair_share: Rearrange get_trip_level()
To:     Daniel Lezcano <daniel.lezcano@linaro.org>
Cc:     "Rafael J. Wysocki" <rafael@kernel.org>,
        "Rafael J. Wysocki" <rjw@rjwysocki.net>,
        Linux PM <linux-pm@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux ACPI <linux-acpi@vger.kernel.org>,
        Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
        Zhang Rui <rui.zhang@intel.com>,
        Lukasz Luba <lukasz.luba@arm.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8BIT
Precedence: bulk

On Wed, Sep 27, 2023 at 5:37 PM Daniel Lezcano
<daniel.lezcano@linaro.org> wrote:
>
> On 27/09/2023 17:06, Rafael J. Wysocki wrote:
> > On Wed, Sep 27, 2023 at 5:00 PM Daniel Lezcano
> > <daniel.lezcano@linaro.org> wrote:
> >>
> >> On 21/09/2023 19:54, Rafael J. Wysocki wrote:
> >>> From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> >>>
> >>> Make get_trip_level() access the thermal zone's trip table directly
> >>> instead of using __thermal_zone_get_trip() which adds overhead related
> >>> to the unnecessary bounds checking and copying the trip point data.
> >>>
> >>> Also rearrange the code in it to make it somewhat easier to follow.
> >>>
> >>> The general functionality is not expected to be changed.
> >>>
> >>> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> >>> ---
> >>>    drivers/thermal/gov_fair_share.c |   22 ++++++++++------------
> >>>    1 file changed, 10 insertions(+), 12 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
> >>> @@ -21,23 +21,21 @@
> >>>     */
> >>>    static int get_trip_level(struct thermal_zone_device *tz)
> >>>    {
> >>> -     struct thermal_trip trip;
> >>> -     int count;
> >>> +     const struct thermal_trip *trip = tz->trips;
> >>> +     int i;
> >>>
> >>> -     for (count = 0; count < tz->num_trips; count++) {
> >>> -             __thermal_zone_get_trip(tz, count, &trip);
> >>> -             if (tz->temperature < trip.temperature)
> >>> +     if (tz->temperature < trip->temperature)
> >>> +             return 0;
> >>> +
> >>> +     for (i = 0; i < tz->num_trips - 1; i++) {
> >>> +             trip++;
> >>> +             if (tz->temperature < trip->temperature)
> >>>                        break;
> >>>        }
> >>
> >> Is it possible to use for_each_thermal_trip() instead ? That would make
> >> the code more self-encapsulate
> >
> > It is possible in principle, but this is a governor which is regarded
> > as part of the core, isn't it?
> >
> > So is an extra overhead related to using a callback (which may be
> > subject to retpolines and such) really justified in this case?
>
>  From my POV, all trip points browsing should be replaced by
> for_each_thermal_trip() so any change in the future in how we go through
> the existing thermal trips will impact one place.
>
> If the routine needs to be optimized, that is something we can do also
> (may be an inline the callback?)

OK