On Thu, Mar 9, 2023 at 6:39 PM Daniel Lezcano <daniel.lezcano@xxxxxxxxxx> wrote:
>
> On 09/03/2023 05:40, Chen-Yu Tsai wrote:
> > On Wed, Mar 8, 2023 at 12:46 AM <bchihi@xxxxxxxxxxxx> wrote:
> >>
> >> From: Balsam CHIHI <bchihi@xxxxxxxxxxxx>
> >>
> >> Add AP Domain to LVTS thermal controllers dt-binding definition for mt8195.
> >>
> >> Signed-off-by: Balsam CHIHI <bchihi@xxxxxxxxxxxx>
> >> ---
> >> include/dt-bindings/thermal/mediatek,lvts-thermal.h | 10 ++++++++++
> >> 1 file changed, 10 insertions(+)
> >>
> >> diff --git a/include/dt-bindings/thermal/mediatek,lvts-thermal.h b/include/dt-bindings/thermal/mediatek,lvts-thermal.h
> >> index c09398920468..8fa5a46675c4 100644
> >> --- a/include/dt-bindings/thermal/mediatek,lvts-thermal.h
> >> +++ b/include/dt-bindings/thermal/mediatek,lvts-thermal.h
> >> @@ -16,4 +16,14 @@
> >> #define MT8195_MCU_LITTLE_CPU2 6
> >> #define MT8195_MCU_LITTLE_CPU3 7
> >>
> >> +#define MT8195_AP_VPU0 8
> >
> > Can't this start from 0? This is a different hardware block. The index
> > namespace is separate. Same question for MT8192.
>
> The ID is used to differentiate the thermal zone identifier in the
> device tree from the driver.
>
> + vpu0-thermal {
> + polling-delay = <0>;
> + polling-delay-passive = <0>;
> + thermal-sensors = <&lvts_ap MT8195_AP_VPU0>;
> +
> + trips {
> + vpu0_crit: trip-crit {
> + temperature = <100000>;
> + hysteresis = <2000>;
> + type = "critical";
> + };
> + };
> + };
>
> If MT8195_AP_VPU0 is 0, then the code won't be able to differentiate
> MT8195_AP_VPU0 and MT8195_MCU_BIG_CPU0
>
> The LVTS driver will call devm_thermal_of_zone_register() with the
> sensor id. If MT8195_MCU_BIG_CPU0 and MT8195_AP_VPU0 have the same id,
> then at the moment of registering the MT8195_AP_VPU0, the underlying OF
> thermal framework code will use MT8195_MCU_BIG_CPU0 description instead
> because it will be the first to be find in the DT.
>
> If MT8195_AP_VPU0 is described in DT before, then the same will happen
> when registering MT8195_MCU_BIG_CPU0, MT8195_AP_VPU0 will be registered
> instead.
>
> IOW all ids must be different.
I see. I didn't realize the lookup namespace covered the whole platform.
In that case, please ignore my request.
ChenYu
> The namespace is already described by the macro name AFAICS, so whatever
> the values, we see only the macro names and those IDs are private the
> kernel implementation.
>
> If the numbering is really important, may be something like:
>
> #define MT8195_MCU_BIG_CPU0 00
> #define MT8195_MCU_BIG_CPU1 01
> #define MT8195_MCU_BIG_CPU2 02
> #define MT8195_MCU_BIG_CPU3 03
> #define MT8195_MCU_LITTLE_CPU0 04
> #define MT8195_MCU_LITTLE_CPU1 05
> #define MT8195_MCU_LITTLE_CPU2 06
> #define MT8195_MCU_LITTLE_CPU3 07
>
> #define MT8195_AP_VPU1 10
> #define MT8195_AP_GPU0 11
> #define MT8195_AP_GPU1 12
> #define MT8195_AP_VDEC 13
> #define MT8195_AP_IMG 14
> #define MT8195_AP_INFRA 15
> #define MT8195_AP_CAM0 16
> #define MT8195_AP_CAM1 17
>
> But I would suggest considering this change as a separate patch after
> the AP domain is added.
>
>
> --
> <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs
>
> Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook |
> <http://twitter.com/#!/linaroorg> Twitter |
> <http://www.linaro.org/linaro-blog/> Blog
>
