Re: [PATCH v3 01/12] dt-bindings: clock: renesas,r9a08g045-vbattb: Document VBATTB

[claudiu beznea <claudiu.beznea@tuxon.dev>]

On 03.09.2024 10:36, Biju Das wrote:
> Hi Claudiu,
> 
>> -----Original Message-----
>> From: claudiu beznea <claudiu.beznea@tuxon.dev>
>> Sent: Tuesday, September 3, 2024 8:28 AM
>> Subject: Re: [PATCH v3 01/12] dt-bindings: clock: renesas,r9a08g045-vbattb: Document VBATTB
>>
>>
>>
>> On 03.09.2024 09:58, Biju Das wrote:
>>> Hi Claudiu,
>>>
>>>> -----Original Message-----
>>>> From: Claudiu <claudiu.beznea@tuxon.dev>
>>>> Sent: Friday, August 30, 2024 2:02 PM
>>>> Subject: [PATCH v3 01/12] dt-bindings: clock:
>>>> renesas,r9a08g045-vbattb: Document VBATTB
>>>>
>>>> From: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
>>>>
>>>> The VBATTB IP of the Renesas RZ/G3S SoC controls the clock for RTC,
>>>> the tamper detector and a small general usage memory of 128B. Add documentation for it.
>>>>
>>>> Signed-off-by: Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
>>>> ---
>>>>
>>>> Changes in v3:
>>>> - moved the file to clock dt bindings directory as it is the
>>>>   only functionality supported at the moment; the other functionalities
>>>>   (tamper detector, SRAM) are offered though register spreaded
>>>>   though the address space of the VBATTB IP and not actually
>>>>   individual devices; the other functionalities are not
>>>>   planned to be supported soon and if they will be I think they
>>>>   fit better on auxiliary bus than MFD
>>>> - dropped interrupt names as requested in the review process
>>>> - dropped the inner node for clock controller
>>>> - added #clock-cells
>>>> - added rtx clock
>>>> - updated description for renesas,vbattb-load-nanofarads
>>>> - included dt-bindings/interrupt-controller/irq.h in examples section
>>>>
>>>> Changes in v2:
>>>> - changed file name and compatible
>>>> - updated title, description sections
>>>> - added clock controller part documentation and drop dedicated file
>>>>   for it included in v1
>>>> - used items to describe interrupts, interrupt-names, clocks, clock-names,
>>>>   resets
>>>> - dropped node labels and status
>>>> - updated clock-names for clock controller to cope with the new
>>>>   logic on detecting the necessity to setup bypass
>>>>
>>>>  .../clock/renesas,r9a08g045-vbattb.yaml       | 81 +++++++++++++++++++
>>>>  1 file changed, 81 insertions(+)
>>>>  create mode 100644
>>>> Documentation/devicetree/bindings/clock/renesas,r9a08g045-vbattb.yaml
>>>>
>>>> diff --git
>>>> a/Documentation/devicetree/bindings/clock/renesas,r9a08g045-vbattb.ya
>>>> ml
>>>> b/Documentation/devicetree/bindings/clock/renesas,r9a08g045-vbattb.ya
>>>> ml
>>>> new file mode 100644
>>>> index 000000000000..29df0e01fae5
>>>> --- /dev/null
>>>> +++ b/Documentation/devicetree/bindings/clock/renesas,r9a08g045-vbatt
>>>> +++ b.y
>>>> +++ aml
>>>> @@ -0,0 +1,81 @@
>>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) %YAML 1.2
>>>> +---
>>>> +$id:
>>>> +http://devicetree.org/schemas/clock/renesas,r9a08g045-vbattb.yaml#
>>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>>> +
>>>> +title: Renesas Battery Backup Function (VBATTB)
>>>> +
>>>> +description:
>>>> +  Renesas VBATTB is an always on powered module (backed by battery)
>>>> +which
>>>> +  controls the RTC clock (VBATTCLK), tamper detection logic and a
>>>> +small
>>>> +  general usage memory (128B).
>>>> +
>>>> +maintainers:
>>>> +  - Claudiu Beznea <claudiu.beznea.uj@bp.renesas.com>
>>>> +
>>>> +properties:
>>>> +  compatible:
>>>> +    const: renesas,r9a08g045-vbattb
>>>> +
>>>> +  reg:
>>>> +    maxItems: 1
>>>> +
>>>> +  interrupts:
>>>> +    items:
>>>> +      - description: tamper detector interrupt
>>>> +
>>>> +  clocks:
>>>> +    items:
>>>> +      - description: VBATTB module clock
>>>> +      - description: RTC input clock (crystal oscillator or external
>>>> + clock device)
>>>> +
>>>> +  clock-names:
>>>> +    items:
>>>> +      - const: bclk
>>>> +      - const: rtx
>>>> +
>>>> +  '#clock-cells':
>>>> +    const: 1
>>>> +
>>>> +  power-domains:
>>>> +    maxItems: 1
>>>
>>> Not sure, you need to document "PD_VBATT" power domain as per Table
>>> 41.2, this LSI supports 3 power domains(PD_ISOVCC, PD_VCC, PD_VBATT)
>>>
>>> Power Mode PD_ISOVCC PD_VCC PD_VBATT
>>> ALL_ON      ON          ON    ON
>>> AWO         OFF         ON    ON
>>> VBATT       OFF         OFF   ON
>>> ALL_OFF     OFF         OFF   OFF
>>>
>>> PD_VBATT domain is the area where the RTC/backup register is located,
>>> works on battery power when the power of PD_VCC and PD_ISOVCC domain are turned off.
>>
>> In Linux, the CPG is the power domain provider for all the IPs in RZ/G3S SoC (modeled though MSTOP CPG
>> support). This is how it is currently implemented.
>>
>> Then groups of IPs are part of power domains PD_ISOVCC, PD_VCC, PD_VBATT.
>> These power domains are i2c controlled with the help of firmware (at least at the moment).
>>
>> From HW manual:
>> - PD_VCC domain always powered on area.
>>
>> - PD_ISOVCC domain is the area where the power can be turned off.
>>
>> - PD_VBATT domain is the area where the RTC/backup register is located,
>>   works on battery power when the power of .
>>
>> The power to these domains are controlled with the help of firmware. Linux cannot do control itself as
>> the CPU is in the PD_ISOVCC. If you look at picture 41.3 Power mode transition [1] it is mentioned the
>> relation b/w these power domains (controlled by PMIC though firmware) and the supported power saving
>> modes: ALL_ON, AWO, VBATT.
>>
> 
> DT describes hardware. So, the question was, from that perspective, do we need to document PD_VBATT domain,
> as it can be controlled outside linux??

The control to these domains is passed to firmware.

From my point of view these will never be used by Linux becuase:
- the PD_ISOVCC is where the CPU resides and it cannot cut itself its power
- the PD_VCC is a domain where critical IPs like CPG, SYSC resides
- the VBATT is where the RTC resides, RTC that should stay on forever

Should we document something that will be never used?

> 
> Cheers,
> Biju