[PATCH v2 5/5] pmdomain: add Rockchip power domain support

[Ahmad Fatoum]

Hi Sascha,

On 10/28/25 11:46 AM, Sascha Hauer wrote:
> On Tue, Oct 28, 2025 at 11:41:47AM +0100, Ahmad Fatoum wrote:
>> Hi,
>>
>> On 10/28/25 10:48 AM, Sascha Hauer wrote:
>>> On Tue, Oct 28, 2025 at 10:04:06AM +0100, Ahmad Fatoum wrote:
>>>>
>>>>
>>>> On 10/28/25 9:57 AM, Sascha Hauer wrote:
>>>>> On Mon, Oct 27, 2025 at 08:28:20PM +0100, Ahmad Fatoum wrote:
>>>>>> +static int rockchip_pm_add_subdomain(struct rockchip_pmu *pmu,
>>>>>> +				     struct device_node *parent)
>>>>>> +{
>>>>>> +	struct device_node *np;
>>>>>> +
>>>>>> +	for_each_child_of_node(parent, np)
>>>>>> +		return -ENOSYS;
>>>>>> +
>>>>>
>>>>> I gave it a quick try on RK3588 and driver probe fails here. RK3588 has
>>>>> several subdomains in its power controller.
>>>>
>>>> I need it on rk3566 though, otherwise the probe hangs. Should I drop the
>>>> rk358x compatibles?
>>>
>>> That leaves us with the warnings introduced with the last patch:
>>>
>>> WARNING: dwc3 fc000000.usb at fc000000.of: power domain not found (ignoring)
>>> WARNING: rockchip-dw-pcie a40c00000.pcie at fe180000.of: power domain not found (ignoring)
>>> WARNING: rockchip-dw-pcie a41000000.pcie at fe190000.of: power domain not found (ignoring)
>>> WARNING: dw_mmc fe2c0000.mmc at fe2c0000.of: power domain not found (ignoring)
>>> WARNING: dwc3 fc400000.usb at fc400000.of: power domain not found (ignoring)
>>> WARNING: rockchip-dw-pcie a40400000.pcie at fe160000.of: power domain not found (ignoring)
>>
>> Is PCIe working right now?
> 
> Not this particular port which still has a missing phy, but generally
> yes. The Rock5t has ethernet connected to PCIe.
> 
>>
>>> Maybe add a dummy driver where people can add their compatible to when
>>> they encounter the warning knowing that their SoC doesn't need a power
>>> domain driver?
>>
>> Are you sure the power domains are indeed unneeded?
> 
> No, I am not, but currently it seems all power domains we need are
> turned on.
> 
> If it's more convenient for you you could of course still add power
> domain support for RK3588 ;)

Not intending to anytime soon, sorry.

> Having a dummy driver might turn out useful for other SoCs, I am not
> sure how many SoCs we have which lack a power domain driver, but have
> power domains in specified in the device tree.

Yes, absolutely. I'd prefer people actually familiar with the hardware
adding matches to the dummy driver though...

I will do this for v3. Can 1, 2 and 4 already go to next?

Thanks,
Ahmad

> 
> Sascha
> 

-- 
Pengutronix e.K.                  |                             |
Steuerwalder Str. 21              | http://www.pengutronix.de/  |
31137 Hildesheim, Germany         | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686  | Fax:   +49-5121-206917-5555 |