Re: [PATCH 2/4] arm64: dts: rockchip: rk3399-pinephone-pro: DW9714 coil requires higher voltage

[David Heidelberg <david@ixit.cz>]

On 06/08/2026 13:19, Ondřej Jirman wrote:
> Hi,
> 
> On Tue, Aug 04, 2026 at 11:12:21AM +0200, David Heidelberg wrote:
>> On 04/08/2026 09:39, Ondřej Jirman wrote:
>>> Hi,
>>>
>>> On Sat, Aug 01, 2026 at 02:35:21AM +0200, David Heidelberg via B4 Relay wrote:
>>>> From: David Heidelberg <david@ixit.cz>
>>>>
>>>> By datasheet the VDD (in-kernel named vcc) range is between 2.3 to 4.3 V.
>>>> Usually it's around 2.8 V. Here it matches the avdd2v8_dvp.
>>>>
>>>> Fixes: 9336eb829a18 ("arm64: dts: rockchip: describe I2c Bus 1 and IMX258 world camera on PinePhone Pro")
>>>> Cc: stable@vger.kernel.org
>>>> Signed-off-by: David Heidelberg <david@ixit.cz>
>>>> ---
>>>>    arch/arm64/boot/dts/rockchip/rk3399-pinephone-pro.dts | 2 +-
>>>>    1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/rockchip/rk3399-pinephone-pro.dts b/arch/arm64/boot/dts/rockchip/rk3399-pinephone-pro.dts
>>>> index d46cdfe3f7849..bde35d9b2263e 100644
>>>> --- a/arch/arm64/boot/dts/rockchip/rk3399-pinephone-pro.dts
>>>> +++ b/arch/arm64/boot/dts/rockchip/rk3399-pinephone-pro.dts
>>>> @@ -473,17 +473,17 @@ wcam_out: endpoint {
>>>>    			};
>>>>    		};
>>>>    	};
>>>>    	wcam_lens: camera-lens@c {
>>>>    		compatible = "dongwoon,dw9714";
>>>>    		reg = <0x0c>;
>>>>    		/* Same I2c bus as both cameras, depends on vcca1v8_codec for power. */
>>>> -		vcc-supply = <&vcc1v8_dvp>;
>>>> +		vcc-supply = <&avdd2v8_dvp>;
>>>
>>> This does not match reality, schematic says it's VCC2V8_DVP:
>>>
>>> 	https://xff.cz/dl/tmp/6253a0465329df41.png
>>
>> Ahoj,
>>
>> thanks for spotting this, since I know you have the device, would be too
>> much to ask you to send standalone fix for PP Pro?
>>
>> I would drop the change from the series.
> 
> Ok. Just for the reference, vcc1v8_dvp here feeds an enable pin for VCC2V8_DVP
> LDO regulator (which is supplied from VCC3V3_SYS and not described in DT, yet).
> So in practice enabling vcc1v8_dvp enables both VCC2V8_DVP and AVDD2V8_DVP and
> while DT does not describe the HW in detail, it describes the only regulator
> that is actually controllable and enables all needed power rails for the camera
> modules all by itself in proper order/timing.
> 
> Not sure what's the DT maintainers idea here about whether DT has to describe
> every detail of power rails routing, or just what's pragmatic.

My perception is that DT syntax should describe everything, except passive 
components.

No additional GPIO is needed to enable VCC and AVDD?

Here, it seems to me that explanation like this in DT could cover it. But I'm no 
authority here :)

David

> 
> Best regards,
> 	o.j.
> 
>> David
>>
>> P.S. I noticed I didn't mark this series RFC. Too late.
>>
>>>
>>> regards,
>>> 	o.
>>>
>>>>    	};
>>>>    	ucam: camera@36 {
>>>>    		compatible = "ovti,ov8858";
>>>>    		reg = <0x36>;
>>>>    		clocks = <&cru SCLK_CIF_OUT>; /* MIPI_MCLK1, derived from CIF_CLK0 */
>>>>    		clock-names = "xvclk";
>>>>    		dovdd-supply = <&vcc1v8_dvp>;
>>>>
>>>> -- 
>>>> 2.53.0
>>>>
>>>>
>>
>> -- 
>> David Heidelberg
>>

-- 
David Heidelberg