Re: [PATCH] arm64: dts: qcom: ipq5210: add the bootph-all property

[Krzysztof Kozlowski <krzk@kernel.org>]

On 10/04/2026 10:41, Kathiravan Thirumoorthy wrote:
> 
> On 4/9/2026 3:33 PM, Krzysztof Kozlowski wrote:
>> On 09/04/2026 11:28, Kathiravan Thirumoorthy wrote:
>>> Add the bootph-all property to the nodes which are utilized by the
>>> bootloaders.
>> Uh oh, so it started for qcom too? I really don't like how these bootph
>> properties spread all over, so please provide arguments - which pure
>> upstream bootloaders exactly and why exactly these nodes.
> 
> This is needed for U-Boot SPL for IPQ5210 [1].

Mention U-Boot SPL in commit msg and add Link for commit msg (or
changelog if other prefers that).

> 
> [1] - 
> https://lore.kernel.org/u-boot/20260408091136.2794546-1-varadarajan.narayanan@oss.qualcomm.com/

> 
>>> Signed-off-by: Kathiravan Thirumoorthy <kathiravan.thirumoorthy@oss.qualcomm.com>
>>> ---
>>>   arch/arm64/boot/dts/qcom/ipq5210-rdp504.dts |  5 +++++
>>>   arch/arm64/boot/dts/qcom/ipq5210.dtsi       | 10 ++++++++++
>>>   2 files changed, 15 insertions(+)
>>>
>>> diff --git a/arch/arm64/boot/dts/qcom/ipq5210-rdp504.dts b/arch/arm64/boot/dts/qcom/ipq5210-rdp504.dts
>>> index 941f866ecfe9..56dbc506da78 100644
>>> --- a/arch/arm64/boot/dts/qcom/ipq5210-rdp504.dts
>>> +++ b/arch/arm64/boot/dts/qcom/ipq5210-rdp504.dts
>>> @@ -17,6 +17,7 @@ aliases {
>>>   
>>>   	chosen {
>>>   		stdout-path = "serial0";
>>> +		bootph-all;
>> Since when do we add bootph-all to chosen? Which broken bootloader
>> ignores chosen?
>>
>> This really makes me wonder that you do all this for some downstream forks.
> 
> U-Boot doesn't need it, but U-Boot SPL needs it.

I understand, but let me rephrase my question better. Which bootloader
build process ignores chosen when preparing final DTB for the target?
Maybe that build process needs to be fixed.

> 
>>>   	};
>>>   };
>>>   
>>> @@ -41,6 +42,7 @@ qup_uart1_default_state: qup-uart1-default-state {
>>>   		function = "qup_se1";
>>>   		drive-strength = <6>;
>>>   		bias-pull-down;
>>> +		bootph-all;
>>>   	};
>> And that's a pin, not a device. What is the point of marking it? The
>> device needing this pin will have phandle which must pull the node.
> 
> This is because of the fdtgrep tool's limitations, which removes these 
> nodes in the final DTB if this property is not present.

fdtgrep should be fixed. I see no point in marking obvious dependencies.

Best regards,
Krzysztof