Re: [PATCH 4/5] ARM: dts: lpc3250-phy3250: add at25 required properties

[Vladimir Zapolskiy <vz@mleia.com>]

Hi Frank.

On 11/14/25 18:14, Frank Li wrote:
> On Fri, Nov 14, 2025 at 01:47:48AM +0200, Vladimir Zapolskiy wrote:
>> Hi Frank.
>>
>> On 10/29/25 22:28, Frank Li wrote:
>>> Add at25 required properties (size, address-width and pagesize), which
>>> duplicate deprecated properties.
>>>
>>> Change nodename 'at25' to 'eeprom'.
>>>
>>> Fix below CHECK_DTB warning:
>>>     arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dtb: at25@0 (atmel,at25): 'pagesize' is a required property
>>>     arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dtb: at25@0 (atmel,at25): $nodename: 'anyOf' conditional failed, one must be fixed:
>>>
>>> Signed-off-by: Frank Li <Frank.Li@nxp.com>
>>> ---
>>>    arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dts | 6 +++++-
>>>    1 file changed, 5 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dts b/arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dts
>>> index 21a6d0bca1e8a..15463a96938cc 100644
>>> --- a/arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dts
>>> +++ b/arch/arm/boot/dts/nxp/lpc/lpc3250-phy3250.dts
>>> @@ -200,7 +200,7 @@ &ssp0 {
>>>    	cs-gpios = <&gpio 3 5 0>;
>>>    	status = "okay";
>>> -	eeprom: at25@0 {
>>> +	eeprom: eeprom@0 {
>>>    		compatible = "atmel,at25";
>>>    		reg = <0>;
>>>    		spi-max-frequency = <5000000>;
>>> @@ -216,6 +216,10 @@ eeprom: at25@0 {
>>>    		at25,byte-len = <0x8000>;
>>>    		at25,addr-mode = <2>;
>>>    		at25,page-size = <64>;
>>> +
>>> +		size = <0x8000>;
>>> +		address-width = <16>;
>>> +		pagesize = <64>;
>>
>> Is there any need to have both sets of properties?
>>
> 
> Generally, keep old one just for back compatible. It is quite common for
> this case.
> 
> I am NOT sure if it is safe enough to remove deprecated properties now.
> 

It should be safe to remove the deprecated properties, it is fine.

This way the backward compatibility is preserved, and while the forward
compatibility is not preserved, it is not supposed to be preserved anyway.

>> I'd suggest to remove the deprecated set in a separate patch, note
>> that your done change is not described in the commit message.
> 
> Which part is not described?
> 

Sorry for it, I reread the commit message, and both changes are mentined,
but please split heterogeneous changes into separate patches.

-- 
Best wishes,
Vladimir