[PATCH 2/2] arm64: dts: rockchip: Make eeprom read-only for Radxa ROCK 5A

FUKAUMI Naoki naoki at radxa.com
Mon Nov 10 00:03:04 PST 2025


Hi Dragan,

On 11/10/25 16:38, Dragan Simic wrote:
> On Monday, November 10, 2025 08:05 CET, FUKAUMI Naoki <naoki at radxa.com> wrote:
>> On 11/10/25 14:17, Dragan Simic wrote:
>>> On Monday, November 10, 2025 04:54 CET, FUKAUMI Naoki <naoki at radxa.com> wrote:
>>>> Make EEPROM read-only as it may contain factory-programmed
>>>> board-specific data.
>>>>
>>>> Fixes: 89c880808cff8 ("arm64: dts: rockchip: add I2C EEPROM to rock-5a")
>>>> Signed-off-by: FUKAUMI Naoki <naoki at radxa.com>
>>>> ---
>>>>    arch/arm64/boot/dts/rockchip/rk3588s-rock-5a.dts | 1 +
>>>>    1 file changed, 1 insertion(+)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/rockchip/rk3588s-rock-5a.dts b/arch/arm64/boot/dts/rockchip/rk3588s-rock-5a.dts
>>>> index 19a08f7794e67..ae7b03488c9e7 100644
>>>> --- a/arch/arm64/boot/dts/rockchip/rk3588s-rock-5a.dts
>>>> +++ b/arch/arm64/boot/dts/rockchip/rk3588s-rock-5a.dts
>>>> @@ -254,6 +254,7 @@ eeprom: eeprom at 50 {
>>>>    		compatible = "belling,bl24c16a", "atmel,24c16";
>>>>    		reg = <0x50>;
>>>>    		pagesize = <16>;
>>>> +		read-only;
>>>>    	};
>>>>    };
>>>
>>> I've checked the board schematic and making this BL24C16 chip
>>> read-only makes sense to me.  However, I think that the patch
>>> description should be expanded to contain a better description
>>> of what might be found in the chip, such as a factory-programmed
>>> MAC address.  Also, providing a link to the board schematic, as
>>> a reference, would be a good idea.
>>
>> ----
>> arm64: dts: rockchip: Make eeprom read-only for Radxa ROCK 3C/5A/5C
>>
>> Make the EEPROM[1][2][3] read-only. Board-specific data written to the
>> EEPROM at the factory includes, but is not limited to, the SKU (~
>> product name), BOM version (~ board version), and MAC address.
> 
> Thanks for the improvements!  I'd just suggest that the opening
> sentence above is rewritten like this:
> 
>    Make the BL24C16 EEPROM chips found on Radxa ROCK 3C, 5A and
>    5C SBCs [1][2][3] read-only, because they contain factory-
>    programmed data that isn't supposed to be modified by the end
>    users.
> 
> This would also be perfect timing for Radxa to officially release
> the layout of the board-specific data that goes into these EPROM
> chips, so the data can actually be used.

  https://github.com/radxa/u-boot/blob/next-dev-v2024.10/drivers/misc/radxa-i2c-eeprom.c

The actual data in the EEPROMs of my boards has some content 
inconsistencies with the code descriptions, but the format is 
consistent. The SKU, BOM, and MAC mentioned previously should be available.

Best regards,

--
FUKAUMI Naoki
Radxa Computer (Shenzhen) Co., Ltd.

>> [1]
>> https://dl.radxa.com/rock3/docs/hw/3c/v1400/radxa_rock_3c_v1400_schematic.pdf
>> p.13
>> [2] https://dl.radxa.com/rock5/5a/docs/hw/radxa_rock5a_V1.1_sch.pdf p.19
>> [3]
>> https://dl.radxa.com/rock5/5c/docs/hw/v1100/radxa_rock_5c_schematic_v1100.pdf
>> p.18
>> ----
>>
>>> With these remarks addressed in a v2, please feel free to include
>>>
>>> Reviewed-by: Dragan Simic <dsimic at manjaro.org>
>>>
>>> Additionally, it would make sense to squash this patch and your
>>> other patch that addresses the ROCK 5C together.  They're very
>>> similar and having them together in a single commit might actually
>>> help the people browsing the repository understand it better.
>>
>> I'll drop eeprom patch from ROCK 5C v7 patch series, and squash eeprom
>> patches for ROCK 3C/5A/5C.
> 
> Thanks!
> 
> 





More information about the Linux-rockchip mailing list