Hi Quentin,
On 7/14/2025 6:35 PM, Quentin Schulz wrote:
> Hi Jonas,
>
> On 7/12/25 1:06 AM, Jonas Karlman wrote:
>> The Radxa ROCK 4D is a compact single-board computer (SBC) featuring
>> numerous top-tier functions, features, and expansion options.
>>
>> Equipped with the Rockchip RK3576 or RK3576J SoC, the ROCK 4D boasts an
>> octa-core CPU (4x Cortex-A72 + 4x Cortex-A53), Mali-G52 GPU, and a
>> powerful 6 TOPS NPU, making it ideal for AI and multimedia tasks.
>>
>> Features tested on a Radxa ROCK 4D v1.112:
>> - SPI Flash boot
>> - Ethernet
>>
>> ROCK 4D boards with SPI Flash is configured to boot from FSPI0->UFS->USB,
>> or directly from USB when the MASKROM button is pressed, booting
>> directly from SD-card is not possible on these boards.
>>
>> Signed-off-by: Jonas Karlman <jo...@kwiboo.se>
>> ---
>> arch/arm/dts/rk3576-rock-4d-u-boot.dtsi | 15 +++++
>> arch/arm/mach-rockchip/rk3576/MAINTAINERS | 6 ++
>> configs/rock-4d-rk3576_defconfig | 68 +++++++++++++++++++++++
>> doc/board/rockchip/rockchip.rst | 1 +
>> 4 files changed, 90 insertions(+)
>> create mode 100644 arch/arm/dts/rk3576-rock-4d-u-boot.dtsi
>> create mode 100644 configs/rock-4d-rk3576_defconfig
>>
>> diff --git a/arch/arm/dts/rk3576-rock-4d-u-boot.dtsi
>> b/arch/arm/dts/rk3576-rock-4d-u-boot.dtsi
>> new file mode 100644
>> index 000000000000..5056a1cecb83
>> --- /dev/null
>> +++ b/arch/arm/dts/rk3576-rock-4d-u-boot.dtsi
>> @@ -0,0 +1,15 @@
>> +// SPDX-License-Identifier: (GPL-2.0+ OR MIT)
>> +
>> +#include "rk3576-u-boot.dtsi"
>> +
>> +&rgmii_phy0 {
>> + /delete-property/ reset-gpio;
>> + reset-gpios = <&gpio2 RK_PB5 GPIO_ACTIVE_LOW>;
>
> Why?
>
> If the upstream DT is wrong, please fix the upstream DT first and
> cherry-pick.
This has already been fixed in upstream DT [1], however it is targeted
for v6.17 so not pickable for another 4-6 weeks.
This u-boot.dtsi override fixes it so Ethernet PHY reset can work in
U-Boot until the prop-rename fix has fully landed.
[1]
https://lore.kernel.org/all/20250704-rk3576-rock4d-phy-handling-fixes-v1-1-1d64130c4...@kernel.org/
>
> If it isn't wrong, then can't we simply patch the PHY subsystem/driver
> to be able to read reset-gpio too instead?
reset-gpios is the correct prop, so fixing DT is what should be done.
Regards,
Jonas
>
> Cheers,
> Quentin
