[PATCH] arm64: dts: ti: k3-am62a7-sk: Add bootph-all property in cpsw_mac_syscon node
Andrew Davis
afd at ti.com
Tue Jul 14 07:16:59 PDT 2026
On 6/30/26 1:54 PM, Chintan Vankar wrote:
> Hello Andrew,
>
> On 26/06/26 02:18, Andrew Davis wrote:
>> On 6/25/26 6:32 AM, Chintan Vankar wrote:
>>> Ethernet boot requires CPSW node to be present starting from R5 SPL stage.
>>> Add "bootph-all" property in CPSW MAC's eFuse node "cpsw_mac_syscon" to
>>> enable this node during SPL stage along with later boot stage so that CPSW
>>> port will get static MAC address.
>>>
>>> Signed-off-by: Chintan Vankar <c-vankar at ti.com>
>>> ---
>>>
>>> Hello All,
>>>
>>> This patch is based on linux-next tagged next-20260623.
>>>
>>> arch/arm64/boot/dts/ti/k3-am62a7-sk.dts | 4 ++++
>>> 1 file changed, 4 insertions(+)
>>>
>>> diff --git a/arch/arm64/boot/dts/ti/k3-am62a7-sk.dts b/arch/arm64/ boot/dts/ti/k3-am62a7-sk.dts
>>> index 821a9705bb7d..d3b3675e7a8f 100644
>>> --- a/arch/arm64/boot/dts/ti/k3-am62a7-sk.dts
>>> +++ b/arch/arm64/boot/dts/ti/k3-am62a7-sk.dts
>>> @@ -230,6 +230,10 @@ AM62AX_MCU_IOPAD(0x0030, PIN_OUTPUT, 0) /* (C8) WKUP_UART0_RTSn */
>>> };
>>> };
>>> +&cpsw_mac_syscon {
>>> + bootph-all;
>>
>> Seems you need this because cpsw_port1 uses it though a phandle reference.
>> cpsw_port1 has bootph-all, why is this property not transitive though
>> phandles? Would not having that cause missing references when the phandles
>> are resolved to nodes that get dropped for some given boot stage?
>>
>
> Yes, the bootph-all property is not automatically transitive through phandle references in the U-Boot SPL DT. Nodes that are only referenced by phandle from a bootph-annotated node are not themselves retained
> unless they also carry a bootph-* property. This is because the way
> fdtgrep works[1], it only keeps node with the tags present and implies
> that property to the parent nodes and not the nodes referenced by
> "phandle".
>
Okay then, sounds like something worth investigating to see if this
can be handled automatically in the future, for now given the current
state of fdtgrep scripting, this patch LGTM,
Acked-by: Andrew Davis <afd at ti.com>
> Without bootph-all in cpsw_mac_syscon, the SPL device tree will drop
> that node, leaving the phandle in cpsw_port1 unresolved. And the above
> claim can be validated with the current conifguration where "bootph-all"
> tag is not present in cpsw_mac_syscon, causing CPSW to fail retrieve MAC
> address.
>
> [1]: https://github.com/u-boot/u-boot/blob/master/scripts/Makefile.lib#L688
>
> Regards,
> Chintan.
>
>> Andrew
>>
>>> +};
>>> +
>>> /* WKUP UART0 is used for DM firmware logs */
>>> &wkup_uart0 {
>>> pinctrl-names = "default";
>>
>
More information about the linux-arm-kernel
mailing list