Re: [PATCH 2/2] arm64: dts: socfpga: agilex7-gen2: Add initial device tree
From: Krzysztof Kozlowski
Date: Wed Jun 24 2026 - 06:28:08 EST
On 24/06/2026 12:17, Nazle Asmade, Muhammad Nazim Amirul wrote:
>> SoC without any interface, serial or storage or network, is close to
>> useless one.
>>
>> I don't see a point in having it in mainline. Serial is usually ABSOLUTE
>> minimum.
>>
>> Best regards,
>> Krzysztof
>>
> Hi Krzysztof,
>
> Thank you for the review and fast response!
>
> I ran both dt_binding_check and dtbs_check (with CHECK_DTBS=y) locally —
> both passed without errors. Could you clarify which specific test you
> believe is failing?
I would expect simple-bus schema warning or W=1, because node is placed
outside of soc@, but maybe there is no such.
>
> Regarding "MMIO goes to MMIO" — are you referring to the GIC
Comments are placed in very specific and intentional place. Please read
guides how mailing list in-line review works before posting patches.
> (interrupt-controller@7000000) being placed at the root level instead of
> under the soc bus node?
>
> Regarding the serial console — the platform clock driver is not yet
> upstream, so the UART depends on clkmgr. Would adding the UART with
> clock-frequency be acceptable as an interim solution?
Add complete working serial. Why can't you use fixed placeholder clock?
There are probably multiple ways to solve it, not necessary
clock-frequency and I do not even remember if clock-frequency is allowed.
But if you cannot bring serial, then my comment stays valid: this is
unusable upstream thus is not ready to be posted and merged.
Best regards,
Krzysztof