Re: [PATCH 2/3] dt-bindings: net: Add support for J-Core EMAC

[Rob Herring]

On Mon, Aug 18, 2025 at 10:55:51PM +0900, D. Jeff Dionne wrote:
> Something like:

Please don't top post to maillists.

> J-Core SoCs are assembled with an SoC generator tool from standard 
> components.  An SoC has a ROM from soc_gen with a Device Tree binary 
> included.  Therefore, J-Core SoC devices are designed to ‘just work’ 
> with linux, but this means the DT entires are generic, slightly 
> different than standard device tree practice.

Yes. Though doesn't the SoC generator evolve/change? New features in the 
IP blocks, bug fixes, etc. Soft IP for FPGAs is similar I think. There 
we typically just require the versioning schema be documented and 
correlate to the IP versions (vs. made up v1, v2, v3).

This is all pretty niche I think, so I'm not too concerned about what 
you do here.

Rob

> 
> J
> 
> > On Aug 18, 2025, at 22:41, Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote:
> > 
> > On 18/08/2025 12:57, Geert Uytterhoeven wrote:
> >>>> 
> >>>> No.  It’s a generic IP core for multiple SoCs, which do have names.
> >>> 
> >>> Then you need other SoCs compatibles, because we do not allow generic
> >>> items. See writing bindings.
> >>> 
> >>>> This is the correct naming scheme.  All compatible devices and SoCs match properly.
> >>> 
> >>> No, it is not a correct naming scheme. Please read writing bindings.
> >> 
> >> Can we please relax this for this specific compatible value?
> > 
> > We can...
> > 
> >> All other devices in this specific hardware implementation were
> >> accepted without SoC-specific compatible values ca. 9 years ago. AFAIK
> >> the Ethernet MAC was the sole missing piece, because its Linux driver
> >> was never attempted to be upstreamed before.
> > 
> > ...just provide some context and rationale in the commit msg.
> > 
> > Some (different) people pick up some irrelevant commits and use them as
> > argument in different discussions in style: it was allowed there, so I
> > can do the same.
> > 
> > Best regards,
> > Krzysztof
>