[PATCH v2 2/2] dt-bindings: arm: List more IXP4xx devices

Rob Herring robh at kernel.org
Tue Sep 26 06:13:02 PDT 2023


On Tue, Sep 26, 2023 at 11:20:01AM +0100, Conor Dooley wrote:
> On Tue, Sep 26, 2023 at 11:04:26AM +0200, Linus Walleij wrote:
> > On Tue, Sep 26, 2023 at 10:26 AM Conor Dooley <conor at kernel.org> wrote:
> > 
> > > > +      - items:
> > > > +          - const: linksys,wrv54g
> > > > +          - const: gemtek,gtwx5715
> > > > +          - const: intel,ixp42x
> > >
> > > One question, not so much about this particular case, but is there no
> > > "standalone" version of the gtwx5715 sold by gemtek? Asking as I don't
> > > see it in the enum above. The description sounds like it is both a
> > > product in its own right & sold rebadged.
> > 
> > Yeah this one is a special headache. It's two different brandings
> > of the same router. The Linksys version can still be bought new
> > from Amazon:
> > https://www.amazon.com/Cisco-Linksys-WRV54G-Wireless-G-VPN-Router/dp/B0000AR8Z1
> > 
> > I'm leaning toward listing them as two different devices instead
> > of what we have now, which is
> > 
> >   compatible = "linksys,wrv54g", "gemtek,gtwx5715", "intel,ixp42x";
> > 
> > We can just decide that one of them is the canon device, I guess
> > Linksys, and the other we can just use that device tree, or we
> > create a device tree that includes the former and just override
> > the compatible.
> > 
> > I don't really know what to do here.
> 
> Ahh, you were intentionally doing this so that for both devices you
> would list all 3 compatibles? I had it in my head that the OEM device
> would have 2 compatibles & the rebadged one would have 3.
> I'd probably be slightly in favour of your second option, since you'd
> likely want to set the model too. I suppose it's the decision of the
> maintainer for the platforms, which IIRC is you, as given the devices
> differ only by the badge...

I would do a combination. Use the same compatible, but different "model" 
if you want to override it.

Rob



More information about the linux-arm-kernel mailing list