On Mon, Nov 04, 2024 at 11:18:48AM +0100, AngeloGioacchino Del Regno wrote:
> Il 01/11/24 16:20, Nícolas F. R. A. Prado ha scritto:
> > This series fixes the inverted handling of the mediatek,mac-wol DT
> > property while keeping backward compatibility. It does so by introducing
> > a new property on patch 1 and updating the driver to handle it on patch
> > 2. Patch 3 adds this property on the Genio 700 EVK DT, where this issue
> > was noticed, to get WOL working on that platform. Patch 4 adds the new
> > property on all DTs with the MediaTek DWMAC ethernet node enabled
> > and inverts the presence of mediatek,mac-wol to maintain the
> > current behavior and have it match the description in the binding.
> >
>
> Actually, I'm sure that all of these boards *do* need MAC WOL and *not* PHY WOL.
>
> The only one I'm unsure about is MT2712, but that's an evaluation board and not
> a retail product with "that kind of diffusion".
>
> I think you can just fix the bug in the driver without getting new properties
> and such. One commit, two lines.
Alright, since you're sure all the boards need MAC WOL this is simpler: I can
just fix the driver handling, and also update all the DTs to make sure they use
MAC WOL. (Right now some are enabling PHY WOL, mt8395-genio-1200-evk.dts and
mt8395-radxa-nio-12l.dts, while the others are enabling the MAC WOL)
The MT2712 EVB currently enables MAC WOL, so even though you're not sure about
that one, I think it makes sense to keep the current behavior ie add the
mediatek,mac-wol property just like for the others, so it keeps using MAC WOL.
I'll wait a couple days in case there are more comments and then I'll send v2
with those changes.
Thanks,
Nícolas