Hi,

On Sat, Feb 03, 2018 at 03:23:28PM +0800, Icenowy Zheng wrote:
> 于 2018年2月3日 GMT+08:00 上午6:13:01, Maxime Ripard <maxime.ripard@bootlin.com> 写到:
> >On Sat, Feb 03, 2018 at 02:04:54AM +0800, Icenowy Zheng wrote:
> >> The V3s is just a differently packaged version of the V3 chip, which
> >has
> >> a MAC with the same capability with H3. The V3s just doesn't wire out
> >> the external MII/RMII/RGMII bus. (V3 wired out it).
> >> 
> >> Drop the compatible string of V3s in the dwmac-sun8i driver, and add
> >a
> >> V3 compatible string, which has all capabilities.
> >> 
> >> Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
> >
> >This breaks the DT ABI, so NAK.
> 
> I have asked this at IRC.

One more reason why no one should ask questions like this on IRC.

> The V3s compatible string is never used in any mainline
> kernel, even not in any RC version.

$ git grep allwinner,sun8i-v3s-emac v4.15 | wc -l 
5

It is there already, and the fact that we have or don't have a DT in
tree that use it doesn't matter. One could very well have written a DT
for it and never submitted it.

Maxime

-- 
Maxime Ripard, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com