Re: Re: realtek PHY commit bbc4d71d63549 causes regression

On Sat, 14 Nov 2020 at 01:40, Andrew Lunn [off-list ref] wrote:

quoted

One question that still has not been answered is how many actual
platforms were fixed by backporting Realtek's follow up fix to
-stable. My suspicion is none. That by itself should be enough
justification to revert the backport of that change.

I think i've already said that would be a good idea. It makes the
problem less critical. But the problem is still there, we are just
kicking the can down the road. I've not seen much activity actually
fixing the broken DT. So i suspect when we catch up with the can, we
will mostly still be in the same place. Actually, maybe worse, because
broken DTs have been copy/pasted for new boards?

I don't see how that matters. If the new board ships with a stable
kernel, things should simply work as they did before. If the new board
ships with a new kernel, things won't work in the first place, so it
is unlikely to cause a regression in the field.