Re: [PATCH net-next v5 09/13] net: phylink: Use phy_caps_lookup for fixed-link configuration

[Maxime Chevallier <maxime.chevallier@bootlin.com>]

> > Basically we still need the value to be screened by the pl->supported.
> > The one change is that we have to run the extra screening on the
> > intersect instead of skipping the screening, or doing it before we even
> > start providing bits.
> > 
> > With this approach we will even allow people to use non twisted pair
> > setups regardless of speed as long as they don't provide any twisted
> > pair modes in the standard set.
> > 
> > I will try to get this tested today and if it works out I will submit
> > it for net. I just need to test this and an SFP ksettings_set issue I
> > found when we aren't using autoneg.  
> 
> This code used to be so simple... and that makes me wonder whether
> Maxime's work is really the best approach. It seems that the old way
> was better precisely because it was more simple.

Sorry to hear you say that. Fixed-link was the main pain point with
this work, I've stressed it out. I agree that for fixed-link, it
ends-up not looking too good, hopefully the rest of the series
compensate for that.

Maxime