Re: [PATCH v4 1/3] dt-bindings: net: adin: document phy clock

[Jakub Kicinski <kuba@kernel.org>]

On Sun, 15 May 2022 10:16:47 +0300 Josua Mayer wrote:
> Am 13.05.22 um 01:44 schrieb Jakub Kicinski:
> > On Thu, 12 May 2022 23:20:18 +0200 Michael Walle wrote:  
> >>> It's pure speculation on my side. I don't even know if PHYs use
> >>> the recovered clock to clock its output towards the MAC or that's
> >>> a different clock domain.
> >>>
> >>> My concern is that people will start to use DT to configure SyncE which
> >>> is entirely a runtime-controllable thing, and doesn't belong.  
> Okay.
> However phy drivers do not seem to implement runtime control of those 
> clock output pins currently, so they are configured once in DT.

To me that means nobody needs the recovered clock.

> >>> Hence
> >>> my preference to hide the recovered vs free-running detail if we can
> >>> pick one that makes most sense for now.  
> I am not a fan of hiding information. The clock configuration register 
> clearly supports this distinction.

Unless you expose all registers as a direct API to the user you'll be
"hiding information". I don't think we are exposing all possible
registers for this PHY, the two bits in question are no different.

> Is this a political stance to say users may not "accidentally" enable 
> SyncE by patching DT?
> If so we should print a warning message when someone selects it?

Why would we add a feature and then print a warning? We can always add 
the support later, once we have a use case for it.

> >> I see. That makes sense, but then wouldn't it make more sense to pick
> >> the (simple) free-running one? As for SyncE you'd need the recovered
> >> clock.  
> > 
> > Sounds good.  
> 
> Yep, it seems recovered clock is only for SyncE - and only if there is a 
> master clock on the network. Sadly however documentation is sparse and I 
> do not know if the adi phys would fall back to using their internal 
> clock, or just refuse to operate at all.