Re: [PATCH 0/6] mvneta: SGMII-based in-band link state signaling

27.03.2015 16:39, Andrew Lunn пишет:

On Fri, Mar 27, 2015 at 04:28:28PM +0300, Stas Sergeev wrote:

quoted

Hello.

Currently the fixed-link DT binding is pre-configured and
cannot be changed in run-time. This means the cable unplug
events are not being detected, and the link parameters can't
be negotiated.

O.K, i will ask the dumb question....

Isn't fixed-link supposed to be used for links which are fixed?
Why would a fixed-link change?

Hi, my current understanding of fixed-link DT binding is that
it is actually just an MDIO-less link, but not more fixed than that.
I made a patch out of that assumption.
Why I think that way is because fixed-link internally uses a fixed_phy
driver - the one that actually allows a state updates and emulates MDIO
on top of these. So I just exploited that already coded capability.
I think any other approach will require much more work.

The use cases i've seen for this is when you are connecting the MAC to
a Switch. The link is configured to its fastest mode and then left
alone.

Please could you give some more background information. What do you
have on the other end of this fixed link which keeps changing?
Obviously not a switch.

Normal PHY, not a switch.
But there is no MDIO, because SGMII AFAIK doesn't need MDIO.
SGMII has in-band status, but for some reason it seems currently
linux is not ready for such setup - this is what my patch addresses.
Or maybe I am missing something?