I think that

  ethtool -s <iface> duplex half autoneg on

should be enough, the link should still establish at 100M, I've tested
that on a 1G/FULL 100MHalf+Full interface and this is the result we
get :)

Nice.

But i still think the test should check the autoneg result and do
something sensible if the link does not come up. This probably applies
to all cases where we trigger auto neg.

That said I've tested the following on mcbin, and it seems that acually
nothing in the code currently deals with Half duplex / Pause interaction,
and we don't get any EOPNOTSUPP.

So the broader question is, should we reflect the current behaviour or
an ideal one ?

What 802.3 says. If we come across cases where phylib/phylink is
broken, let me know, and i will fix it. But we will leave driver bugs
to individual driver developers.

But we also need to consider that for some APIs, we have decided that
a configuration can be set now, which does not actually apply in our
current conditions, but it will be stored away for when conditions
change and it is applicable. The half duplex case could fit that. When
the link is currently half duplex, you can configure pause, but you
don't expect it to actually change the current behaviour. It only
kicks in when the link renegotiates to full duplex sometime in the
future. We have to also consider this the other way around. The link
is full duplex and pause is configured by the user. Something happens
with the LP and the link renegotiates to half duplex. The local end
should not throw away the configuration, it simply cannot apply it
given the current situation.

	Andrew