On Thu, 27 Oct 2022 19:50:35 -0700 Florian Fainelli wrote:
> > @@ -723,6 +724,8 @@ struct phy_device {
> >
> > int pma_extable;
> >
> > + unsigned int link_down_events;
>
> Should not this be an u64 to match what the extended link state can
> report? Not that I would hope that anyone had a chance to witness 4
> billion link down events using PHYLIB.
How about I also make the uAPI side 32 bit? I made it 64b because
that feels like the default these days but as you say, re-training
a link 4 billion times is unlikely to happen in one person's lifetime :S
