RE: [PATCH net v3 2/2] net: fec: free the PPS interrupt before tearing down the PHC and netdev

[Wei Fang]

> > > +
> > > +       if (fep->pps_irq >= 0) {
> > > +               free_irq(fep->pps_irq, ndev);
> > > +               fep->pps_irq = -1;
> > > +       }
> > > +
> >
> > I think fep->pps_irq > 0 would be more appropriate here,
> > since the IRQ returned by platform_get_irq_byname_optional() or
> > platform_get_irq_optional() cannot be 0.
> 
> Hi Phuc,
> 
> You're right. I hadn't checked whether these helpers could return
> IRQ 0.
> 
> I saw the existing irq >= 0 check in the request path and wanted
> the release path to match it. I used -1 to distinguish "no IRQ
> requested" from a possible IRQ 0, recording the IRQ only after
> request_irq() succeeded.
> 
> Since both helpers reject IRQ 0, that extra handling is unnecessary.
> Keeping pps_irq at its initial value of 0 and checking > 0 before
> free_irq() is sufficient.
> 
> Wei, I plan to drop the -1 initialization, use > 0 for the release
> check, and reset pps_irq to 0 after freeing it. Do you have any
> concerns with that change?
> 

The only concern is that your current series is conflicted with Phuc.
You'd better to rebase your patches after Phuc's patches are applied
to the net-next tree, and then post them for review. :)