Re: interrupt cleanup #2

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Tue, Aug 28, 2018 at 01:53:44PM +0530, Cherry G.Mathew wrote:
> Cherry G.Mathew <cherry%zyx.in@localhost> writes:
> 
> > Manuel Bouyer <bouyer%antioche.eu.org@localhost> writes:
> >
> >> On Tue, Aug 28, 2018 at 10:44:32AM +0530, Cherry G.Mathew wrote:
> >>> [...]
> >>> +	extern int pirq_interrupt(void *);
> >>> +	if (event_set_handler(evtchn, pirq_interrupt, pih, level,
> >>> +		intrstr, xname) != 0) {
> >>> +		kmem_free(pih, sizeof(struct pintrhand));
> >>> +		return NULL;
> >>> +	}
> >>
> >> Please don't do this. pirq_interrupt() should be declared in a header file.
> >
> > Hi Manuel,
> >
> > I'm trying to remove pirq_interrupt() eventually. Let me send the rest
> > of the patches so you can comment with the full pictures.
> 
> Just to clarify - are you objecting to the semantic, or the design ? ie;
> are you ok with the direction of the patch other than the fact that
> pirq_interrupt() isn't declared in a header file ?

I'm not sure why you're masking irq with 0xff as pirq.
Can't this lead to conflicts or false routing on systems with multiple
IOPICs ?

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--