Re: [PATCH] genirq: spurious irq detection for threaded irqs

["Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>]

On Thu, Mar 04, 2010 at 11:26:20PM +0100, Thomas Gleixner wrote:
> On Thu, 4 Mar 2010, Uwe Kleine-König wrote:
> > -void note_interrupt(unsigned int irq, struct irq_desc *desc,
> > -		    irqreturn_t action_ret)
> > +void note_threaded_interrupt(unsigned int irq, struct irq_desc *desc,
> > +		irqreturn_t action_ret)
> >  {
> >  	if (unlikely(action_ret != IRQ_HANDLED)) {
> >  		/*
> > @@ -262,6 +262,19 @@ void note_interrupt(unsigned int irq, struct irq_desc *desc,
> >  	desc->irqs_unhandled = 0;
> >  }
> >  
> > +void note_interrupt(unsigned int irq, struct irq_desc *desc,
> > +		irqreturn_t action_ret)
> > +{
> > +	if (action_ret == IRQ_WAKE_THREAD)
> > +		/* handled in irq_thread() when the threaded handler returns */
> > +		return;
> > +
> > +	/* don't report IRQ_WAKE_THREAD | IRQ_HANDLED as bogus return value */
> > +	action_ret &= ~IRQ_WAKE_THREAD;
> > +
> > +	note_threaded_interrupt(irq, desc, action_ret);
> > +}
> > +
> 
> We don't need an extra function for that. A simple
> 
> 	if (action_ret & IRQ_WAKE_THREAD)
> 		return;
> 
> in note_interrupt() is sufficient to cover everything.
With your suggestion if action_ret == IRQ_WAKE_THREAD | IRQ_HANDLED then
the IRQ_HANDLED bit doesn't get noted.  Does that matter?

But still you're right about the extra function.  Assuming a threaded
handler doesn't return IRQ_WAKE_THREAD my note_interrupt does the same
as note_threaded_interrupt.

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |