Hello,

> >Anyway, I moved gc_enable_intr() behind the actual processing so it would
> >catch lost interrupts - it occasionally does so, I couldn't find any
> >negative side effects so far ( which of course doesn't mean there are none
> >) but now fsck'ing an USB disk doesn't disrupt audio anymore so I'll leave
> >it in place to see if everything else works as it should.
> 
> Could you be more specific about "behind?" Where in the code did you move
> it to?

Same position as the OpenPIC version.
( ok, logically the same position since only one of them will ever be called )

	while ((hwpend = (ci->ci_ipending & ~pcpl & HWIRQ_MASK)) != 0) {
		irq = 31 - cntlzw(hwpend);
		is = &intrsources[irq];
#if 0
		if (!have_openpic)
			gc_enable_irq(is->is_hwirq);
#endif
		ci->ci_ipending &= ~(1 << irq);
		splraise(is->is_mask);
		mtmsr(emsr);
		KERNEL_LOCK(LK_CANRECURSE|LK_EXCLUSIVE);
		ih = is->is_hand;
		while (ih) {
			(*ih->ih_fun)(ih->ih_arg);
			ih = ih->ih_next;
		}
		KERNEL_UNLOCK();
		mtmsr(dmsr);
		ci->ci_cpl = pcpl;
		is->is_ev.ev_count++;
		if (have_openpic)
			openpic_enable_irq(is->is_hwirq, is->is_type);
#if 1
		if (!have_openpic)
			gc_enable_irq(is->is_hwirq);
#endif

As I said earlier I changed gc_enable_irq() to collect lost interrupts and mark them pending again so this will loop over the handlers again if it finds one.

have fun
Michael