[PATCH v5 06/27] irq_domain/powerpc: eliminate irq_map; use irq_alloc_desc() instead

[Benjamin Herrenschmidt]

On Fri, 2012-04-06 at 13:51 +0200, Andreas Schwab wrote:
> Benjamin Herrenschmidt <benh at kernel.crashing.org> writes:
> 
> > Ok. Doesn't matter anyway, this shouldn't be the problem in this
> > specific case. IE. we shouldn't be setting that interrupt to NONE.
> 
> After removinge the call to irq_set_irq_type in mpic_host_map the irq
> problem disappears.
> 
> Instead of the irq_set_irq_type(,NONE) calls from mpic_host_map I see
> corresponding irq_set_irq_type(,LEVEL_LOW) calls now.

It's arguable that this irq_set_irq_type(,NONE) shouln't be there but
still ... it's been around for ever and things worked :-) So something
-else- is causing the problem and I'd like to understand what exactly.

(When I say arguable, I do mean it. There are some reasons to keep it
even if we agree that it's not "setting a default" but marking the
interrupt as somewhat disabled as Russell wants, that's fine with me, an
MPIC external interrupt should -always- have a proper type set before
being used).

First we need to fix that business with NR_IRQS/nr_irqs everywhere, then
if the problem persists, check why we aren't calling the proper
irq_set_irq_type() after the mapping is established (we should be).

Cheers,
Ben.