Re: [PATCH 9/25] irq: Add a dynamic irq creation API

[Jeff Garzik]

Benjamin Herrenschmidt wrote:
> I have neither b) nor c) nowadays on powerpc.... "linux" irq numbers are
> purely a virtual thing, that is an index in irq_desc array and something
> we give to drivers to do request_irq() from. They can map onto hw
> interrupts, MSI-like messages, environment interrupts, could be
> hypervisor messgaes, in fact, it could be anything that remotely looks
> like an interrupt and the concept of "hw vector" is very blurry here...
> every interrupt controller defines it's own hardware vector space. On
> pSeries, hardware vectors are fairly big numbers that can encode the
> geographical location of the slot where the device is connected to, on
> some other hypervisor, they are 64 bits "tokens" representing an
> hypervisor object that can send events, etc etc....

Indeed...  The return value from return_irq() is purely a cookie, and 
has been for quite some time.

	Jeff


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/