Re: [Xen-devel] [PATCH for-4.5 7/8] xen/irq: Handle multiple action per IRQ

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Wed, 2014-03-19 at 17:18 +0000, Julien Grall wrote:
> What is the final decision for the interrupt handling? Stefano would
> prefer to let the SMMU drivers handle a such case. Ian, do you have any
> opinion?

It does sound like the end result is that shared interrupts need to be
handled somehow.

One potential issue is that a driver needs to be somewhat aware that it
can be using a shared interrupt (and e.g. not log spurious looking
interrupts). Linux solves this by having an explicit IRQF_SHARED which
the driver must pass iff it is prepared to cope with this situation and
the core should detect mismatches, which isn't a bad solution.

I think it's pretty clear from the discussion that routing an IRQ to the
guest must be equivalent to a driver which does not pass IRQF_SHARED.

There's also the question of error handling if/when a spurious interrupt
does occur (i.e. none of the multiple handlers dealt with it). Interrupt
handlers should perhaps return a status to say whether or not they did
anything with the interrupt, which the core code can then deal with
(logging, quenching etc).

Ian.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel