[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 1/4] x86/hvm: Centralize and simplify the RTC IRQ logic.

On 28/03/2013 15:30, "Tim Deegan" <tim@xxxxxxx> wrote:

>> And it further doesn't help that we don't even have
>> vioapic_irq_negative_edge() as counterpart to
>> vioapic_irq_positive_edge(), i.e. we're not even capable of truly
>> delivering an active low IRQ.
> I get the impression that the xen IRQ model doesn't actually include a
> concept of 'active high' vs 'active low', just 'asserted' or 'not
> asserted'.  Keir?

Yes, this is correct. Possibly the naming of v[ioa]pic_irq_positive_edge is
unfortunate! Really it is indicating an asserting edge, regardless of
whether that edge is driving to a high or low voltage.

 -- Keir

>> Which sadly makes me feel even more
>> nervous about your change here. Plus if IRQ8 is active low in our
>> emulation, then the if and else bodies would need to be switched
>> (which then wouldn't work because of deassert_irq() being too
>> simplistic).
> OK, so should I just respin without patch 1/4 and pretend I never saw
> any of this mess? :)
> Tim.

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.