Re: [Embedded-pv-devel] Driver domain under Xen

[Julien Grall <julien.grall@xxxxxxxxxx>]

On 23 January 2015 at 16:20, Andrii Tseglytskyi
<andrii.tseglytskyi@xxxxxxxxxxxxxxx> wrote:
> This is why I didn't want to send you code right now. All these
> changes work for us to have 1 to 1 IRQ mapping in driver domain and to
> release them correctly in case of domain crash. Before get them
> upstreamed I would prefer to think twice on each of them and rework if
> needed.

I was mostly interesting to know if you have found new race condition in
the guest physical IRQ management. I spent the last couple of months fixing
multiple race condition in my series. I don't consider the v1 neither v2 safe to
use for interrupts passthrough.

For instance, your clear_bit(_IRQ_DISABLED, &desc->status) is too
workaround a latent bug in gic_remove_irq_from_guest().

Anyway, it's fine to use as long as it's used for a prototype.

Regards,

-- 
Julien Grall

_______________________________________________
Embedded-pv-devel mailing list
Embedded-pv-devel@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/embedded-pv-devel