Re: [PATCH v2] xen/arm: implement GICD_I[S/C]ACTIVER reads

[Julien Grall <julien@xxxxxxx>]

On 17/04/2020 17:31, Stefano Stabellini wrote:
> On Fri, 17 Apr 2020, Julien Grall wrote:
> > Hi Bertrand,
> >
> > On 17/04/2020 16:16, Bertrand Marquis wrote:
> > > It seems that the problem is a bit bigger then expected and will need more
> > > discussion and thinking.
> > > I did some research on my side and there are several design possible
> > > depending on what should be the goal performance or real-time behaviour
> > > (going from just give the status we have to fire a service interrupts when
> > > any interrupts is acked by a guest to refresh our internal status).
> > >
> > > In the short term, could we not agree to fix the typo by returning always 0
> > > and start a discussion on the vgic implementation ?
> > I have already pointed out multiple time now ([1], [2]) that I would not
> > oppose the temporary solution. I think this is a matter of someone (Stefano?)
> > to submit it :).
> >
> > Cheers,
> >
> > [1] https://lists.xenproject.org/archives/html/xen-devel/2019-11/msg01642.html
> > [2] https://lists.xenproject.org/archives/html/xen-devel/2020-04/msg00459.html
> I can submit it. Julien, would you prefer the plain always return 0
> patch, or would you prefer if I tried to get the latest ISACTIVER
> information (like this patch does) but only for the local processor?
Always returning 0.

Cheers,

--
Julien Grall