Re: [Xen-devel] [PATCH QEMU] xen/pt: allow QEMU to request MSI unmasking at bind time

[Roger Pau Monne <roger.pau@xxxxxxxxxx>]

On Thu, Aug 24, 2017 at 04:13:58AM -0600, Jan Beulich wrote:
> >>> On 24.08.17 at 12:06, <roger.pau@xxxxxxxxxx> wrote:
> > On Thu, Aug 24, 2017 at 03:54:21AM -0600, Jan Beulich wrote:
> >> >>> On 24.08.17 at 11:47, <roger.pau@xxxxxxxxxx> wrote:
> >> > @@ -274,7 +278,7 @@ int xen_pt_msi_update(XenPCIPassthroughState *s)
> >> >  {
> >> >      XenPTMSI *msi = s->msi;
> >> >      return msi_msix_update(s, msi_addr64(msi), msi->data, msi->pirq,
> >> > -                           false, 0, &msi->pirq);
> >> > +                           false, 0, &msi->pirq, false);
> >> >  }
> >> 
> >> I don't think this is correct when the device has mask bits.
> > 
> > Right, I thought I modified this. I've already changed
> > pt_irq_create_bind so that the original behavior is keep if the unmask
> > bit is not set. In this case this should be 'true' in order to keep
> > the previous behavior, which was correct for MSI.
> 
> Wouldn't you want to pass the state of the mask bit here,
> rather than uniformly hard coding true or false?

Yes, I think so. I've overlooked the MSI code because I thought we
allowed QEMU to directly write to the mask register, but that's not
true, it's trapped by Xen.

Thanks, Roger.

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