Re: [PATCH 2/4] kvm: Add missing KVM_IRQFD API documentation

["Michael S. Tsirkin" <mst@redhat.com>]

On Sun, Jun 24, 2012 at 08:56:17AM -0600, Alex Williamson wrote:
> On Sun, 2012-06-24 at 11:34 +0300, Michael S. Tsirkin wrote:
> > On Fri, Jun 22, 2012 at 04:15:50PM -0600, Alex Williamson wrote:
> > > Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
> > > ---
> > > 
> > >  Documentation/virtual/kvm/api.txt |   17 +++++++++++++++++
> > >  1 file changed, 17 insertions(+)
> > > 
> > > diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> > > index 310fe50..9b4cb2b 100644
> > > --- a/Documentation/virtual/kvm/api.txt
> > > +++ b/Documentation/virtual/kvm/api.txt
> > > @@ -1965,6 +1965,23 @@ return the hash table order in the parameter.  (If the guest is using
> > >  the virtualized real-mode area (VRMA) facility, the kernel will
> > >  re-create the VMRA HPTEs on the next KVM_RUN of any vcpu.)
> > >  
> > > +4.76 KVM_IRQFD
> > > +
> > > +Capability: KVM_CAP_IRQFD
> > > +Architectures: x86
> > > +Type: vm ioctl
> > > +Parameters: struct kvm_irqfd (in)
> > > +Returns: 0 on success, -1 on error
> > > +
> > > +Allows setting an eventfd to directly trigger a guest interrupt
> > > +kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
> > > +kvm_irqfd.gsi specifies the irqchip pin toggled by this event.  By
> > > +default this interface only supports edge triggered interrupts,
> > > +meaning the specified gsi is asserted and immediately de-asserted
> > > +when the eventfd is triggered.
> > 
> > That's a bit confusing. This assert/deassert only has effect for level.
> > Do we or do we not want to maintain this assert/deassert behaviour
> > for level irqfds? If yes we shouldn't say it's unsupported if
> > no we should not document it.
> > 
> > Gleb, Avi, any thoughts?
> 
> AIUI the assert/deassert manages to work for level (I actually had vfio
> injecting these for a while and it seemed to work), but I don't think it
> was ever designed for level interrupts and attempting to do that is
> unsupported.  Thanks,
> 
> Alex

So don't document assert/deassert then.

> > >  Specifying KVM_IRQFD_FLAG_DEASSIGN
> > > +removes the previously set irqfd matching both kvm_irqfd.fd and
> > > +kvm_irqfd.gsi.
> > > +
> > >  
> > >  5. The kvm_run structure
> > >  ------------------------
> 
>