Re: [PATCH net-next v1 1/3] vm_sockets: Include flag field in the vsock address data structure

[Stefano Garzarella]

On Thu, Dec 03, 2020 at 12:32:08PM +0200, Paraschiv, Andra-Irina wrote:
> On 03/12/2020 11:21, Stefan Hajnoczi wrote:
> > On Tue, Dec 01, 2020 at 05:25:03PM +0200, Andra Paraschiv wrote:
> > > vsock enables communication between virtual machines and the host they
> > > are running on. With the multi transport support (guest->host and
> > > host->guest), nested VMs can also use vsock channels for communication.
> > >
> > > In addition to this, by default, all the vsock packets are forwarded to
> > > the host, if no host->guest transport is loaded. This behavior can be
> > > implicitly used for enabling vsock communication between sibling VMs.
> > >
> > > Add a flag field in the vsock address data structure that can be used to
> > > explicitly mark the vsock connection as being targeted for a certain
> > > type of communication. This way, can distinguish between nested VMs and
> > > sibling VMs use cases and can also setup them at the same time. Till
> > > now, could either have nested VMs or sibling VMs at a time using the
> > > vsock communication stack.
> > >
> > > Use the already available "svm_reserved1" field and mark it as a flag
> > > field instead. This flag can be set when initializing the vsock address
> > > variable used for the connect() call.
> > >
> > > Signed-off-by: Andra Paraschiv <andraprs@xxxxxxxxxx>
> > > ---
> > >  include/uapi/linux/vm_sockets.h | 18 +++++++++++++++++-
> > >  1 file changed, 17 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/include/uapi/linux/vm_sockets.h b/include/uapi/linux/vm_sockets.h
> > > index fd0ed7221645d..58da5a91413ac 100644
> > > --- a/include/uapi/linux/vm_sockets.h
> > > +++ b/include/uapi/linux/vm_sockets.h
> > > @@ -114,6 +114,22 @@
> > >  #define VMADDR_CID_HOST 2
> > > +/* This sockaddr_vm flag value covers the current default use case:
> > > + * local vsock communication between guest and host and nested VMs setup.
> > > + * In addition to this, implicitly, the vsock packets are forwarded to the host
> > > + * if no host->guest vsock transport is set.
> > > + */
> > > +#define VMADDR_FLAG_DEFAULT_COMMUNICATION	0x0000
> > > +
> > > +/* Set this flag value in the sockaddr_vm corresponding field if the vsock
> > > + * channel needs to be setup between two sibling VMs running on the same host.
> > > + * This way can explicitly distinguish between vsock channels created for nested
> > > + * VMs (or local communication between guest and host) and the ones created for
> > > + * sibling VMs. And vsock channels for multiple use cases (nested / sibling VMs)
> > > + * can be setup at the same time.
> > > + */
> > > +#define VMADDR_FLAG_SIBLING_VMS_COMMUNICATION	0x0001
> > vsock has the h2g and g2h concept. It would be more general to call this
> > flag VMADDR_FLAG_G2H or less cryptically VMADDR_FLAG_TO_HOST.

I agree, VMADDR_FLAG_TO_HOST is more general and it's clearer that is up 
to the host where to forward the packet (sibling if supported, or 
whatever).

Thanks,
Stefano

> Thanks for the feedback, Stefan.
>
> I can update the naming to be more general, such as  "_TO_HOST", and 
> keep the use cases (e.g. guest <-> host / nested / sibling VMs 
> communication) mention in the comments so that would relate more to 
> the motivation behind it.
>
> Andra
>
> > That way it just tells the driver in which direction to send packets
> > without implying that sibling communication is possible (it's not
> > allowed by default on any transport).
> >
> > I don't have a strong opinion on this but wanted to suggest the idea.
> >
> > Stefan
>
>
>
>
> Amazon Development Center (Romania) S.R.L. registered office: 27A Sf. Lazar Street, UBC5, floor 2, Iasi, Iasi County, 700045, Romania. Registered in Romania. Registration number J22/2621/2005.