On Mon, Aug 20, 2018 at 1:17 PM Ben Pfaff <[email protected]> wrote:
>
> [dropping iovisor-dev as I'm not on that list]
>
> On Fri, Aug 17, 2018 at 05:29:33PM -0700, William Tu wrote:
> > The patch series introduces AF_XDP support for OVS netdev.
> > AF_XDP is a new address family working together with eBPF.
> > In short, a socket with AF_XDP family can receive and send
> > packets from an eBPF/XDP program attached to the netdev.
> > For more details about AF_XDP, please see linux kernel's
> > Documentation/networking/af_xdp.rst
> >
> > OVS has a couple of netdev types, i.e., system, tap, or
> > internal. The patch first adds a new netdev types called
> > "afxdp", and implement its configuration, packet reception,
> > and transmit functions. Since the AF_XDP socket, xsk,
> > operates in userspace, once ovs-vswitchd receives packets
> > from xsk, the proposed architecture re-uses the existing
> > userspace dpif-netdev datapath. As a result, most of
> > the packet processing happens at the userspace instead of
> > linux kernel.
>
> This is really cool. BPF and XDP and AF_XDP are all fascinating
> technology and I'd like to see OVS able to take advantage of them.
>
> What do you see as the next step for this patch series? I see that it's
> an RFC and that it depends on a fairly long patch series. Is the next
> step to rebase the series and start to get parts of it into the OVS
> tree? Or is it to refine this AF_XDP patch series first? Or do you
> have some other approach in mind?
Hi Ben,
There are still lots to do, my plan is below:
v2: remove the extra copy by adding a buffer management.
provide some preliminary performance number
v3: add more test cases to "make check-afxdp" and try the
zero copy mode on supported NIC
v4: remove the long dependency on eBPF datapath patch
Thanks,
William
>
> Thanks,
>
> Ben.
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
