>
> ~RTE_VHOST_USER_CLIENT means that every other bit is set to one, which
> affects other flags.
> I think the current code is ok: it's not wrokg to assume that the newly
> initialized structure has every member set to 0.
> If you feel strongly about it, I think it would be better to set the single
> bit in
> vhost_driver_flags.
Ok - at the moment in DPDK if the last bit in the flags is zero the others
aren't checked, but it is probably better to just toggle the last bit as you
suggested.
Anyway, if we can rely on the zero initialisation like you said, this patch
isn't needed.
Thanks,
Ciara
> Thanks,
> Daniele
>
> 2016-08-18 5:29 GMT-07:00 Loftus, Ciara <ciara.lof...@intel.com>:
> >
> > rte_vhost_driver_register accepts a 'mode' flag which sets server(0) or
> > client(1) mode. vHost devices are registered in 'server' mode (0) when
> > initially created. Before this patch the flags provided to this
> > function were the device's 'vhost_driver_flags' which hadn't been set
> > yet when the register function is called but happen to be zero. Provide
> > an explicit flag instead for better practice.
> >
> > Fixes: ("c1ff66ac80b5: netdev-dpdk: vHost client mode and reconnect")
>
> Apologies for mis-format. Should read:
> Fixes: c1ff66ac80b5 ("netdev-dpdk: vHost client mode and reconnect")
>
> > Signed-off-by: Ciara Loftus <ciara.lof...@intel.com>
> > ---
> > lib/netdev-dpdk.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
> > index e5f2cdd..32853ce 100644
> > --- a/lib/netdev-dpdk.c
> > +++ b/lib/netdev-dpdk.c
> > @@ -918,7 +918,7 @@ netdev_dpdk_vhost_construct(struct netdev
> > *netdev)
> > vhost_sock_dir, name);
> >
> > err = rte_vhost_driver_register(dev->vhost_server_id,
> > - dev->vhost_driver_flags);
> > + ~RTE_VHOST_USER_CLIENT);
> > if (err) {
> > VLOG_ERR("vhost-user socket device setup failure for socket %s\n",
> > dev->vhost_server_id);
> > --
> > 2.4.3
> >
> > _______________________________________________
> > dev mailing list
> > dev@openvswitch.org
> > http://openvswitch.org/mailman/listinfo/dev
> _______________________________________________
> dev mailing list
> dev@openvswitch.org
> http://openvswitch.org/mailman/listinfo/dev
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
