On Thu, Jul 17, 2014 at 3:46 PM, Ansis Atteka <aatt...@nicira.com> wrote: > diff --git a/datapath/flow_netlink.c b/datapath/flow_netlink.c > index 5f975a1..a4108c0 100644 > --- a/datapath/flow_netlink.c > +++ b/datapath/flow_netlink.c > @@ -1034,7 +1034,7 @@ int ovs_nla_put_flow(struct datapath *dp, const struct > sw_flow_key *swkey, > struct vport *in_port; > > in_port = ovs_vport_ovsl_rcu(dp, swkey->phy.in_port); > - if (in_port->ops->type == OVS_VPORT_TYPE_GENEVE) > + if (in_port && in_port->ops->type == > OVS_VPORT_TYPE_GENEVE)
I think is probably not only an issue for Geneve ports (since the crash happens before we get to check the type) but for tunnel ports in general. I also guess that the scenario is not that the port hasn't been created yet but that the port has been deleted and the flow still exists. One thing that I worry about is that this has the possibility to change how the flow is reported - before the flow deletion it has Geneve options but immediately after the flow still exists but without options. OVS will likely deal with this but it doesn't seem like a great thing. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
