openvswitch: Fix pop_vlan action for double tagged frames
Subsystems (3)
PATCH v2 1/1 openvswitch: Fix pop_vlan action for double tagged frames
skb_vlan_pop() expects skb->protocol to be a valid TPID for double tagged frames, but skb->protocol is set to the ethertype by key_extract(). So temporarily set it to the TPID when doing a pop_vlan.
--- net/openvswitch/actions.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
net/openvswitch/actions.c +10 -0
@@ -280,6 +280,13 @@ static int set_mpls(struct sk_buff *skb, struct sw_flow_key *flow_key,static int pop_vlan (struct sk_buff *skb,struct sw_flow_key *key) {int err; +__be16 proto = skb->protocol; + +/* skb->protocol is set to the inner most parsed ethertype. To satisfy +* skb_vlan_pop() for multi-tagged frames we must set it to the tpid. +*/ +if (is_flow_key_valid (key) && key->eth.vlan.tci && key->eth.cvlan.tci) + skb->protocol = key->eth.cvlan.tpid; err =skb_vlan_pop (skb);if (skb_vlan_tag_present (skb)) {
@@ -288,6 +295,9 @@ static int pop_vlan(struct sk_buff *skb, struct sw_flow_key *key) key->eth.vlan.tci =0 ; key->eth.vlan.tpid =0 ; } + + skb->protocol = proto; +return err; }
Signature · 1 line
2.12.0
net/openvswitch/actions.c static int set_mpls(struct sk_buff *skb, struct sw_flow_key *flow_key, · line 288
Quoted 1 line · Eric Garver
net/openvswitch/actions.c +1
-0+if (is_flow_key_valid (key) && key->eth.vlan.tci && key->eth.cvlan.tci)
Maybe (key->eth.vlan.tci & htons(VLAN_TAG_PRESENT)) for consistency with the rest of the code? But it's just nitpicking.
The real problem here is when a double tagged packet leaves the ovs bridge, it won't have the skb->protocol that the kernel expects: it will be ethertype of the payload, while my understanding is it should be the inner tpid, right?
This patch fixes that nicely for the pop vlan case. But what about other cases? It seems to me that we need to add the logic to key_extract.
Thanks!
Jiri
Quoted 10 lines · Jiri Benc
net/openvswitch/actions.c static int set_mpls(struct sk_buff *skb, struct sw_flow_key *flow_key, · line 288
Quoted 1 line · Eric Garver
net/openvswitch/actions.c +1
-0+if (is_flow_key_valid (key) && key->eth.vlan.tci && key->eth.cvlan.tci)Maybe (
key->eth.vlan.tci&htons(VLAN_TAG_PRESENT)) for consistency with the rest of the code? But it's just nitpicking.The real problem here is when a double tagged packet leaves the ovs bridge, it won't have the
skb->protocolthat the kernel expects: it will be ethertype of the payload, while my understanding is it should be the inner tpid, right?
Right.
Quoted 5 lines
This patch fixes that nicely for the pop vlan case. But what about other cases? It seems to me that we need to add the logic to
key_extract.
The part I was missing is; before encap into the L3 tunnel all the VLAN tags must be explicitly popped.
Setting skb->protocol to the TPID for double tagged frames means the pop operations shift the payload ethertype into skb->protocol.
I'll send a v3 that does this is key_extract().
net/openvswitch/actions.c static int set_mpls(struct sk_buff *skb, struct sw_flow_key *flow_key, · line 288
Quoted 1 line · Eric Garver
net/openvswitch/actions.c +1
-0+if (is_flow_key_valid (key) && key->eth.vlan.tci && key->eth.cvlan.tci)
Maybe (key->eth.vlan.tci & htons(VLAN_TAG_PRESENT)) for consistency with the rest of the code? But it's just nitpicking.
The real problem here is when a double tagged packet leaves the ovs bridge, it won't have the skb->protocol that the kernel expects: it will be ethertype of the payload, while my understanding is it should be the inner tpid, right?
This patch fixes that nicely for the pop vlan case. But what about other cases? It seems to me that we need to add the logic to key_extract.
Thanks!
Jiri