[PATCH net 0/3] openvswitch: vlan fixes

Fix three issues introduced by the 802.1AD patchset.

Jiri Benc (3):
  openvswitch: vlan: remove wrong likely statement
  openvswitch: fix vlan subtraction from packet length
  openvswitch: add NETIF_F_HW_VLAN_STAG_TX to internal dev

 net/openvswitch/flow.c               | 2 +-
 net/openvswitch/vport-internal_dev.c | 2 +-
 net/openvswitch/vport.c              | 3 ++-
 3 files changed, 4 insertions(+), 3 deletions(-)

-- 
1.8.3.1

This code is called whenever flow key is being extracted from the packet.
The packet may be as likely vlan tagged as not.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>
---
 net/openvswitch/flow.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index c8c82e109c68..22087062bd10 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c

@@ -343,7 +343,7 @@ static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
 	key->eth.cvlan.tci = 0;
 	key->eth.cvlan.tpid = 0;
 
-	if (likely(skb_vlan_tag_present(skb))) {
+	if (skb_vlan_tag_present(skb)) {
 		key->eth.vlan.tci = htons(skb->vlan_tci);
 		key->eth.vlan.tpid = skb->vlan_proto;
 	} else {

-- 
1.8.3.1

When the packet has its vlan tag in skb->vlan_tci, the length of the VLAN
header is not counted in skb->len. It doesn't make sense to subtract it.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>
---
 net/openvswitch/vport.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/openvswitch/vport.c b/net/openvswitch/vport.c
index 8f198437c724..7387418ac514 100644
--- a/net/openvswitch/vport.c
+++ b/net/openvswitch/vport.c

@@ -485,7 +485,8 @@ static unsigned int packet_length(const struct sk_buff *skb)
 {
 	unsigned int length = skb->len - ETH_HLEN;
 
-	if (skb_vlan_tagged(skb))
+	if (!skb_vlan_tag_present(skb) &&
+	    eth_type_vlan(skb->protocol))
 		length -= VLAN_HLEN;
 
 	/* Don't subtract for multiple VLAN tags. Most (all?) drivers allow

-- 
1.8.3.1

The internal device does support 802.1AD offloading since 018c1dda5ff1
("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink
attributes").

Signed-off-by: Jiri Benc <redacted>
---
 net/openvswitch/vport-internal_dev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/openvswitch/vport-internal_dev.c b/net/openvswitch/vport-internal_dev.c
index 95c36147a6e1..e7da29021b38 100644
--- a/net/openvswitch/vport-internal_dev.c
+++ b/net/openvswitch/vport-internal_dev.c

@@ -176,7 +176,7 @@ static void do_setup(struct net_device *netdev)
 
 	netdev->vlan_features = netdev->features;
 	netdev->hw_enc_features = netdev->features;
-	netdev->features |= NETIF_F_HW_VLAN_CTAG_TX;
+	netdev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX;
 	netdev->hw_features = netdev->features & ~NETIF_F_LLTX;
 
 	eth_hw_addr_random(netdev);

-- 
1.8.3.1

On Mon, Oct 10, 2016 at 8:02 AM, Jiri Benc [off-list ref] wrote:

This code is called whenever flow key is being extracted from the packet.
The packet may be as likely vlan tagged as not.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>

Acked-by: Pravin B Shelar <redacted>

On Mon, Oct 10, 2016 at 8:02 AM, Jiri Benc [off-list ref] wrote:

The internal device does support 802.1AD offloading since 018c1dda5ff1
("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink
attributes").

Signed-off-by: Jiri Benc <redacted>

Acked-by: Pravin B Shelar <redacted>

On Mon, Oct 10, 2016 at 8:02 AM, Jiri Benc [off-list ref] wrote:

When the packet has its vlan tag in skb->vlan_tci, the length of the VLAN
header is not counted in skb->len. It doesn't make sense to subtract it.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>

Acked-by: Pravin B Shelar <redacted>

On Mon, Oct 10, 2016 at 05:02:42PM +0200, Jiri Benc wrote:

This code is called whenever flow key is being extracted from the packet.
The packet may be as likely vlan tagged as not.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>

Acked-by: Eric Garver <redacted>

On Mon, Oct 10, 2016 at 05:02:43PM +0200, Jiri Benc wrote:

When the packet has its vlan tag in skb->vlan_tci, the length of the VLAN
header is not counted in skb->len. It doesn't make sense to subtract it.

Fixes: 018c1dda5ff1 ("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink attributes")
Signed-off-by: Jiri Benc <redacted>

Acked-by: Eric Garver <redacted>

On Mon, Oct 10, 2016 at 05:02:44PM +0200, Jiri Benc wrote:

The internal device does support 802.1AD offloading since 018c1dda5ff1
("openvswitch: 802.1AD Flow handling, actions, vlan parsing, netlink
attributes").

Signed-off-by: Jiri Benc <redacted>

Acked-by: Eric Garver <redacted>

From: Jiri Benc <redacted>
Date: Mon, 10 Oct 2016 17:02:41 +0200

Fix three issues introduced by the 802.1AD patchset.

Series applied, thanks Jiri.