Commit 5108bbad authored by Jiri Benc's avatar Jiri Benc Committed by David S. Miller

openvswitch: add processing of L3 packets

Support receiving, extracting flow key and sending of L3 packets (packets
without an Ethernet header).

Note that even after this patch, non-Ethernet interfaces are still not
allowed to be added to bridges. Similarly, netlink interface for sending and
receiving L3 packets to/from user space is not in place yet.

Based on previous versions by Lorand Jakab and Simon Horman.
Signed-off-by: default avatarLorand Jakab <lojakab@cisco.com>
Signed-off-by: default avatarSimon Horman <simon.horman@netronome.com>
Signed-off-by: default avatarJiri Benc <jbenc@redhat.com>
Acked-by: default avatarPravin B Shelar <pshelar@ovn.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 1560a074
...@@ -560,7 +560,6 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info) ...@@ -560,7 +560,6 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
struct sw_flow *flow; struct sw_flow *flow;
struct sw_flow_actions *sf_acts; struct sw_flow_actions *sf_acts;
struct datapath *dp; struct datapath *dp;
struct ethhdr *eth;
struct vport *input_vport; struct vport *input_vport;
u16 mru = 0; u16 mru = 0;
int len; int len;
...@@ -581,17 +580,6 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info) ...@@ -581,17 +580,6 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
nla_memcpy(__skb_put(packet, len), a[OVS_PACKET_ATTR_PACKET], len); nla_memcpy(__skb_put(packet, len), a[OVS_PACKET_ATTR_PACKET], len);
skb_reset_mac_header(packet);
eth = eth_hdr(packet);
/* Normally, setting the skb 'protocol' field would be handled by a
* call to eth_type_trans(), but it assumes there's a sending
* device, which we may not have. */
if (eth_proto_is_802_3(eth->h_proto))
packet->protocol = eth->h_proto;
else
packet->protocol = htons(ETH_P_802_2);
/* Set packet's mru */ /* Set packet's mru */
if (a[OVS_PACKET_ATTR_MRU]) { if (a[OVS_PACKET_ATTR_MRU]) {
mru = nla_get_u16(a[OVS_PACKET_ATTR_MRU]); mru = nla_get_u16(a[OVS_PACKET_ATTR_MRU]);
...@@ -618,6 +606,7 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info) ...@@ -618,6 +606,7 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
rcu_assign_pointer(flow->sf_acts, acts); rcu_assign_pointer(flow->sf_acts, acts);
packet->priority = flow->key.phy.priority; packet->priority = flow->key.phy.priority;
packet->mark = flow->key.phy.skb_mark; packet->mark = flow->key.phy.skb_mark;
packet->protocol = flow->key.eth.type;
rcu_read_lock(); rcu_read_lock();
dp = get_dp_rcu(net, ovs_header->dp_ifindex); dp = get_dp_rcu(net, ovs_header->dp_ifindex);
......
...@@ -334,14 +334,17 @@ static int parse_vlan_tag(struct sk_buff *skb, struct vlan_head *key_vh) ...@@ -334,14 +334,17 @@ static int parse_vlan_tag(struct sk_buff *skb, struct vlan_head *key_vh)
return 1; return 1;
} }
static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key) static void clear_vlan(struct sw_flow_key *key)
{ {
int res;
key->eth.vlan.tci = 0; key->eth.vlan.tci = 0;
key->eth.vlan.tpid = 0; key->eth.vlan.tpid = 0;
key->eth.cvlan.tci = 0; key->eth.cvlan.tci = 0;
key->eth.cvlan.tpid = 0; key->eth.cvlan.tpid = 0;
}
static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
{
int res;
if (skb_vlan_tag_present(skb)) { if (skb_vlan_tag_present(skb)) {
key->eth.vlan.tci = htons(skb->vlan_tci); key->eth.vlan.tci = htons(skb->vlan_tci);
...@@ -483,17 +486,20 @@ static int parse_icmpv6(struct sk_buff *skb, struct sw_flow_key *key, ...@@ -483,17 +486,20 @@ static int parse_icmpv6(struct sk_buff *skb, struct sw_flow_key *key,
* *
* Returns 0 if successful, otherwise a negative errno value. * Returns 0 if successful, otherwise a negative errno value.
* *
* Initializes @skb header pointers as follows: * Initializes @skb header fields as follows:
* *
* - skb->mac_header: the Ethernet header. * - skb->mac_header: the L2 header.
* *
* - skb->network_header: just past the Ethernet header, or just past the * - skb->network_header: just past the L2 header, or just past the
* VLAN header, to the first byte of the Ethernet payload. * VLAN header, to the first byte of the L2 payload.
* *
* - skb->transport_header: If key->eth.type is ETH_P_IP or ETH_P_IPV6 * - skb->transport_header: If key->eth.type is ETH_P_IP or ETH_P_IPV6
* on output, then just past the IP header, if one is present and * on output, then just past the IP header, if one is present and
* of a correct length, otherwise the same as skb->network_header. * of a correct length, otherwise the same as skb->network_header.
* For other key->eth.type values it is left untouched. * For other key->eth.type values it is left untouched.
*
* - skb->protocol: the type of the data starting at skb->network_header.
* Equals to key->eth.type.
*/ */
static int key_extract(struct sk_buff *skb, struct sw_flow_key *key) static int key_extract(struct sk_buff *skb, struct sw_flow_key *key)
{ {
...@@ -505,28 +511,35 @@ static int key_extract(struct sk_buff *skb, struct sw_flow_key *key) ...@@ -505,28 +511,35 @@ static int key_extract(struct sk_buff *skb, struct sw_flow_key *key)
skb_reset_mac_header(skb); skb_reset_mac_header(skb);
/* Link layer. We are guaranteed to have at least the 14 byte Ethernet /* Link layer. */
* header in the linear data area. clear_vlan(key);
*/ if (key->mac_proto == MAC_PROTO_NONE) {
eth = eth_hdr(skb); if (unlikely(eth_type_vlan(skb->protocol)))
ether_addr_copy(key->eth.src, eth->h_source); return -EINVAL;
ether_addr_copy(key->eth.dst, eth->h_dest);
__skb_pull(skb, 2 * ETH_ALEN); skb_reset_network_header(skb);
/* We are going to push all headers that we pull, so no need to } else {
* update skb->csum here. eth = eth_hdr(skb);
*/ ether_addr_copy(key->eth.src, eth->h_source);
ether_addr_copy(key->eth.dst, eth->h_dest);
if (unlikely(parse_vlan(skb, key))) __skb_pull(skb, 2 * ETH_ALEN);
return -ENOMEM; /* We are going to push all headers that we pull, so no need to
* update skb->csum here.
*/
key->eth.type = parse_ethertype(skb); if (unlikely(parse_vlan(skb, key)))
if (unlikely(key->eth.type == htons(0))) return -ENOMEM;
return -ENOMEM;
skb->protocol = parse_ethertype(skb);
if (unlikely(skb->protocol == htons(0)))
return -ENOMEM;
skb_reset_network_header(skb); skb_reset_network_header(skb);
__skb_push(skb, skb->data - skb_mac_header(skb));
}
skb_reset_mac_len(skb); skb_reset_mac_len(skb);
__skb_push(skb, skb->data - skb_mac_header(skb)); key->eth.type = skb->protocol;
/* Network layer. */ /* Network layer. */
if (key->eth.type == htons(ETH_P_IP)) { if (key->eth.type == htons(ETH_P_IP)) {
...@@ -721,9 +734,25 @@ int ovs_flow_key_update(struct sk_buff *skb, struct sw_flow_key *key) ...@@ -721,9 +734,25 @@ int ovs_flow_key_update(struct sk_buff *skb, struct sw_flow_key *key)
return key_extract(skb, key); return key_extract(skb, key);
} }
static int key_extract_mac_proto(struct sk_buff *skb)
{
switch (skb->dev->type) {
case ARPHRD_ETHER:
return MAC_PROTO_ETHERNET;
case ARPHRD_NONE:
if (skb->protocol == htons(ETH_P_TEB))
return MAC_PROTO_ETHERNET;
return MAC_PROTO_NONE;
}
WARN_ON_ONCE(1);
return -EINVAL;
}
int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info, int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info,
struct sk_buff *skb, struct sw_flow_key *key) struct sk_buff *skb, struct sw_flow_key *key)
{ {
int res;
/* Extract metadata from packet. */ /* Extract metadata from packet. */
if (tun_info) { if (tun_info) {
key->tun_proto = ip_tunnel_info_af(tun_info); key->tun_proto = ip_tunnel_info_af(tun_info);
...@@ -751,7 +780,10 @@ int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info, ...@@ -751,7 +780,10 @@ int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info,
key->phy.skb_mark = skb->mark; key->phy.skb_mark = skb->mark;
ovs_ct_fill_key(skb, key); ovs_ct_fill_key(skb, key);
key->ovs_flow_hash = 0; key->ovs_flow_hash = 0;
key->mac_proto = MAC_PROTO_ETHERNET; res = key_extract_mac_proto(skb);
if (res < 0)
return res;
key->mac_proto = res;
key->recirc_id = 0; key->recirc_id = 0;
return key_extract(skb, key); return key_extract(skb, key);
...@@ -768,5 +800,29 @@ int ovs_flow_key_extract_userspace(struct net *net, const struct nlattr *attr, ...@@ -768,5 +800,29 @@ int ovs_flow_key_extract_userspace(struct net *net, const struct nlattr *attr,
if (err) if (err)
return err; return err;
if (ovs_key_mac_proto(key) == MAC_PROTO_NONE) {
/* key_extract assumes that skb->protocol is set-up for
* layer 3 packets which is the case for other callers,
* in particular packets recieved from the network stack.
* Here the correct value can be set from the metadata
* extracted above.
*/
skb->protocol = key->eth.type;
} else {
struct ethhdr *eth;
skb_reset_mac_header(skb);
eth = eth_hdr(skb);
/* Normally, setting the skb 'protocol' field would be
* handled by a call to eth_type_trans(), but it assumes
* there's a sending device, which we may not have.
*/
if (eth_proto_is_802_3(eth->h_proto))
skb->protocol = eth->h_proto;
else
skb->protocol = htons(ETH_P_802_2);
}
return key_extract(skb, key); return key_extract(skb, key);
} }
...@@ -485,6 +485,25 @@ void ovs_vport_send(struct vport *vport, struct sk_buff *skb, u8 mac_proto) ...@@ -485,6 +485,25 @@ void ovs_vport_send(struct vport *vport, struct sk_buff *skb, u8 mac_proto)
{ {
int mtu = vport->dev->mtu; int mtu = vport->dev->mtu;
switch (vport->dev->type) {
case ARPHRD_NONE:
if (mac_proto == MAC_PROTO_ETHERNET) {
skb_reset_network_header(skb);
skb_reset_mac_len(skb);
skb->protocol = htons(ETH_P_TEB);
} else if (mac_proto != MAC_PROTO_NONE) {
WARN_ON_ONCE(1);
goto drop;
}
break;
case ARPHRD_ETHER:
if (mac_proto != MAC_PROTO_ETHERNET)
goto drop;
break;
default:
goto drop;
}
if (unlikely(packet_length(skb, vport->dev) > mtu && if (unlikely(packet_length(skb, vport->dev) > mtu &&
!skb_is_gso(skb))) { !skb_is_gso(skb))) {
net_warn_ratelimited("%s: dropped over-mtu packet: %d > %d\n", net_warn_ratelimited("%s: dropped over-mtu packet: %d > %d\n",
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment