openvswitch: Add conntrack action

Expose the kernel connection tracker via OVS. Userspace components can
make use of the CT action to populate the connection state (ct_state)
field for a flow. This state can be subsequently matched.

Exposed connection states are OVS_CS_F_*:
- NEW (0x01) - Beginning of a new connection.
- ESTABLISHED (0x02) - Part of an existing connection.
- RELATED (0x04) - Related to an established connection.
- INVALID (0x20) - Could not track the connection for this packet.
- REPLY_DIR (0x40) - This packet is in the reply direction for the flow.
- TRACKED (0x80) - This packet has been sent through conntrack.

When the CT action is executed by itself, it will send the packet
through the connection tracker and populate the ct_state field with one
or more of the connection state flags above. The CT action will always
set the TRACKED bit.

When the COMMIT flag is passed to the conntrack action, this specifies
that information about the connection should be stored. This allows
subsequent packets for the same (or related) connections to be
correlated with this connection. Sending subsequent packets for the
connection through conntrack allows the connection tracker to consider
the packets as ESTABLISHED, RELATED, and/or REPLY_DIR.

The CT action may optionally take a zone to track the flow within. This
allows connections with the same 5-tuple to be kept logically separate
from connections in other zones. If the zone is specified, then the
"ct_zone" match field will be subsequently populated with the zone id.

IP fragments are handled by transparently assembling them as part of the
CT action. The maximum received unit (MRU) size is tracked so that
refragmentation can occur during output.

IP frag handling contributed by Andy Zhou.

Based on original design by Justin Pettit.

Signed-off-by: Joe Stringer <joestringer@nicira.com>
Signed-off-by: Justin Pettit <jpettit@nicira.com>
Signed-off-by: Andy Zhou <azhou@nicira.com>
Acked-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/uapi/linux/openvswitch.h

@@ -164,6 +164,9 @@ enum ovs_packet_cmd {
  * %OVS_USERSPACE_ATTR_EGRESS_TUN_PORT attribute, which is sent only if the
  * output port is actually a tunnel port. Contains the output tunnel key
  * extracted from the packet as nested %OVS_TUNNEL_KEY_ATTR_* attributes.
+ * @OVS_PACKET_ATTR_MRU: Present for an %OVS_PACKET_CMD_ACTION and
+ * %OVS_PACKET_ATTR_USERSPACE action specify the Maximum received fragment
+ * size.
  *
  * These attributes follow the &struct ovs_header within the Generic Netlink
  * payload for %OVS_PACKET_* commands.
@@ -180,6 +183,7 @@ enum ovs_packet_attr {
 	OVS_PACKET_ATTR_UNUSED2,
 	OVS_PACKET_ATTR_PROBE,      /* Packet operation is a feature probe,
 				       error logging should be suppressed. */
+	OVS_PACKET_ATTR_MRU,	    /* Maximum received IP fragment size. */
 	__OVS_PACKET_ATTR_MAX
 };
 
@@ -319,6 +323,8 @@ enum ovs_key_attr {
 	OVS_KEY_ATTR_MPLS,      /* array of struct ovs_key_mpls.
 				 * The implementation may restrict
 				 * the accepted length of the array. */
+	OVS_KEY_ATTR_CT_STATE,	/* u8 bitmask of OVS_CS_F_* */
+	OVS_KEY_ATTR_CT_ZONE,	/* u16 connection tracking zone. */
 
 #ifdef __KERNEL__
 	OVS_KEY_ATTR_TUNNEL_INFO,  /* struct ip_tunnel_info */
@@ -431,6 +437,15 @@ struct ovs_key_nd {
 	__u8	nd_tll[ETH_ALEN];
 };
 
+/* OVS_KEY_ATTR_CT_STATE flags */
+#define OVS_CS_F_NEW               0x01 /* Beginning of a new connection. */
+#define OVS_CS_F_ESTABLISHED       0x02 /* Part of an existing connection. */
+#define OVS_CS_F_RELATED           0x04 /* Related to an established
+					 * connection. */
+#define OVS_CS_F_INVALID           0x20 /* Could not track connection. */
+#define OVS_CS_F_REPLY_DIR         0x40 /* Flow is in the reply direction. */
+#define OVS_CS_F_TRACKED           0x80 /* Conntrack has occurred. */
+
 /**
  * enum ovs_flow_attr - attributes for %OVS_FLOW_* commands.
  * @OVS_FLOW_ATTR_KEY: Nested %OVS_KEY_ATTR_* attributes specifying the flow
@@ -594,6 +609,28 @@ struct ovs_action_hash {
 	uint32_t  hash_basis;
 };
 
+/**
+ * enum ovs_ct_attr - Attributes for %OVS_ACTION_ATTR_CT action.
+ * @OVS_CT_ATTR_FLAGS: u32 connection tracking flags.
+ * @OVS_CT_ATTR_ZONE: u16 connection tracking zone.
+ */
+enum ovs_ct_attr {
+	OVS_CT_ATTR_UNSPEC,
+	OVS_CT_ATTR_FLAGS,      /* u8 bitmask of OVS_CT_F_*. */
+	OVS_CT_ATTR_ZONE,       /* u16 zone id. */
+	__OVS_CT_ATTR_MAX
+};
+
+#define OVS_CT_ATTR_MAX (__OVS_CT_ATTR_MAX - 1)
+
+/*
+ * OVS_CT_ATTR_FLAGS flags - bitmask of %OVS_CT_F_*
+ * @OVS_CT_F_COMMIT: Commits the flow to the conntrack table. This allows
+ * future packets for the same connection to be identified as 'established'
+ * or 'related'.
+ */
+#define OVS_CT_F_COMMIT		0x01
+
 /**
  * enum ovs_action_attr - Action types.
  *
@@ -623,6 +660,8 @@ struct ovs_action_hash {
  * indicate the new packet contents. This could potentially still be
  * %ETH_P_MPLS if the resulting MPLS label stack is not empty.  If there
  * is no MPLS label stack, as determined by ethertype, no action is taken.
+ * @OVS_ACTION_ATTR_CT: Track the connection. Populate the conntrack-related
+ * entries in the flow key.
  *
  * Only a single header can be set with a single %OVS_ACTION_ATTR_SET.  Not all
  * fields within a header are modifiable, e.g. the IPv4 protocol and fragment
@@ -648,6 +687,7 @@ enum ovs_action_attr {
 				       * data immediately followed by a mask.
 				       * The data must be zero for the unmasked
 				       * bits. */
+	OVS_ACTION_ATTR_CT,           /* One nested OVS_CT_ATTR_* . */
 
 	__OVS_ACTION_ATTR_MAX,	      /* Nothing past this will be accepted
 				       * from userspace. */

net/openvswitch/Kconfig

@@ -31,6 +31,17 @@ config OPENVSWITCH
 
 	  If unsure, say N.
 
+config OPENVSWITCH_CONNTRACK
+	bool "Open vSwitch conntrack action support"
+	depends on OPENVSWITCH
+	depends on NF_CONNTRACK
+	default OPENVSWITCH
+	---help---
+	  If you say Y here, then Open vSwitch module will be able to pass
+	  packets through conntrack.
+
+	  Say N to exclude this support and reduce the binary size.
+
 config OPENVSWITCH_GRE
 	tristate "Open vSwitch GRE tunneling support"
 	depends on OPENVSWITCH

net/openvswitch/Makefile

@@ -15,6 +15,8 @@ openvswitch-y := \
 	vport-internal_dev.o \
 	vport-netdev.o
 
+openvswitch-$(CONFIG_OPENVSWITCH_CONNTRACK) += conntrack.o
+
 obj-$(CONFIG_OPENVSWITCH_VXLAN)+= vport-vxlan.o
 obj-$(CONFIG_OPENVSWITCH_GENEVE)+= vport-geneve.o
 obj-$(CONFIG_OPENVSWITCH_GRE)	+= vport-gre.o

net/openvswitch/actions.c

@@ -22,13 +22,15 @@
 #include <linux/in.h>
 #include <linux/ip.h>
 #include <linux/openvswitch.h>
+#include <linux/netfilter_ipv6.h>
 #include <linux/sctp.h>
 #include <linux/tcp.h>
 #include <linux/udp.h>
 #include <linux/in6.h>
 #include <linux/if_arp.h>
 #include <linux/if_vlan.h>
 
+#include <net/dst.h>
 #include <net/ip.h>
 #include <net/ipv6.h>
 #include <net/checksum.h>
@@ -38,6 +40,7 @@
 
 #include "datapath.h"
 #include "flow.h"
+#include "conntrack.h"
 #include "vport.h"
 
 static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
@@ -52,6 +55,20 @@ struct deferred_action {
 	struct sw_flow_key pkt_key;
 };
 
+#define MAX_L2_LEN	(VLAN_ETH_HLEN + 3 * MPLS_HLEN)
+struct ovs_frag_data {
+	unsigned long dst;
+	struct vport *vport;
+	struct ovs_skb_cb cb;
+	__be16 inner_protocol;
+	__u16 vlan_tci;
+	__be16 vlan_proto;
+	unsigned int l2_len;
+	u8 l2_data[MAX_L2_LEN];
+};
+
+static DEFINE_PER_CPU(struct ovs_frag_data, ovs_frag_data_storage);
+
 #define DEFERRED_ACTION_FIFO_SIZE 10
 struct action_fifo {
 	int head;
@@ -602,14 +619,145 @@ static int set_sctp(struct sk_buff *skb, struct sw_flow_key *flow_key,
 	return 0;
 }
 
-static void do_output(struct datapath *dp, struct sk_buff *skb, int out_port)
+static int ovs_vport_output(struct sock *sock, struct sk_buff *skb)
+{
+	struct ovs_frag_data *data = this_cpu_ptr(&ovs_frag_data_storage);
+	struct vport *vport = data->vport;
+
+	if (skb_cow_head(skb, data->l2_len) < 0) {
+		kfree_skb(skb);
+		return -ENOMEM;
+	}
+
+	__skb_dst_copy(skb, data->dst);
+	*OVS_CB(skb) = data->cb;
+	skb->inner_protocol = data->inner_protocol;
+	skb->vlan_tci = data->vlan_tci;
+	skb->vlan_proto = data->vlan_proto;
+
+	/* Reconstruct the MAC header.  */
+	skb_push(skb, data->l2_len);
+	memcpy(skb->data, &data->l2_data, data->l2_len);
+	ovs_skb_postpush_rcsum(skb, skb->data, data->l2_len);
+	skb_reset_mac_header(skb);
+
+	ovs_vport_send(vport, skb);
+	return 0;
+}
+
+static unsigned int
+ovs_dst_get_mtu(const struct dst_entry *dst)
+{
+	return dst->dev->mtu;
+}
+
+static struct dst_ops ovs_dst_ops = {
+	.family = AF_UNSPEC,
+	.mtu = ovs_dst_get_mtu,
+};
+
+/* prepare_frag() is called once per (larger-than-MTU) frame; its inverse is
+ * ovs_vport_output(), which is called once per fragmented packet.
+ */
+static void prepare_frag(struct vport *vport, struct sk_buff *skb)
+{
+	unsigned int hlen = skb_network_offset(skb);
+	struct ovs_frag_data *data;
+
+	data = this_cpu_ptr(&ovs_frag_data_storage);
+	data->dst = skb->_skb_refdst;
+	data->vport = vport;
+	data->cb = *OVS_CB(skb);
+	data->inner_protocol = skb->inner_protocol;
+	data->vlan_tci = skb->vlan_tci;
+	data->vlan_proto = skb->vlan_proto;
+	data->l2_len = hlen;
+	memcpy(&data->l2_data, skb->data, hlen);
+
+	memset(IPCB(skb), 0, sizeof(struct inet_skb_parm));
+	skb_pull(skb, hlen);
+}
+
+static void ovs_fragment(struct vport *vport, struct sk_buff *skb, u16 mru,
+			 __be16 ethertype)
+{
+	if (skb_network_offset(skb) > MAX_L2_LEN) {
+		OVS_NLERR(1, "L2 header too long to fragment");
+		return;
+	}
+
+	if (ethertype == htons(ETH_P_IP)) {
+		struct dst_entry ovs_dst;
+		unsigned long orig_dst;
+
+		prepare_frag(vport, skb);
+		dst_init(&ovs_dst, &ovs_dst_ops, NULL, 1,
+			 DST_OBSOLETE_NONE, DST_NOCOUNT);
+		ovs_dst.dev = vport->dev;
+
+		orig_dst = skb->_skb_refdst;
+		skb_dst_set_noref(skb, &ovs_dst);
+		IPCB(skb)->frag_max_size = mru;
+
+		ip_do_fragment(skb->sk, skb, ovs_vport_output);
+		refdst_drop(orig_dst);
+	} else if (ethertype == htons(ETH_P_IPV6)) {
+		const struct nf_ipv6_ops *v6ops = nf_get_ipv6_ops();
+		unsigned long orig_dst;
+		struct rt6_info ovs_rt;
+
+		if (!v6ops) {
+			kfree_skb(skb);
+			return;
+		}
+
+		prepare_frag(vport, skb);
+		memset(&ovs_rt, 0, sizeof(ovs_rt));
+		dst_init(&ovs_rt.dst, &ovs_dst_ops, NULL, 1,
+			 DST_OBSOLETE_NONE, DST_NOCOUNT);
+		ovs_rt.dst.dev = vport->dev;
+
+		orig_dst = skb->_skb_refdst;
+		skb_dst_set_noref(skb, &ovs_rt.dst);
+		IP6CB(skb)->frag_max_size = mru;
+
+		v6ops->fragment(skb->sk, skb, ovs_vport_output);
+		refdst_drop(orig_dst);
+	} else {
+		WARN_ONCE(1, "Failed fragment ->%s: eth=%04x, MRU=%d, MTU=%d.",
+			  ovs_vport_name(vport), ntohs(ethertype), mru,
+			  vport->dev->mtu);
+		kfree_skb(skb);
+	}
+}
+
+static void do_output(struct datapath *dp, struct sk_buff *skb, int out_port,
+		      struct sw_flow_key *key)
 {
 	struct vport *vport = ovs_vport_rcu(dp, out_port);
 
-	if (likely(vport))
-		ovs_vport_send(vport, skb);
-	else
+	if (likely(vport)) {
+		u16 mru = OVS_CB(skb)->mru;
+
+		if (likely(!mru || (skb->len <= mru + ETH_HLEN))) {
+			ovs_vport_send(vport, skb);
+		} else if (mru <= vport->dev->mtu) {
+			__be16 ethertype = key->eth.type;
+
+			if (!is_flow_key_valid(key)) {
+				if (eth_p_mpls(skb->protocol))
+					ethertype = skb->inner_protocol;
+				else
+					ethertype = vlan_get_protocol(skb);
+			}
+
+			ovs_fragment(vport, skb, mru, ethertype);
+		} else {
+			kfree_skb(skb);
+		}
+	} else {
 		kfree_skb(skb);
+	}
 }
 
 static int output_userspace(struct datapath *dp, struct sk_buff *skb,
@@ -623,6 +771,7 @@ static int output_userspace(struct datapath *dp, struct sk_buff *skb,
 
 	memset(&upcall, 0, sizeof(upcall));
 	upcall.cmd = OVS_PACKET_CMD_ACTION;
+	upcall.mru = OVS_CB(skb)->mru;
 
 	for (a = nla_data(attr), rem = nla_len(attr); rem > 0;
 		 a = nla_next(a, &rem)) {
@@ -816,6 +965,11 @@ static int execute_masked_set_action(struct sk_buff *skb,
 		err = set_mpls(skb, flow_key, nla_data(a), get_mask(a,
 								    __be32 *));
 		break;
+
+	case OVS_KEY_ATTR_CT_STATE:
+	case OVS_KEY_ATTR_CT_ZONE:
+		err = -EINVAL;
+		break;
 	}
 
 	return err;
@@ -885,7 +1039,7 @@ static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
 			struct sk_buff *out_skb = skb_clone(skb, GFP_ATOMIC);
 
 			if (out_skb)
-				do_output(dp, out_skb, prev_port);
+				do_output(dp, out_skb, prev_port, key);
 
 			prev_port = -1;
 		}
@@ -942,6 +1096,15 @@ static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
 		case OVS_ACTION_ATTR_SAMPLE:
 			err = sample(dp, skb, key, a, attr, len);
 			break;
+
+		case OVS_ACTION_ATTR_CT:
+			err = ovs_ct_execute(ovs_dp_get_net(dp), skb, key,
+					     nla_data(a));
+
+			/* Hide stolen IP fragments from user space. */
+			if (err == -EINPROGRESS)
+				return 0;
+			break;
 		}
 
 		if (unlikely(err)) {
@@ -951,7 +1114,7 @@ static int do_execute_actions(struct datapath *dp, struct sk_buff *skb,
 	}
 
 	if (prev_port != -1)
-		do_output(dp, skb, prev_port);
+		do_output(dp, skb, prev_port, key);
 	else
 		consume_skb(skb);