openvswitch: Add vxlan tunneling support.

Following patch adds vxlan vport type for openvswitch using vxlan api. So now there is vxlan dependency for openvswitch. CC: Jesse Gross <jesse@nicira.com> Signed-off-by: Pravin B Shelar <pshelar@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com> Signed-off-by: David S. Miller <davem@davemloft.net>

openvswitch: Add vxlan tunneling support.
Following patch adds vxlan vport type for openvswitch using vxlan api. So now there is vxlan dependency for openvswitch. CC: Jesse Gross <jesse@nicira.com> Signed-off-by: Pravin B Shelar <pshelar@nicira.com> Acked-by: Jesse Gross <jesse@nicira.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Pravin B Shelar · David S. Miller
1 parent 1eaa81785a
Showing 6 changed files with 236 additions and 0 deletions Side-by-side Diff
include/uapi/linux/openvswitch.h
net/openvswitch/Kconfig
net/openvswitch/Makefile
net/openvswitch/vport-vxlan.c
net/openvswitch/vport.c
net/openvswitch/vport.h
@@ -165,6 +165,7 @@
 	OVS_VPORT_TYPE_NETDEV,   /* network device */
 	OVS_VPORT_TYPE_INTERNAL, /* network device implemented by datapath */
 	OVS_VPORT_TYPE_GRE,      /* GRE tunnel. */
+	OVS_VPORT_TYPE_VXLAN,	 /* VXLAN tunnel. */
 	__OVS_VPORT_TYPE_MAX
 };
  
@@ -210,6 +211,16 @@
 };
  
 #define OVS_VPORT_ATTR_MAX (__OVS_VPORT_ATTR_MAX - 1)
+
+/* OVS_VPORT_ATTR_OPTIONS attributes for tunnels.
+ */
+enum {
+	OVS_TUNNEL_ATTR_UNSPEC,
+	OVS_TUNNEL_ATTR_DST_PORT, /* 16-bit UDP port, used by L4 tunnels. */
+	__OVS_TUNNEL_ATTR_MAX
+};
+
+#define OVS_TUNNEL_ATTR_MAX (__OVS_TUNNEL_ATTR_MAX - 1)
  
 /* Flows. */
  
@@ -40,4 +40,17 @@
 	  Say N to exclude this support and reduce the binary size.
  
 	  If unsure, say Y.
+
+config OPENVSWITCH_VXLAN
+	bool "Open vSwitch VXLAN tunneling support"
+	depends on INET
+	depends on OPENVSWITCH
+	depends on VXLAN && !(OPENVSWITCH=y && VXLAN=m)
+	default y
+	---help---
+	  If you say Y here, then the Open vSwitch will be able create vxlan vport.
+
+	  Say N to exclude this support and reduce the binary size.
+
+	  If unsure, say Y.
@@ -13,4 +13,8 @@
 	vport-gre.o \
 	vport-internal_dev.o \
 	vport-netdev.o
+
+ifneq ($(CONFIG_OPENVSWITCH_VXLAN),)
+openvswitch-y += vport-vxlan.o
+endif
+/*
+ * Copyright (c) 2013 Nicira, Inc.
+ * Copyright (c) 2013 Cisco Systems, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/in.h>
+#include <linux/ip.h>
+#include <linux/net.h>
+#include <linux/rculist.h>
+#include <linux/udp.h>
+
+#include <net/icmp.h>
+#include <net/ip.h>
+#include <net/udp.h>
+#include <net/ip_tunnels.h>
+#include <net/udp.h>
+#include <net/rtnetlink.h>
+#include <net/route.h>
+#include <net/dsfield.h>
+#include <net/inet_ecn.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
+#include <net/vxlan.h>
+
+#include "datapath.h"
+#include "vport.h"
+
+/**
+ * struct vxlan_port - Keeps track of open UDP ports
+ * @vs: vxlan_sock created for the port.
+ * @name: vport name.
+ */
+struct vxlan_port {
+	struct vxlan_sock *vs;
+	char name[IFNAMSIZ];
+};
+
+static inline struct vxlan_port *vxlan_vport(const struct vport *vport)
+{
+	return vport_priv(vport);
+}
+
+/* Called with rcu_read_lock and BH disabled. */
+static void vxlan_rcv(struct vxlan_sock *vs, struct sk_buff *skb, __be32 vx_vni)
+{
+	struct ovs_key_ipv4_tunnel tun_key;
+	struct vport *vport = vs->data;
+	struct iphdr *iph;
+	__be64 key;
+
+	/* Save outer tunnel values */
+	iph = ip_hdr(skb);
+	key = cpu_to_be64(ntohl(vx_vni) >> 8);
+	ovs_flow_tun_key_init(&tun_key, iph, key, TUNNEL_KEY);
+
+	ovs_vport_receive(vport, skb, &tun_key);
+}
+
+static int vxlan_get_options(const struct vport *vport, struct sk_buff *skb)
+{
+	struct vxlan_port *vxlan_port = vxlan_vport(vport);
+	__be16 dst_port = inet_sk(vxlan_port->vs->sock->sk)->inet_sport;
+
+	if (nla_put_u16(skb, OVS_TUNNEL_ATTR_DST_PORT, ntohs(dst_port)))
+		return -EMSGSIZE;
+	return 0;
+}
+
+static void vxlan_tnl_destroy(struct vport *vport)
+{
+	struct vxlan_port *vxlan_port = vxlan_vport(vport);
+
+	vxlan_sock_release(vxlan_port->vs);
+
+	ovs_vport_deferred_free(vport);
+}
+
+static struct vport *vxlan_tnl_create(const struct vport_parms *parms)
+{
+	struct net *net = ovs_dp_get_net(parms->dp);
+	struct nlattr *options = parms->options;
+	struct vxlan_port *vxlan_port;
+	struct vxlan_sock *vs;
+	struct vport *vport;
+	struct nlattr *a;
+	u16 dst_port;
+	int err;
+
+	if (!options) {
+		err = -EINVAL;
+		goto error;
+	}
+	a = nla_find_nested(options, OVS_TUNNEL_ATTR_DST_PORT);
+	if (a && nla_len(a) == sizeof(u16)) {
+		dst_port = nla_get_u16(a);
+	} else {
+		/* Require destination port from userspace. */
+		err = -EINVAL;
+		goto error;
+	}
+
+	vport = ovs_vport_alloc(sizeof(struct vxlan_port),
+				&ovs_vxlan_vport_ops, parms);
+	if (IS_ERR(vport))
+		return vport;
+
+	vxlan_port = vxlan_vport(vport);
+	strncpy(vxlan_port->name, parms->name, IFNAMSIZ);
+
+	vs = vxlan_sock_add(net, htons(dst_port), vxlan_rcv, vport, true);
+	if (IS_ERR(vs)) {
+		ovs_vport_free(vport);
+		return (void *)vs;
+	}
+	vxlan_port->vs = vs;
+
+	return vport;
+
+error:
+	return ERR_PTR(err);
+}
+
+static int vxlan_tnl_send(struct vport *vport, struct sk_buff *skb)
+{
+	struct net *net = ovs_dp_get_net(vport->dp);
+	struct vxlan_port *vxlan_port = vxlan_vport(vport);
+	__be16 dst_port = inet_sk(vxlan_port->vs->sock->sk)->inet_sport;
+	struct rtable *rt;
+	struct flowi4 fl;
+	__be16 src_port;
+	int port_min;
+	int port_max;
+	__be16 df;
+	int err;
+
+	if (unlikely(!OVS_CB(skb)->tun_key)) {
+		err = -EINVAL;
+		goto error;
+	}
+
+	/* Route lookup */
+	memset(&fl, 0, sizeof(fl));
+	fl.daddr = OVS_CB(skb)->tun_key->ipv4_dst;
+	fl.saddr = OVS_CB(skb)->tun_key->ipv4_src;
+	fl.flowi4_tos = RT_TOS(OVS_CB(skb)->tun_key->ipv4_tos);
+	fl.flowi4_mark = skb->mark;
+	fl.flowi4_proto = IPPROTO_UDP;
+
+	rt = ip_route_output_key(net, &fl);
+	if (IS_ERR(rt)) {
+		err = PTR_ERR(rt);
+		goto error;
+	}
+
+	df = OVS_CB(skb)->tun_key->tun_flags & TUNNEL_DONT_FRAGMENT ?
+		htons(IP_DF) : 0;
+
+	skb->local_df = 1;
+
+	inet_get_local_port_range(&port_min, &port_max);
+	src_port = vxlan_src_port(port_min, port_max, skb);
+
+	err = vxlan_xmit_skb(net, vxlan_port->vs, rt, skb,
+			     fl.saddr, OVS_CB(skb)->tun_key->ipv4_dst,
+			     OVS_CB(skb)->tun_key->ipv4_tos,
+			     OVS_CB(skb)->tun_key->ipv4_ttl, df,
+			     src_port, dst_port,
+			     htonl(be64_to_cpu(OVS_CB(skb)->tun_key->tun_id) << 8));
+	if (err < 0)
+		ip_rt_put(rt);
+error:
+	return err;
+}
+
+static const char *vxlan_get_name(const struct vport *vport)
+{
+	struct vxlan_port *vxlan_port = vxlan_vport(vport);
+	return vxlan_port->name;
+}
+
+const struct vport_ops ovs_vxlan_vport_ops = {
+	.type		= OVS_VPORT_TYPE_VXLAN,
+	.create		= vxlan_tnl_create,
+	.destroy	= vxlan_tnl_destroy,
+	.get_name	= vxlan_get_name,
+	.get_options	= vxlan_get_options,
+	.send		= vxlan_tnl_send,
+};
@@ -42,6 +42,9 @@
 #ifdef CONFIG_OPENVSWITCH_GRE
 	&ovs_gre_vport_ops,
 #endif
+#ifdef CONFIG_OPENVSWITCH_VXLAN
+	&ovs_vxlan_vport_ops,
+#endif
 };
  
 /* Protected by RCU read lock for reading, ovs_mutex for writing. */
@@ -199,6 +199,7 @@
 extern const struct vport_ops ovs_netdev_vport_ops;
 extern const struct vport_ops ovs_internal_vport_ops;
 extern const struct vport_ops ovs_gre_vport_ops;
+extern const struct vport_ops ovs_vxlan_vport_ops;
  
 static inline void ovs_skb_postpush_rcsum(struct sk_buff *skb,
 				      const void *start, unsigned int len)
...	...	@@ -165,6 +165,7 @@
165	165	OVS_VPORT_TYPE_NETDEV, /* network device */
166	166	OVS_VPORT_TYPE_INTERNAL, /* network device implemented by datapath */
167	167	OVS_VPORT_TYPE_GRE, /* GRE tunnel. */
	168	+ OVS_VPORT_TYPE_VXLAN, /* VXLAN tunnel. */
168	169	__OVS_VPORT_TYPE_MAX
169	170	};
170	171
...	...	@@ -210,6 +211,16 @@
210	211	};
211	212
212	213	#define OVS_VPORT_ATTR_MAX (__OVS_VPORT_ATTR_MAX - 1)
	214	+
	215	+/* OVS_VPORT_ATTR_OPTIONS attributes for tunnels.
	216	+ */
	217	+enum {
	218	+ OVS_TUNNEL_ATTR_UNSPEC,
	219	+ OVS_TUNNEL_ATTR_DST_PORT, /* 16-bit UDP port, used by L4 tunnels. */
	220	+ __OVS_TUNNEL_ATTR_MAX
	221	+};
	222	+
	223	+#define OVS_TUNNEL_ATTR_MAX (__OVS_TUNNEL_ATTR_MAX - 1)
213	224
214	225	/* Flows. */
215	226
...	...	@@ -40,4 +40,17 @@
40	40	Say N to exclude this support and reduce the binary size.
41	41
42	42	If unsure, say Y.
	43	+
	44	+config OPENVSWITCH_VXLAN
	45	+ bool "Open vSwitch VXLAN tunneling support"
	46	+ depends on INET
	47	+ depends on OPENVSWITCH
	48	+ depends on VXLAN && !(OPENVSWITCH=y && VXLAN=m)
	49	+ default y
	50	+ ---help---
	51	+ If you say Y here, then the Open vSwitch will be able create vxlan vport.
	52	+
	53	+ Say N to exclude this support and reduce the binary size.
	54	+
	55	+ If unsure, say Y.
...	...	@@ -13,4 +13,8 @@
13	13	vport-gre.o \
14	14	vport-internal_dev.o \
15	15	vport-netdev.o
	16	+
	17	+ifneq ($(CONFIG_OPENVSWITCH_VXLAN),)
	18	+openvswitch-y += vport-vxlan.o
	19	+endif
	1	+/*
	2	+ * Copyright (c) 2013 Nicira, Inc.
	3	+ * Copyright (c) 2013 Cisco Systems, Inc.
	4	+ *
	5	+ * This program is free software; you can redistribute it and/or
	6	+ * modify it under the terms of version 2 of the GNU General Public
	7	+ * License as published by the Free Software Foundation.
	8	+ *
	9	+ * This program is distributed in the hope that it will be useful, but
	10	+ * WITHOUT ANY WARRANTY; without even the implied warranty of
	11	+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	12	+ * General Public License for more details.
	13	+ *
	14	+ * You should have received a copy of the GNU General Public License
	15	+ * along with this program; if not, write to the Free Software
	16	+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
	17	+ * 02110-1301, USA
	18	+ */
	19	+
	20	+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
	21	+
	22	+#include <linux/in.h>
	23	+#include <linux/ip.h>
	24	+#include <linux/net.h>
	25	+#include <linux/rculist.h>
	26	+#include <linux/udp.h>
	27	+
	28	+#include <net/icmp.h>
	29	+#include <net/ip.h>
	30	+#include <net/udp.h>
	31	+#include <net/ip_tunnels.h>
	32	+#include <net/udp.h>
	33	+#include <net/rtnetlink.h>
	34	+#include <net/route.h>
	35	+#include <net/dsfield.h>
	36	+#include <net/inet_ecn.h>
	37	+#include <net/net_namespace.h>
	38	+#include <net/netns/generic.h>
	39	+#include <net/vxlan.h>
	40	+
	41	+#include "datapath.h"
	42	+#include "vport.h"
	43	+
	44	+/**
	45	+ * struct vxlan_port - Keeps track of open UDP ports
	46	+ * @vs: vxlan_sock created for the port.
	47	+ * @name: vport name.
	48	+ */
	49	+struct vxlan_port {
	50	+ struct vxlan_sock *vs;
	51	+ char name[IFNAMSIZ];
	52	+};
	53	+
	54	+static inline struct vxlan_port vxlan_vport(const struct vport vport)
	55	+{
	56	+ return vport_priv(vport);
	57	+}
	58	+
	59	+/* Called with rcu_read_lock and BH disabled. */
	60	+static void vxlan_rcv(struct vxlan_sock vs, struct sk_buff skb, __be32 vx_vni)
	61	+{
	62	+ struct ovs_key_ipv4_tunnel tun_key;
	63	+ struct vport *vport = vs->data;
	64	+ struct iphdr *iph;
	65	+ __be64 key;
	66	+
	67	+ /* Save outer tunnel values */
	68	+ iph = ip_hdr(skb);
	69	+ key = cpu_to_be64(ntohl(vx_vni) >> 8);
	70	+ ovs_flow_tun_key_init(&tun_key, iph, key, TUNNEL_KEY);
	71	+
	72	+ ovs_vport_receive(vport, skb, &tun_key);
	73	+}
	74	+
	75	+static int vxlan_get_options(const struct vport vport, struct sk_buff skb)
	76	+{
	77	+ struct vxlan_port *vxlan_port = vxlan_vport(vport);
	78	+ __be16 dst_port = inet_sk(vxlan_port->vs->sock->sk)->inet_sport;
	79	+
	80	+ if (nla_put_u16(skb, OVS_TUNNEL_ATTR_DST_PORT, ntohs(dst_port)))
	81	+ return -EMSGSIZE;
	82	+ return 0;
	83	+}
	84	+
	85	+static void vxlan_tnl_destroy(struct vport *vport)
	86	+{
	87	+ struct vxlan_port *vxlan_port = vxlan_vport(vport);
	88	+
	89	+ vxlan_sock_release(vxlan_port->vs);
	90	+
	91	+ ovs_vport_deferred_free(vport);
	92	+}
	93	+
	94	+static struct vport vxlan_tnl_create(const struct vport_parms parms)
	95	+{
	96	+ struct net *net = ovs_dp_get_net(parms->dp);
	97	+ struct nlattr *options = parms->options;
	98	+ struct vxlan_port *vxlan_port;
	99	+ struct vxlan_sock *vs;
	100	+ struct vport *vport;
	101	+ struct nlattr *a;
	102	+ u16 dst_port;
	103	+ int err;
	104	+
	105	+ if (!options) {
	106	+ err = -EINVAL;
	107	+ goto error;
	108	+ }
	109	+ a = nla_find_nested(options, OVS_TUNNEL_ATTR_DST_PORT);
	110	+ if (a && nla_len(a) == sizeof(u16)) {
	111	+ dst_port = nla_get_u16(a);
	112	+ } else {
	113	+ /* Require destination port from userspace. */
	114	+ err = -EINVAL;
	115	+ goto error;
	116	+ }
	117	+
	118	+ vport = ovs_vport_alloc(sizeof(struct vxlan_port),
	119	+ &ovs_vxlan_vport_ops, parms);
	120	+ if (IS_ERR(vport))
	121	+ return vport;
	122	+
	123	+ vxlan_port = vxlan_vport(vport);
	124	+ strncpy(vxlan_port->name, parms->name, IFNAMSIZ);
	125	+
	126	+ vs = vxlan_sock_add(net, htons(dst_port), vxlan_rcv, vport, true);
	127	+ if (IS_ERR(vs)) {
	128	+ ovs_vport_free(vport);
	129	+ return (void *)vs;
	130	+ }
	131	+ vxlan_port->vs = vs;
	132	+
	133	+ return vport;
	134	+
	135	+error:
	136	+ return ERR_PTR(err);
	137	+}
	138	+
	139	+static int vxlan_tnl_send(struct vport vport, struct sk_buff skb)
	140	+{
	141	+ struct net *net = ovs_dp_get_net(vport->dp);
	142	+ struct vxlan_port *vxlan_port = vxlan_vport(vport);
	143	+ __be16 dst_port = inet_sk(vxlan_port->vs->sock->sk)->inet_sport;
	144	+ struct rtable *rt;
	145	+ struct flowi4 fl;
	146	+ __be16 src_port;
	147	+ int port_min;
	148	+ int port_max;
	149	+ __be16 df;
	150	+ int err;
	151	+
	152	+ if (unlikely(!OVS_CB(skb)->tun_key)) {
	153	+ err = -EINVAL;
	154	+ goto error;
	155	+ }
	156	+
	157	+ /* Route lookup */
	158	+ memset(&fl, 0, sizeof(fl));
	159	+ fl.daddr = OVS_CB(skb)->tun_key->ipv4_dst;
	160	+ fl.saddr = OVS_CB(skb)->tun_key->ipv4_src;
	161	+ fl.flowi4_tos = RT_TOS(OVS_CB(skb)->tun_key->ipv4_tos);
	162	+ fl.flowi4_mark = skb->mark;
	163	+ fl.flowi4_proto = IPPROTO_UDP;
	164	+
	165	+ rt = ip_route_output_key(net, &fl);
	166	+ if (IS_ERR(rt)) {
	167	+ err = PTR_ERR(rt);
	168	+ goto error;
	169	+ }
	170	+
	171	+ df = OVS_CB(skb)->tun_key->tun_flags & TUNNEL_DONT_FRAGMENT ?
	172	+ htons(IP_DF) : 0;
	173	+
	174	+ skb->local_df = 1;
	175	+
	176	+ inet_get_local_port_range(&port_min, &port_max);
	177	+ src_port = vxlan_src_port(port_min, port_max, skb);
	178	+
	179	+ err = vxlan_xmit_skb(net, vxlan_port->vs, rt, skb,
	180	+ fl.saddr, OVS_CB(skb)->tun_key->ipv4_dst,
	181	+ OVS_CB(skb)->tun_key->ipv4_tos,
	182	+ OVS_CB(skb)->tun_key->ipv4_ttl, df,
	183	+ src_port, dst_port,
	184	+ htonl(be64_to_cpu(OVS_CB(skb)->tun_key->tun_id) << 8));
	185	+ if (err < 0)
	186	+ ip_rt_put(rt);
	187	+error:
	188	+ return err;
	189	+}
	190	+
	191	+static const char vxlan_get_name(const struct vport vport)
	192	+{
	193	+ struct vxlan_port *vxlan_port = vxlan_vport(vport);
	194	+ return vxlan_port->name;
	195	+}
	196	+
	197	+const struct vport_ops ovs_vxlan_vport_ops = {
	198	+ .type = OVS_VPORT_TYPE_VXLAN,
	199	+ .create = vxlan_tnl_create,
	200	+ .destroy = vxlan_tnl_destroy,
	201	+ .get_name = vxlan_get_name,
	202	+ .get_options = vxlan_get_options,
	203	+ .send = vxlan_tnl_send,
	204	+};
...	...	@@ -42,6 +42,9 @@
42	42	#ifdef CONFIG_OPENVSWITCH_GRE
43	43	&ovs_gre_vport_ops,
44	44	#endif
	45	+#ifdef CONFIG_OPENVSWITCH_VXLAN
	46	+ &ovs_vxlan_vport_ops,
	47	+#endif
45	48	};
46	49
47	50	/* Protected by RCU read lock for reading, ovs_mutex for writing. */
...	...	@@ -199,6 +199,7 @@
199	199	extern const struct vport_ops ovs_netdev_vport_ops;
200	200	extern const struct vport_ops ovs_internal_vport_ops;
201	201	extern const struct vport_ops ovs_gre_vport_ops;
	202	+extern const struct vport_ops ovs_vxlan_vport_ops;
202	203
203	204	static inline void ovs_skb_postpush_rcsum(struct sk_buff *skb,
204	205	const void *start, unsigned int len)