* This code is derived from kernel vxlan module.
*/
+#include <linux/version.h>
+#if LINUX_VERSION_CODE < KERNEL_VERSION(3,12,0)
+
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/kernel.h>
#include <net/netns/generic.h>
#include <net/vxlan.h>
-#include "checksum.h"
#include "compat.h"
+#include "datapath.h"
#include "gso.h"
#include "vlan.h"
-#define PORT_HASH_BITS 8
-#define PORT_HASH_SIZE (1<<PORT_HASH_BITS)
-
-#define VXLAN_N_VID (1u << 24)
-#define VXLAN_VID_MASK (VXLAN_N_VID - 1)
-/* IP header + UDP + VXLAN + Ethernet header */
-#define VXLAN_HEADROOM (20 + 8 + 8 + 14)
#define VXLAN_HLEN (sizeof(struct udphdr) + sizeof(struct vxlanhdr))
#define VXLAN_FLAGS 0x08000000 /* struct vxlanhdr.vx_flags required value. */
__be32 vx_vni;
};
-static int vxlan_net_id;
-
-/* per-network namespace private data for this module */
-struct vxlan_net {
- struct hlist_head sock_list[PORT_HASH_SIZE];
- struct mutex sock_lock; /* RTNL lock nests inside this lock. */
-};
-
-/* Socket hash table head */
-static inline struct hlist_head *vs_head(struct net *net, __be16 port)
-{
- struct vxlan_net *vn = net_generic(net, vxlan_net_id);
-
- return &vn->sock_list[hash_32(ntohs(port), PORT_HASH_BITS)];
-}
-
-/* Find VXLAN socket based on network namespace and UDP port */
-static struct vxlan_sock *vxlan_find_port(struct net *net, __be16 port)
-{
- struct vxlan_sock *vs;
-
- hlist_for_each_entry_rcu(vs, vs_head(net, port), hlist) {
- if (inet_sport(vs->sock->sk) == port)
- return vs;
- }
- return NULL;
-}
-
/* Callback from net/ipv4/udp.c to receive packets */
static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
{
- struct vxlan_handler *vh;
struct vxlan_sock *vs;
struct vxlanhdr *vxh;
if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB)))
goto drop;
- vs = vxlan_find_port(sock_net(sk), inet_sport(sk));
+ vs = rcu_dereference_sk_user_data(sk);
if (!vs)
goto drop;
- list_for_each_entry_rcu(vh, &vs->handler_list, node) {
- if (vh->rcv(vh, skb, vxh->vx_vni) == PACKET_RCVD)
- return 0;
- }
+ vs->rcv(vs, skb, vxh->vx_vni);
+ return 0;
drop:
/* Consume bad packet */
unsigned int range = (port_max - port_min) + 1;
u32 hash;
- hash = skb_get_rxhash(skb);
+ hash = skb_get_hash(skb);
if (!hash)
hash = jhash(skb->data, 2 * ETH_ALEN,
(__force u32) skb->protocol);
return 0;
}
-int vxlan_xmit_skb(struct net *net, struct vxlan_handler *vh,
+int vxlan_xmit_skb(struct vxlan_sock *vs,
struct rtable *rt, struct sk_buff *skb,
__be32 src, __be32 dst, __u8 tos, __u8 ttl, __be16 df,
__be16 src_port, __be16 dst_port, __be32 vni)
int min_headroom;
int err;
- skb_reset_inner_headers(skb);
-
min_headroom = LL_RESERVED_SPACE(rt_dst(rt).dev) + rt_dst(rt).header_len
+ VXLAN_HLEN + sizeof(struct iphdr)
+ (vlan_tx_tag_present(skb) ? VLAN_HLEN : 0);
if (unlikely(err))
return err;
- if (unlikely(vlan_deaccel_tag(skb)))
- return -ENOMEM;
+ if (vlan_tx_tag_present(skb)) {
+ if (unlikely(!__vlan_put_tag(skb,
+ skb->vlan_proto,
+ vlan_tx_tag_get(skb))))
+ return -ENOMEM;
+
+ vlan_set_tci(skb, 0);
+ }
+
+ skb_reset_inner_headers(skb);
vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh));
vxh->vx_flags = htonl(VXLAN_FLAGS);
uh->len = htons(skb->len);
uh->check = 0;
- vxlan_set_owner(vh->vs->sock->sk, skb);
+ vxlan_set_owner(vs->sock->sk, skb);
err = handle_offloads(skb);
if (err)
return err;
- return iptunnel_xmit(net, rt, skb, src, dst,
- IPPROTO_UDP, tos, ttl, df);
+ return iptunnel_xmit(rt, skb, src, dst, IPPROTO_UDP, tos, ttl, df, false);
}
-static struct vxlan_sock *vxlan_socket_create(struct net *net, __be16 port)
+static void rcu_free_vs(struct rcu_head *rcu)
+{
+ struct vxlan_sock *vs = container_of(rcu, struct vxlan_sock, rcu);
+
+ kfree(vs);
+}
+
+static void vxlan_del_work(struct work_struct *work)
+{
+ struct vxlan_sock *vs = container_of(work, struct vxlan_sock, del_work);
+
+ sk_release_kernel(vs->sock->sk);
+ call_rcu(&vs->rcu, rcu_free_vs);
+}
+
+static struct vxlan_sock *vxlan_socket_create(struct net *net, __be16 port,
+ vxlan_rcv_t *rcv, void *data)
{
struct vxlan_sock *vs;
struct sock *sk;
int rc;
vs = kmalloc(sizeof(*vs), GFP_KERNEL);
- if (!vs)
+ if (!vs) {
+ pr_debug("memory alocation failure\n");
return ERR_PTR(-ENOMEM);
+ }
+
+ INIT_WORK(&vs->del_work, vxlan_del_work);
/* Create UDP socket for encapsulation receive. */
rc = sock_create_kern(AF_INET, SOCK_DGRAM, IPPROTO_UDP, &vs->sock);
sk_change_net(sk, net);
rc = kernel_bind(vs->sock, (struct sockaddr *) &vxlan_addr,
- sizeof(vxlan_addr));
+ sizeof(vxlan_addr));
if (rc < 0) {
pr_debug("bind for UDP socket %pI4:%u (%d)\n",
- &vxlan_addr.sin_addr, ntohs(vxlan_addr.sin_port), rc);
+ &vxlan_addr.sin_addr, ntohs(vxlan_addr.sin_port), rc);
sk_release_kernel(sk);
kfree(vs);
return ERR_PTR(rc);
}
+ vs->rcv = rcv;
+ vs->data = data;
/* Disable multicast loopback */
inet_sk(sk)->mc_loop = 0;
- INIT_LIST_HEAD(&vs->handler_list);
- hlist_add_head_rcu(&vs->hlist, vs_head(net, port));
+ rcu_assign_sk_user_data(vs->sock->sk, vs);
/* Mark socket as an encapsulation socket. */
udp_sk(sk)->encap_type = 1;
udp_sk(sk)->encap_rcv = vxlan_udp_encap_recv;
udp_encap_enable();
-
return vs;
}
-static void rcu_free_vs_callback(struct rcu_head *rcu)
-{
- struct vxlan_sock *vs = container_of(rcu, struct vxlan_sock, rcu);
-
- kfree(vs);
-}
-
-static void vxlan_socket_del(struct vxlan_sock *vs)
-{
- if (list_empty(&vs->handler_list)) {
- hlist_del_rcu(&vs->hlist);
-
- sk_release_kernel(vs->sock->sk);
- call_rcu(&vs->rcu, rcu_free_vs_callback);
- }
-}
-
-static int vxlan_init_module(void);
-static void vxlan_cleanup_module(void);
-
-static void rcu_free_vh_callback(struct rcu_head *rcu)
+struct vxlan_sock *vxlan_sock_add(struct net *net, __be16 port,
+ vxlan_rcv_t *rcv, void *data,
+ bool no_share, bool ipv6)
{
- struct vxlan_handler *vh = container_of(rcu, struct vxlan_handler, rcu);
-
- kfree(vh);
+ return vxlan_socket_create(net, port, rcv, data);
}
-static void vh_del_work(struct work_struct *work)
+void vxlan_sock_release(struct vxlan_sock *vs)
{
- struct vxlan_handler *vh = container_of(work, struct vxlan_handler, del_work);
- struct vxlan_sock *vs = vh->vs;
- struct net *net = sock_net(vs->sock->sk);
- struct vxlan_net *vn = net_generic(net, vxlan_net_id);
+ ASSERT_OVSL();
+ rcu_assign_sk_user_data(vs->sock->sk, NULL);
- mutex_lock(&vn->sock_lock);
-
- list_del_rcu(&vh->node);
- call_rcu(&vh->rcu, rcu_free_vh_callback);
- vxlan_socket_del(vs);
-
- mutex_unlock(&vn->sock_lock);
-
- vxlan_cleanup_module();
+ queue_work(system_wq, &vs->del_work);
}
-struct vxlan_handler *vxlan_handler_add(struct net *net,
- __be16 portno, vxlan_rcv_t *rcv,
- void *data, int priority, bool create)
-{
- struct vxlan_net *vn;
- struct vxlan_sock *vs;
- struct vxlan_handler *vh;
- struct vxlan_handler *new;
- int err;
-
- err = vxlan_init_module();
- if (err)
- return ERR_PTR(err);
-
- vn = net_generic(net, vxlan_net_id);
- mutex_lock(&vn->sock_lock);
- /* Look to see if can reuse socket */
- vs = vxlan_find_port(net, portno);
- if (!vs) {
- vs = vxlan_socket_create(net, portno);
- if (IS_ERR(vs)) {
- new = (void *) vs;
- goto out;
- }
- }
-
- /* Try existing vxlan hanlders for this socket. */
- list_for_each_entry(vh, &vs->handler_list, node) {
- if (vh->rcv == rcv) {
- if (create) {
- vxlan_socket_del(vs);
- new = ERR_PTR(-EEXIST);
- goto out;
- }
- atomic_inc(&vh->refcnt);
- new = vh;
- goto out;
- }
- }
-
- new = kzalloc(sizeof(*new), GFP_KERNEL);
- if (!new) {
- vxlan_socket_del(vs);
- new = ERR_PTR(-ENOMEM);
- goto out;
- }
-
- new->rcv = rcv;
- new->vs = vs;
- atomic_set(&new->refcnt, 1);
- INIT_WORK(&new->del_work, vh_del_work);
- new->data = data;
- new->priority = priority;
-
- list_for_each_entry(vh, &vs->handler_list, node) {
- if (vh->priority > priority) {
- list_add_tail_rcu(&new->node, &vh->node);
- goto out;
- }
- }
-
- list_add_tail_rcu(&new->node, &vs->handler_list);
-out:
- mutex_unlock(&vn->sock_lock);
- return new;
-}
-
-void vxlan_handler_put(struct vxlan_handler *vh)
-{
- BUG_ON(!vh->vs);
-
- if (atomic_dec_and_test(&vh->refcnt))
- queue_work(&vh->del_work);
-}
-
-static __net_init int vxlan_init_net(struct net *net)
-{
- struct vxlan_net *vn = net_generic(net, vxlan_net_id);
- unsigned int h;
-
- mutex_init(&vn->sock_lock);
-
- for (h = 0; h < PORT_HASH_SIZE; ++h)
- INIT_HLIST_HEAD(&vn->sock_list[h]);
-
- return 0;
-}
-
-static struct pernet_operations vxlan_net_ops = {
- .init = vxlan_init_net,
- .id = &vxlan_net_id,
- .size = sizeof(struct vxlan_net),
-};
-
-static int refcnt;
-static DEFINE_MUTEX(init_lock);
-DEFINE_COMPAT_PNET_REG_FUNC(device);
-
-static int vxlan_init_module(void)
-{
- int err = 0;
-
- mutex_lock(&init_lock);
- if (refcnt)
- goto out;
- err = register_pernet_device(&vxlan_net_ops);
-out:
- if (!err)
- refcnt++;
- mutex_unlock(&init_lock);
- return err;
-}
-
-static void vxlan_cleanup_module(void)
-{
- mutex_lock(&init_lock);
- refcnt--;
- if (refcnt)
- goto out;
- unregister_pernet_device(&vxlan_net_ops);
-out:
- mutex_unlock(&init_lock);
-}
+#endif /* 3.12 */
git://git.onelab.eu / sliver-openvswitch.git / blobdiff