GRE: Refactor GRE tunneling code.
authorPravin B Shelar <pshelar@nicira.com>
Mon, 25 Mar 2013 14:49:35 +0000 (14:49 +0000)
committerDavid S. Miller <davem@davemloft.net>
Tue, 26 Mar 2013 16:27:18 +0000 (12:27 -0400)
Following patch refactors GRE code into ip tunneling code and GRE
specific code. Common tunneling code is moved to ip_tunnel module.
ip_tunnel module is written as generic library which can be used
by different tunneling implementations.

ip_tunnel module contains following components:
 - packet xmit and rcv generic code. xmit flow looks like
   (gre_xmit/ipip_xmit)->ip_tunnel_xmit->ip_local_out.
 - hash table of all devices.
 - lookup for tunnel devices.
 - control plane operations like device create, destroy, ioctl, netlink
   operations code.
 - registration for tunneling modules, like gre, ipip etc.
 - define single pcpu_tstats dev->tstats.
 - struct tnl_ptk_info added to pass parsed tunnel packet parameters.

ipip.h header is renamed to ip_tunnel.h

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 files changed:
drivers/net/vxlan.c
include/net/gre.h
include/net/ip6_tunnel.h
include/net/ip_tunnels.h [new file with mode: 0644]
include/net/ipip.h [deleted file]
net/ipv4/Kconfig
net/ipv4/Makefile
net/ipv4/af_inet.c
net/ipv4/gre.c
net/ipv4/ip_gre.c
net/ipv4/ip_tunnel.c [new file with mode: 0644]
net/ipv4/ip_vti.c
net/ipv4/ipip.c
net/ipv4/ipmr.c
net/ipv6/af_inet6.c
net/ipv6/ip6_gre.c
net/ipv6/ip6_tunnel.c
net/ipv6/sit.c

index 33427fd62515f0ca51cfe3f5bfeb9ef73919b6e4..fe9ea7d14951b0db0d0ab081c8808a2e04abdcd6 100644 (file)
@@ -33,7 +33,7 @@
 #include <net/arp.h>
 #include <net/ndisc.h>
 #include <net/ip.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/icmp.h>
 #include <net/udp.h>
 #include <net/rtnetlink.h>
index 82665474bcb7c6e1389b2741fd9eda141f617270..9f03a390c826a44a700f42715c48fe6f11eef9ca 100644 (file)
@@ -2,6 +2,7 @@
 #define __LINUX_GRE_H
 
 #include <linux/skbuff.h>
+#include <net/ip_tunnels.h>
 
 #define GREPROTO_CISCO         0
 #define GREPROTO_PPTP          1
@@ -12,7 +13,57 @@ struct gre_protocol {
        void (*err_handler)(struct sk_buff *skb, u32 info);
 };
 
+struct gre_base_hdr {
+       __be16 flags;
+       __be16 protocol;
+};
+#define GRE_HEADER_SECTION 4
+
 int gre_add_protocol(const struct gre_protocol *proto, u8 version);
 int gre_del_protocol(const struct gre_protocol *proto, u8 version);
 
+static inline __be16 gre_flags_to_tnl_flags(__be16 flags)
+{
+       __be16 tflags = 0;
+
+       if (flags & GRE_CSUM)
+               tflags |= TUNNEL_CSUM;
+       if (flags & GRE_ROUTING)
+               tflags |= TUNNEL_ROUTING;
+       if (flags & GRE_KEY)
+               tflags |= TUNNEL_KEY;
+       if (flags & GRE_SEQ)
+               tflags |= TUNNEL_SEQ;
+       if (flags & GRE_STRICT)
+               tflags |= TUNNEL_STRICT;
+       if (flags & GRE_REC)
+               tflags |= TUNNEL_REC;
+       if (flags & GRE_VERSION)
+               tflags |= TUNNEL_VERSION;
+
+       return tflags;
+}
+
+static inline __be16 tnl_flags_to_gre_flags(__be16 tflags)
+{
+       __be16 flags = 0;
+
+       if (tflags & TUNNEL_CSUM)
+               flags |= GRE_CSUM;
+       if (tflags & TUNNEL_ROUTING)
+               flags |= GRE_ROUTING;
+       if (tflags & TUNNEL_KEY)
+               flags |= GRE_KEY;
+       if (tflags & TUNNEL_SEQ)
+               flags |= GRE_SEQ;
+       if (tflags & TUNNEL_STRICT)
+               flags |= GRE_STRICT;
+       if (tflags & TUNNEL_REC)
+               flags |= GRE_REC;
+       if (tflags & TUNNEL_VERSION)
+               flags |= GRE_VERSION;
+
+       return flags;
+}
+
 #endif
index ebdef7f60862586dede971ec26c8c25f995838ee..4da5de10d1d49f62bbbe33e770e4c95f2b5c9ddd 100644 (file)
@@ -3,6 +3,7 @@
 
 #include <linux/ipv6.h>
 #include <linux/netdevice.h>
+#include <linux/if_tunnel.h>
 #include <linux/ip6_tunnel.h>
 
 #define IP6TUNNEL_ERR_TIMEO (30*HZ)
diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h
new file mode 100644 (file)
index 0000000..4b6f0b2
--- /dev/null
@@ -0,0 +1,177 @@
+#ifndef __NET_IP_TUNNELS_H
+#define __NET_IP_TUNNELS_H 1
+
+#include <linux/if_tunnel.h>
+#include <linux/netdevice.h>
+#include <linux/skbuff.h>
+#include <linux/types.h>
+#include <linux/u64_stats_sync.h>
+#include <net/dsfield.h>
+#include <net/gro_cells.h>
+#include <net/inet_ecn.h>
+#include <net/ip.h>
+#include <net/rtnetlink.h>
+
+#if IS_ENABLED(CONFIG_IPV6)
+#include <net/ipv6.h>
+#include <net/ip6_fib.h>
+#include <net/ip6_route.h>
+#endif
+
+/* Keep error state on tunnel for 30 sec */
+#define IPTUNNEL_ERR_TIMEO     (30*HZ)
+
+/* 6rd prefix/relay information */
+#ifdef CONFIG_IPV6_SIT_6RD
+struct ip_tunnel_6rd_parm {
+       struct in6_addr         prefix;
+       __be32                  relay_prefix;
+       u16                     prefixlen;
+       u16                     relay_prefixlen;
+};
+#endif
+
+struct ip_tunnel_prl_entry {
+       struct ip_tunnel_prl_entry __rcu *next;
+       __be32                          addr;
+       u16                             flags;
+       struct rcu_head                 rcu_head;
+};
+
+struct ip_tunnel {
+       struct ip_tunnel __rcu  *next;
+       struct hlist_node hash_node;
+       struct net_device       *dev;
+
+       int             err_count;      /* Number of arrived ICMP errors */
+       unsigned long   err_time;       /* Time when the last ICMP error
+                                        * arrived */
+
+       /* These four fields used only by GRE */
+       __u32           i_seqno;        /* The last seen seqno  */
+       __u32           o_seqno;        /* The last output seqno */
+       int             hlen;           /* Precalculated header length */
+       int             mlink;
+
+       struct ip_tunnel_parm parms;
+
+       /* for SIT */
+#ifdef CONFIG_IPV6_SIT_6RD
+       struct ip_tunnel_6rd_parm ip6rd;
+#endif
+       struct ip_tunnel_prl_entry __rcu *prl;  /* potential router list */
+       unsigned int            prl_count;      /* # of entries in PRL */
+       int                     ip_tnl_net_id;
+       struct gro_cells        gro_cells;
+};
+
+#define TUNNEL_CSUM    __cpu_to_be16(0x01)
+#define TUNNEL_ROUTING __cpu_to_be16(0x02)
+#define TUNNEL_KEY     __cpu_to_be16(0x04)
+#define TUNNEL_SEQ     __cpu_to_be16(0x08)
+#define TUNNEL_STRICT  __cpu_to_be16(0x10)
+#define TUNNEL_REC     __cpu_to_be16(0x20)
+#define TUNNEL_VERSION __cpu_to_be16(0x40)
+#define TUNNEL_NO_KEY  __cpu_to_be16(0x80)
+
+struct tnl_ptk_info {
+       __be16 flags;
+       __be16 proto;
+       __be32 key;
+       __be32 seq;
+};
+
+#define PACKET_RCVD    0
+#define PACKET_REJECT  1
+
+#define IP_TNL_HASH_BITS   10
+#define IP_TNL_HASH_SIZE   (1 << IP_TNL_HASH_BITS)
+
+struct ip_tunnel_net {
+       struct hlist_head *tunnels;
+       struct net_device *fb_tunnel_dev;
+};
+
+int ip_tunnel_init(struct net_device *dev);
+void ip_tunnel_uninit(struct net_device *dev);
+void  ip_tunnel_dellink(struct net_device *dev, struct list_head *head);
+int __net_init ip_tunnel_init_net(struct net *net, int ip_tnl_net_id,
+                                 struct rtnl_link_ops *ops, char *devname);
+
+void __net_exit ip_tunnel_delete_net(struct ip_tunnel_net *itn);
+
+void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
+                   const struct iphdr *tnl_params);
+int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd);
+int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu);
+
+struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev,
+                                               struct rtnl_link_stats64 *tot);
+struct ip_tunnel *ip_tunnel_lookup(struct ip_tunnel_net *itn,
+                                  int link, __be16 flags,
+                                  __be32 remote, __be32 local,
+                                  __be32 key);
+
+int ip_tunnel_rcv(struct ip_tunnel *tunnel, struct sk_buff *skb,
+                 const struct tnl_ptk_info *tpi, bool log_ecn_error);
+int ip_tunnel_changelink(struct net_device *dev, struct nlattr *tb[],
+                        struct ip_tunnel_parm *p);
+int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[],
+                     struct ip_tunnel_parm *p);
+void ip_tunnel_setup(struct net_device *dev, int net_id);
+
+/* Extract dsfield from inner protocol */
+static inline u8 ip_tunnel_get_dsfield(const struct iphdr *iph,
+                                      const struct sk_buff *skb)
+{
+       if (skb->protocol == htons(ETH_P_IP))
+               return iph->tos;
+       else if (skb->protocol == htons(ETH_P_IPV6))
+               return ipv6_get_dsfield((const struct ipv6hdr *)iph);
+       else
+               return 0;
+}
+
+/* Propogate ECN bits out */
+static inline u8 ip_tunnel_ecn_encap(u8 tos, const struct iphdr *iph,
+                                    const struct sk_buff *skb)
+{
+       u8 inner = ip_tunnel_get_dsfield(iph, skb);
+
+       return INET_ECN_encapsulate(tos, inner);
+}
+
+static inline void tunnel_ip_select_ident(struct sk_buff *skb,
+                                         const struct iphdr  *old_iph,
+                                         struct dst_entry *dst)
+{
+       struct iphdr *iph = ip_hdr(skb);
+
+       /* Use inner packet iph-id if possible. */
+       if (skb->protocol == htons(ETH_P_IP) && old_iph->id)
+               iph->id = old_iph->id;
+       else
+               __ip_select_ident(iph, dst,
+                                 (skb_shinfo(skb)->gso_segs ?: 1) - 1);
+}
+
+static inline void iptunnel_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+       int err;
+       int pkt_len = skb->len - skb_transport_offset(skb);
+       struct pcpu_tstats *tstats = this_cpu_ptr(dev->tstats);
+
+       nf_reset(skb);
+
+       err = ip_local_out(skb);
+       if (likely(net_xmit_eval(err) == 0)) {
+               u64_stats_update_begin(&tstats->syncp);
+               tstats->tx_bytes += pkt_len;
+               tstats->tx_packets++;
+               u64_stats_update_end(&tstats->syncp);
+       } else {
+               dev->stats.tx_errors++;
+               dev->stats.tx_aborted_errors++;
+       }
+}
+#endif /* __NET_IP_TUNNELS_H */
diff --git a/include/net/ipip.h b/include/net/ipip.h
deleted file mode 100644 (file)
index 483b91a..0000000
+++ /dev/null
@@ -1,84 +0,0 @@
-#ifndef __NET_IPIP_H
-#define __NET_IPIP_H 1
-
-#include <linux/if_tunnel.h>
-#include <net/gro_cells.h>
-#include <net/ip.h>
-
-/* Keep error state on tunnel for 30 sec */
-#define IPTUNNEL_ERR_TIMEO     (30*HZ)
-
-/* 6rd prefix/relay information */
-struct ip_tunnel_6rd_parm {
-       struct in6_addr         prefix;
-       __be32                  relay_prefix;
-       u16                     prefixlen;
-       u16                     relay_prefixlen;
-};
-
-struct ip_tunnel {
-       struct ip_tunnel __rcu  *next;
-       struct net_device       *dev;
-
-       int                     err_count;      /* Number of arrived ICMP errors */
-       unsigned long           err_time;       /* Time when the last ICMP error arrived */
-
-       /* These four fields used only by GRE */
-       __u32                   i_seqno;        /* The last seen seqno  */
-       __u32                   o_seqno;        /* The last output seqno */
-       int                     hlen;           /* Precalculated GRE header length */
-       int                     mlink;
-
-       struct ip_tunnel_parm   parms;
-
-       /* for SIT */
-#ifdef CONFIG_IPV6_SIT_6RD
-       struct ip_tunnel_6rd_parm       ip6rd;
-#endif
-       struct ip_tunnel_prl_entry __rcu *prl;          /* potential router list */
-       unsigned int                    prl_count;      /* # of entries in PRL */
-
-       struct gro_cells                gro_cells;
-};
-
-struct ip_tunnel_prl_entry {
-       struct ip_tunnel_prl_entry __rcu *next;
-       __be32                          addr;
-       u16                             flags;
-       struct rcu_head                 rcu_head;
-};
-
-static inline void iptunnel_xmit(struct sk_buff *skb, struct net_device *dev)
-{
-       int err;
-       int pkt_len = skb->len - skb_transport_offset(skb);
-       struct pcpu_tstats *tstats = this_cpu_ptr(dev->tstats);
-
-       nf_reset(skb);
-
-       err = ip_local_out(skb);
-       if (likely(net_xmit_eval(err) == 0)) {
-               u64_stats_update_begin(&tstats->syncp);
-               tstats->tx_bytes += pkt_len;
-               tstats->tx_packets++;
-               u64_stats_update_end(&tstats->syncp);
-       } else {
-               dev->stats.tx_errors++;
-               dev->stats.tx_aborted_errors++;
-       }
-}
-
-static inline void tunnel_ip_select_ident(struct sk_buff *skb,
-                                         const struct iphdr  *old_iph,
-                                         struct dst_entry *dst)
-{
-       struct iphdr *iph = ip_hdr(skb);
-
-       /* Use inner packet iph-id if possible. */
-       if (skb->protocol == htons(ETH_P_IP) && old_iph->id)
-               iph->id = old_iph->id;
-       else
-               __ip_select_ident(iph, dst,
-                                 (skb_shinfo(skb)->gso_segs ?: 1) - 1);
-}
-#endif
index 7944df768454caf1cc87905de319a644a6dc3736..2073226a8a63996dea73a90473182eb7b4a17332 100644 (file)
@@ -186,9 +186,14 @@ config NET_IPGRE_DEMUX
         This is helper module to demultiplex GRE packets on GRE version field criteria.
         Required by ip_gre and pptp modules.
 
+config NET_IP_TUNNEL
+       tristate
+       default n
+
 config NET_IPGRE
        tristate "IP: GRE tunnels over IP"
        depends on (IPV6 || IPV6=n) && NET_IPGRE_DEMUX
+       select NET_IP_TUNNEL
        help
          Tunneling means encapsulating data of one protocol type within
          another protocol and sending it over a channel that understands the
index 15ca63ec604ee8e01c1f79103c7fe39265a8831c..089cb9f363871327b027d8ce9246c830256cd456 100644 (file)
@@ -13,6 +13,7 @@ obj-y     := route.o inetpeer.o protocol.o \
             fib_frontend.o fib_semantics.o fib_trie.o \
             inet_fragment.o ping.o
 
+obj-$(CONFIG_NET_IP_TUNNEL) += ip_tunnel.o
 obj-$(CONFIG_SYSCTL) += sysctl_net_ipv4.o
 obj-$(CONFIG_PROC_FS) += proc.o
 obj-$(CONFIG_IP_MULTIPLE_TABLES) += fib_rules.o
index 70b2d4cf58017f7db6ee7db31e2a9da4c7b155f2..93824c57b1082d3fd473ed070249380fc24c45a5 100644 (file)
 #include <net/sock.h>
 #include <net/raw.h>
 #include <net/icmp.h>
-#include <net/ipip.h>
 #include <net/inet_common.h>
 #include <net/xfrm.h>
 #include <net/net_namespace.h>
index 7a4c710c4cddef241766bf5816b20ef768e8a45b..d2d5a99fba0983a4dead0909e30212c84f72154b 100644 (file)
 
 static const struct gre_protocol __rcu *gre_proto[GREPROTO_MAX] __read_mostly;
 static DEFINE_SPINLOCK(gre_proto_lock);
-struct gre_base_hdr {
-       __be16 flags;
-       __be16 protocol;
-};
-#define GRE_HEADER_SECTION 4
 
 int gre_add_protocol(const struct gre_protocol *proto, u8 version)
 {
index 2e94289a17e8572a6e0edddec0acfd2670e36bb9..ad662e906f7e403e895c600d909776807070a016 100644 (file)
@@ -37,7 +37,7 @@
 #include <net/ip.h>
 #include <net/icmp.h>
 #include <net/protocol.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/arp.h>
 #include <net/checksum.h>
 #include <net/dsfield.h>
    fatal route to network, even if it were you who configured
    fatal static route: you are innocent. :-)
 
-
-
-   3. Really, ipv4/ipip.c, ipv4/ip_gre.c and ipv6/sit.c contain
-   practically identical code. It would be good to glue them
-   together, but it is not very evident, how to make them modular.
-   sit is integral part of IPv6, ipip and gre are naturally modular.
-   We could extract common parts (hash table, ioctl etc)
-   to a separate module (ip_tunnel.c).
-
    Alexey Kuznetsov.
  */
 
@@ -126,400 +117,135 @@ MODULE_PARM_DESC(log_ecn_error, "Log packets received with corrupted ECN");
 
 static struct rtnl_link_ops ipgre_link_ops __read_mostly;
 static int ipgre_tunnel_init(struct net_device *dev);
-static void ipgre_tunnel_setup(struct net_device *dev);
-static int ipgre_tunnel_bind_dev(struct net_device *dev);
-
-/* Fallback tunnel: no source, no destination, no key, no options */
-
-#define HASH_SIZE  16
 
 static int ipgre_net_id __read_mostly;
-struct ipgre_net {
-       struct ip_tunnel __rcu *tunnels[4][HASH_SIZE];
-
-       struct net_device *fb_tunnel_dev;
-};
-
-/* Tunnel hash table */
-
-/*
-   4 hash tables:
-
-   3: (remote,local)
-   2: (remote,*)
-   1: (*,local)
-   0: (*,*)
+static int gre_tap_net_id __read_mostly;
 
-   We require exact key match i.e. if a key is present in packet
-   it will match only tunnel with the same key; if it is not present,
-   it will match only keyless tunnel.
-
-   All keysless packets, if not matched configured keyless tunnels
-   will match fallback tunnel.
- */
+static __sum16 check_checksum(struct sk_buff *skb)
+{
+       __sum16 csum = 0;
 
-#define HASH(addr) (((__force u32)addr^((__force u32)addr>>4))&0xF)
+       switch (skb->ip_summed) {
+       case CHECKSUM_COMPLETE:
+               csum = csum_fold(skb->csum);
 
-#define tunnels_r_l    tunnels[3]
-#define tunnels_r      tunnels[2]
-#define tunnels_l      tunnels[1]
-#define tunnels_wc     tunnels[0]
+               if (!csum)
+                       break;
+               /* Fall through. */
 
-static struct rtnl_link_stats64 *ipgre_get_stats64(struct net_device *dev,
-                                                  struct rtnl_link_stats64 *tot)
-{
-       int i;
-
-       for_each_possible_cpu(i) {
-               const struct pcpu_tstats *tstats = per_cpu_ptr(dev->tstats, i);
-               u64 rx_packets, rx_bytes, tx_packets, tx_bytes;
-               unsigned int start;
-
-               do {
-                       start = u64_stats_fetch_begin_bh(&tstats->syncp);
-                       rx_packets = tstats->rx_packets;
-                       tx_packets = tstats->tx_packets;
-                       rx_bytes = tstats->rx_bytes;
-                       tx_bytes = tstats->tx_bytes;
-               } while (u64_stats_fetch_retry_bh(&tstats->syncp, start));
-
-               tot->rx_packets += rx_packets;
-               tot->tx_packets += tx_packets;
-               tot->rx_bytes   += rx_bytes;
-               tot->tx_bytes   += tx_bytes;
+       case CHECKSUM_NONE:
+               skb->csum = 0;
+               csum = __skb_checksum_complete(skb);
+               skb->ip_summed = CHECKSUM_COMPLETE;
+               break;
        }
 
-       tot->multicast = dev->stats.multicast;
-       tot->rx_crc_errors = dev->stats.rx_crc_errors;
-       tot->rx_fifo_errors = dev->stats.rx_fifo_errors;
-       tot->rx_length_errors = dev->stats.rx_length_errors;
-       tot->rx_frame_errors = dev->stats.rx_frame_errors;
-       tot->rx_errors = dev->stats.rx_errors;
-
-       tot->tx_fifo_errors = dev->stats.tx_fifo_errors;
-       tot->tx_carrier_errors = dev->stats.tx_carrier_errors;
-       tot->tx_dropped = dev->stats.tx_dropped;
-       tot->tx_aborted_errors = dev->stats.tx_aborted_errors;
-       tot->tx_errors = dev->stats.tx_errors;
-
-       return tot;
+       return csum;
 }
 
-/* Does key in tunnel parameters match packet */
-static bool ipgre_key_match(const struct ip_tunnel_parm *p,
-                           __be16 flags, __be32 key)
+static int ip_gre_calc_hlen(__be16 o_flags)
 {
-       if (p->i_flags & GRE_KEY) {
-               if (flags & GRE_KEY)
-                       return key == p->i_key;
-               else
-                       return false;   /* key expected, none present */
-       } else
-               return !(flags & GRE_KEY);
-}
+       int addend = 4;
 
-/* Given src, dst and key, find appropriate for input tunnel. */
+       if (o_flags&TUNNEL_CSUM)
+               addend += 4;
+       if (o_flags&TUNNEL_KEY)
+               addend += 4;
+       if (o_flags&TUNNEL_SEQ)
+               addend += 4;
+       return addend;
+}
 
-static struct ip_tunnel *ipgre_tunnel_lookup(struct net_device *dev,
-                                            __be32 remote, __be32 local,
-                                            __be16 flags, __be32 key,
-                                            __be16 gre_proto)
+static int parse_gre_header(struct sk_buff *skb, struct tnl_ptk_info *tpi,
+                           bool *csum_err, int *hdr_len)
 {
-       struct net *net = dev_net(dev);
-       int link = dev->ifindex;
-       unsigned int h0 = HASH(remote);
-       unsigned int h1 = HASH(key);
-       struct ip_tunnel *t, *cand = NULL;
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-       int dev_type = (gre_proto == htons(ETH_P_TEB)) ?
-                      ARPHRD_ETHER : ARPHRD_IPGRE;
-       int score, cand_score = 4;
-
-       for_each_ip_tunnel_rcu(t, ign->tunnels_r_l[h0 ^ h1]) {
-               if (local != t->parms.iph.saddr ||
-                   remote != t->parms.iph.daddr ||
-                   !(t->dev->flags & IFF_UP))
-                       continue;
-
-               if (!ipgre_key_match(&t->parms, flags, key))
-                       continue;
-
-               if (t->dev->type != ARPHRD_IPGRE &&
-                   t->dev->type != dev_type)
-                       continue;
-
-               score = 0;
-               if (t->parms.link != link)
-                       score |= 1;
-               if (t->dev->type != dev_type)
-                       score |= 2;
-               if (score == 0)
-                       return t;
-
-               if (score < cand_score) {
-                       cand = t;
-                       cand_score = score;
-               }
-       }
-
-       for_each_ip_tunnel_rcu(t, ign->tunnels_r[h0 ^ h1]) {
-               if (remote != t->parms.iph.daddr ||
-                   !(t->dev->flags & IFF_UP))
-                       continue;
-
-               if (!ipgre_key_match(&t->parms, flags, key))
-                       continue;
-
-               if (t->dev->type != ARPHRD_IPGRE &&
-                   t->dev->type != dev_type)
-                       continue;
-
-               score = 0;
-               if (t->parms.link != link)
-                       score |= 1;
-               if (t->dev->type != dev_type)
-                       score |= 2;
-               if (score == 0)
-                       return t;
-
-               if (score < cand_score) {
-                       cand = t;
-                       cand_score = score;
-               }
-       }
+       struct iphdr *iph = ip_hdr(skb);
+       struct gre_base_hdr *greh;
+       __be32 *options;
 
-       for_each_ip_tunnel_rcu(t, ign->tunnels_l[h1]) {
-               if ((local != t->parms.iph.saddr &&
-                    (local != t->parms.iph.daddr ||
-                     !ipv4_is_multicast(local))) ||
-                   !(t->dev->flags & IFF_UP))
-                       continue;
-
-               if (!ipgre_key_match(&t->parms, flags, key))
-                       continue;
-
-               if (t->dev->type != ARPHRD_IPGRE &&
-                   t->dev->type != dev_type)
-                       continue;
-
-               score = 0;
-               if (t->parms.link != link)
-                       score |= 1;
-               if (t->dev->type != dev_type)
-                       score |= 2;
-               if (score == 0)
-                       return t;
-
-               if (score < cand_score) {
-                       cand = t;
-                       cand_score = score;
-               }
-       }
+       if (unlikely(!pskb_may_pull(skb, sizeof(struct gre_base_hdr))))
+               return -EINVAL;
 
-       for_each_ip_tunnel_rcu(t, ign->tunnels_wc[h1]) {
-               if (t->parms.i_key != key ||
-                   !(t->dev->flags & IFF_UP))
-                       continue;
-
-               if (t->dev->type != ARPHRD_IPGRE &&
-                   t->dev->type != dev_type)
-                       continue;
-
-               score = 0;
-               if (t->parms.link != link)
-                       score |= 1;
-               if (t->dev->type != dev_type)
-                       score |= 2;
-               if (score == 0)
-                       return t;
-
-               if (score < cand_score) {
-                       cand = t;
-                       cand_score = score;
-               }
-       }
+       greh = (struct gre_base_hdr *)((u8 *)iph + (iph->ihl << 2));
+       if (unlikely(greh->flags & (GRE_VERSION | GRE_ROUTING)))
+               return -EINVAL;
 
-       if (cand != NULL)
-               return cand;
+       tpi->flags = gre_flags_to_tnl_flags(greh->flags);
+       *hdr_len = ip_gre_calc_hlen(tpi->flags);
 
-       dev = ign->fb_tunnel_dev;
-       if (dev->flags & IFF_UP)
-               return netdev_priv(dev);
+       if (!pskb_may_pull(skb, *hdr_len))
+               return -EINVAL;
 
-       return NULL;
-}
+       tpi->proto = greh->protocol;
 
-static struct ip_tunnel __rcu **__ipgre_bucket(struct ipgre_net *ign,
-               struct ip_tunnel_parm *parms)
-{
-       __be32 remote = parms->iph.daddr;
-       __be32 local = parms->iph.saddr;
-       __be32 key = parms->i_key;
-       unsigned int h = HASH(key);
-       int prio = 0;
-
-       if (local)
-               prio |= 1;
-       if (remote && !ipv4_is_multicast(remote)) {
-               prio |= 2;
-               h ^= HASH(remote);
+       options = (__be32 *)(greh + 1);
+       if (greh->flags & GRE_CSUM) {
+               if (check_checksum(skb)) {
+                       *csum_err = true;
+                       return -EINVAL;
+               }
+               options++;
        }
 
-       return &ign->tunnels[prio][h];
-}
-
-static inline struct ip_tunnel __rcu **ipgre_bucket(struct ipgre_net *ign,
-               struct ip_tunnel *t)
-{
-       return __ipgre_bucket(ign, &t->parms);
-}
-
-static void ipgre_tunnel_link(struct ipgre_net *ign, struct ip_tunnel *t)
-{
-       struct ip_tunnel __rcu **tp = ipgre_bucket(ign, t);
+       if (greh->flags & GRE_KEY) {
+               tpi->key = *options;
+               options++;
+       } else
+               tpi->key = 0;
 
-       rcu_assign_pointer(t->next, rtnl_dereference(*tp));
-       rcu_assign_pointer(*tp, t);
-}
+       if (unlikely(greh->flags & GRE_SEQ)) {
+               tpi->seq = *options;
+               options++;
+       } else
+               tpi->seq = 0;
 
-static void ipgre_tunnel_unlink(struct ipgre_net *ign, struct ip_tunnel *t)
-{
-       struct ip_tunnel __rcu **tp;
-       struct ip_tunnel *iter;
-
-       for (tp = ipgre_bucket(ign, t);
-            (iter = rtnl_dereference(*tp)) != NULL;
-            tp = &iter->next) {
-               if (t == iter) {
-                       rcu_assign_pointer(*tp, t->next);
-                       break;
+       /* WCCP version 1 and 2 protocol decoding.
+        * - Change protocol to IP
+        * - When dealing with WCCPv2, Skip extra 4 bytes in GRE header
+        */
+       if (greh->flags == 0 && tpi->proto == htons(ETH_P_WCCP)) {
+               tpi->proto = htons(ETH_P_IP);
+               if ((*(u8 *)options & 0xF0) != 0x40) {
+                       *hdr_len += 4;
+                       if (!pskb_may_pull(skb, *hdr_len))
+                               return -EINVAL;
                }
        }
-}
-
-static struct ip_tunnel *ipgre_tunnel_find(struct net *net,
-                                          struct ip_tunnel_parm *parms,
-                                          int type)
-{
-       __be32 remote = parms->iph.daddr;
-       __be32 local = parms->iph.saddr;
-       __be32 key = parms->i_key;
-       int link = parms->link;
-       struct ip_tunnel *t;
-       struct ip_tunnel __rcu **tp;
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-
-       for (tp = __ipgre_bucket(ign, parms);
-            (t = rtnl_dereference(*tp)) != NULL;
-            tp = &t->next)
-               if (local == t->parms.iph.saddr &&
-                   remote == t->parms.iph.daddr &&
-                   key == t->parms.i_key &&
-                   link == t->parms.link &&
-                   type == t->dev->type)
-                       break;
-
-       return t;
-}
-
-static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
-               struct ip_tunnel_parm *parms, int create)
-{
-       struct ip_tunnel *t, *nt;
-       struct net_device *dev;
-       char name[IFNAMSIZ];
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-
-       t = ipgre_tunnel_find(net, parms, ARPHRD_IPGRE);
-       if (t || !create)
-               return t;
-
-       if (parms->name[0])
-               strlcpy(name, parms->name, IFNAMSIZ);
-       else
-               strcpy(name, "gre%d");
-
-       dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
-       if (!dev)
-               return NULL;
-
-       dev_net_set(dev, net);
-
-       nt = netdev_priv(dev);
-       nt->parms = *parms;
-       dev->rtnl_link_ops = &ipgre_link_ops;
 
-       dev->mtu = ipgre_tunnel_bind_dev(dev);
-
-       if (register_netdevice(dev) < 0)
-               goto failed_free;
-
-       /* Can use a lockless transmit, unless we generate output sequences */
-       if (!(nt->parms.o_flags & GRE_SEQ))
-               dev->features |= NETIF_F_LLTX;
-
-       dev_hold(dev);
-       ipgre_tunnel_link(ign, nt);
-       return nt;
-
-failed_free:
-       free_netdev(dev);
-       return NULL;
-}
-
-static void ipgre_tunnel_uninit(struct net_device *dev)
-{
-       struct net *net = dev_net(dev);
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-
-       ipgre_tunnel_unlink(ign, netdev_priv(dev));
-       dev_put(dev);
+       return 0;
 }
 
-
 static void ipgre_err(struct sk_buff *skb, u32 info)
 {
 
-/* All the routers (except for Linux) return only
-   8 bytes of packet payload. It means, that precise relaying of
-   ICMP in the real Internet is absolutely infeasible.
+       /* All the routers (except for Linux) return only
+          8 bytes of packet payload. It means, that precise relaying of
+          ICMP in the real Internet is absolutely infeasible.
 
-   Moreover, Cisco "wise men" put GRE key to the third word
-   in GRE header. It makes impossible maintaining even soft state for keyed
-   GRE tunnels with enabled checksum. Tell them "thank you".
-
-   Well, I wonder, rfc1812 was written by Cisco employee,
-   what the hell these idiots break standards established
-   by themselves???
- */
+          Moreover, Cisco "wise men" put GRE key to the third word
+          in GRE header. It makes impossible maintaining even soft
+          state for keyed GRE tunnels with enabled checksum. Tell
+          them "thank you".
 
+          Well, I wonder, rfc1812 was written by Cisco employee,
+          what the hell these idiots break standards established
+          by themselves???
+          */
+       struct net *net = dev_net(skb->dev);
+       struct ip_tunnel_net *itn;
        const struct iphdr *iph = (const struct iphdr *)skb->data;
-       __be16       *p = (__be16 *)(skb->data+(iph->ihl<<2));
-       int grehlen = (iph->ihl<<2) + 4;
        const int type = icmp_hdr(skb)->type;
        const int code = icmp_hdr(skb)->code;
        struct ip_tunnel *t;
-       __be16 flags;
-       __be32 key = 0;
+       struct tnl_ptk_info tpi;
+       int hdr_len;
+       bool csum_err = false;
 
-       flags = p[0];
-       if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) {
-               if (flags&(GRE_VERSION|GRE_ROUTING))
+       if (parse_gre_header(skb, &tpi, &csum_err, &hdr_len)) {
+               if (!csum_err)          /* ignore csum errors. */
                        return;
-               if (flags&GRE_KEY) {
-                       grehlen += 4;
-                       if (flags&GRE_CSUM)
-                               grehlen += 4;
-               }
        }
 
-       /* If only 8 bytes returned, keyed message will be dropped here */
-       if (skb_headlen(skb) < grehlen)
-               return;
-
-       if (flags & GRE_KEY)
-               key = *(((__be32 *)p) + (grehlen / 4) - 1);
-
        switch (type) {
        default:
        case ICMP_PARAMETERPROB:
@@ -548,8 +274,13 @@ static void ipgre_err(struct sk_buff *skb, u32 info)
                break;
        }
 
-       t = ipgre_tunnel_lookup(skb->dev, iph->daddr, iph->saddr,
-                               flags, key, p[1]);
+       if (tpi.proto == htons(ETH_P_TEB))
+               itn = net_generic(net, gre_tap_net_id);
+       else
+               itn = net_generic(net, ipgre_net_id);
+
+       t = ip_tunnel_lookup(itn, skb->dev->ifindex, tpi.flags,
+                            iph->daddr, iph->saddr, tpi.key);
 
        if (t == NULL)
                return;
@@ -578,158 +309,33 @@ static void ipgre_err(struct sk_buff *skb, u32 info)
        t->err_time = jiffies;
 }
 
-static inline u8
-ipgre_ecn_encapsulate(u8 tos, const struct iphdr *old_iph, struct sk_buff *skb)
-{
-       u8 inner = 0;
-       if (skb->protocol == htons(ETH_P_IP))
-               inner = old_iph->tos;
-       else if (skb->protocol == htons(ETH_P_IPV6))
-               inner = ipv6_get_dsfield((const struct ipv6hdr *)old_iph);
-       return INET_ECN_encapsulate(tos, inner);
-}
-
 static int ipgre_rcv(struct sk_buff *skb)
 {
+       struct net *net = dev_net(skb->dev);
+       struct ip_tunnel_net *itn;
        const struct iphdr *iph;
-       u8     *h;
-       __be16    flags;
-       __sum16   csum = 0;
-       __be32 key = 0;
-       u32    seqno = 0;
        struct ip_tunnel *tunnel;
-       int    offset = 4;
-       __be16 gre_proto;
-       int    err;
+       struct tnl_ptk_info tpi;
+       int hdr_len;
+       bool csum_err = false;
 
-       if (!pskb_may_pull(skb, 16))
+       if (parse_gre_header(skb, &tpi, &csum_err, &hdr_len) < 0)
                goto drop;
 
-       iph = ip_hdr(skb);
-       h = skb->data;
-       flags = *(__be16 *)h;
-
-       if (flags&(GRE_CSUM|GRE_KEY|GRE_ROUTING|GRE_SEQ|GRE_VERSION)) {
-               /* - Version must be 0.
-                  - We do not support routing headers.
-                */
-               if (flags&(GRE_VERSION|GRE_ROUTING))
-                       goto drop;
-
-               if (flags&GRE_CSUM) {
-                       switch (skb->ip_summed) {
-                       case CHECKSUM_COMPLETE:
-                               csum = csum_fold(skb->csum);
-                               if (!csum)
-                                       break;
-                               /* fall through */
-                       case CHECKSUM_NONE:
-                               skb->csum = 0;
-                               csum = __skb_checksum_complete(skb);
-                               skb->ip_summed = CHECKSUM_COMPLETE;
-                       }
-                       offset += 4;
-               }
-               if (flags&GRE_KEY) {
-                       key = *(__be32 *)(h + offset);
-                       offset += 4;
-               }
-               if (flags&GRE_SEQ) {
-                       seqno = ntohl(*(__be32 *)(h + offset));
-                       offset += 4;
-               }
-       }
+       if (tpi.proto == htons(ETH_P_TEB))
+               itn = net_generic(net, gre_tap_net_id);
+       else
+               itn = net_generic(net, ipgre_net_id);
 
-       gre_proto = *(__be16 *)(h + 2);
+       iph = ip_hdr(skb);
+       tunnel = ip_tunnel_lookup(itn, skb->dev->ifindex, tpi.flags,
+                                 iph->saddr, iph->daddr, tpi.key);
 
-       tunnel = ipgre_tunnel_lookup(skb->dev,
-                                    iph->saddr, iph->daddr, flags, key,
-                                    gre_proto);
        if (tunnel) {
-               struct pcpu_tstats *tstats;
-
-               secpath_reset(skb);
-
-               skb->protocol = gre_proto;
-               /* WCCP version 1 and 2 protocol decoding.
-                * - Change protocol to IP
-                * - When dealing with WCCPv2, Skip extra 4 bytes in GRE header
-                */
-               if (flags == 0 && gre_proto == htons(ETH_P_WCCP)) {
-                       skb->protocol = htons(ETH_P_IP);
-                       if ((*(h + offset) & 0xF0) != 0x40)
-                               offset += 4;
-               }
-
-               skb->mac_header = skb->network_header;
-               __pskb_pull(skb, offset);
-               skb_postpull_rcsum(skb, skb_transport_header(skb), offset);
-               skb->pkt_type = PACKET_HOST;
-#ifdef CONFIG_NET_IPGRE_BROADCAST
-               if (ipv4_is_multicast(iph->daddr)) {
-                       /* Looped back packet, drop it! */
-                       if (rt_is_output_route(skb_rtable(skb)))
-                               goto drop;
-                       tunnel->dev->stats.multicast++;
-                       skb->pkt_type = PACKET_BROADCAST;
-               }
-#endif
-
-               if (((flags&GRE_CSUM) && csum) ||
-                   (!(flags&GRE_CSUM) && tunnel->parms.i_flags&GRE_CSUM)) {
-                       tunnel->dev->stats.rx_crc_errors++;
-                       tunnel->dev->stats.rx_errors++;
-                       goto drop;
-               }
-               if (tunnel->parms.i_flags&GRE_SEQ) {
-                       if (!(flags&GRE_SEQ) ||
-                           (tunnel->i_seqno && (s32)(seqno - tunnel->i_seqno) < 0)) {
-                               tunnel->dev->stats.rx_fifo_errors++;
-                               tunnel->dev->stats.rx_errors++;
-                               goto drop;
-                       }
-                       tunnel->i_seqno = seqno + 1;
-               }
-
-               /* Warning: All skb pointers will be invalidated! */
-               if (tunnel->dev->type == ARPHRD_ETHER) {
-                       if (!pskb_may_pull(skb, ETH_HLEN)) {
-                               tunnel->dev->stats.rx_length_errors++;
-                               tunnel->dev->stats.rx_errors++;
-                               goto drop;
-                       }
-
-                       iph = ip_hdr(skb);
-                       skb->protocol = eth_type_trans(skb, tunnel->dev);
-                       skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
-               }
-
-               __skb_tunnel_rx(skb, tunnel->dev);
-
-               skb_reset_network_header(skb);
-               err = IP_ECN_decapsulate(iph, skb);
-               if (unlikely(err)) {
-                       if (log_ecn_error)
-                               net_info_ratelimited("non-ECT from %pI4 with TOS=%#x\n",
-                                                    &iph->saddr, iph->tos);
-                       if (err > 1) {
-                               ++tunnel->dev->stats.rx_frame_errors;
-                               ++tunnel->dev->stats.rx_errors;
-                               goto drop;
-                       }
-               }
-
-               tstats = this_cpu_ptr(tunnel->dev->tstats);
-               u64_stats_update_begin(&tstats->syncp);
-               tstats->rx_packets++;
-               tstats->rx_bytes += skb->len;
-               u64_stats_update_end(&tstats->syncp);
-
-               gro_cells_receive(&tunnel->gro_cells, skb);
+               ip_tunnel_rcv(tunnel, skb, &tpi, log_ecn_error);
                return 0;
        }
        icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
-
 drop:
        kfree_skb(skb);
        return 0;
@@ -746,7 +352,7 @@ static struct sk_buff *handle_offloads(struct ip_tunnel *tunnel, struct sk_buff
                skb_shinfo(skb)->gso_type |= SKB_GSO_GRE;
                return skb;
        } else if (skb->ip_summed == CHECKSUM_PARTIAL &&
-                  tunnel->parms.o_flags&GRE_CSUM) {
+                  tunnel->parms.o_flags&TUNNEL_CSUM) {
                err = skb_checksum_help(skb);
                if (unlikely(err))
                        goto error;
@@ -760,480 +366,157 @@ error:
        return ERR_PTR(err);
 }
 
-static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
+static struct sk_buff *gre_build_header(struct sk_buff *skb,
+                                       const struct tnl_ptk_info *tpi,
+                                       int hdr_len)
 {
-       struct ip_tunnel *tunnel = netdev_priv(dev);
-       const struct iphdr  *old_iph;
-       const struct iphdr  *tiph;
-       struct flowi4 fl4;
-       u8     tos;
-       __be16 df;
-       struct rtable *rt;                      /* Route to the other host */
-       struct net_device *tdev;                /* Device to other host */
-       struct iphdr  *iph;                     /* Our new IP header */
-       unsigned int max_headroom;              /* The extra header space needed */
-       int    gre_hlen;
-       __be32 dst;
-       int    mtu;
-       u8     ttl;
-       int    err;
-
-       skb = handle_offloads(tunnel, skb);
-       if (IS_ERR(skb)) {
-               dev->stats.tx_dropped++;
-               return NETDEV_TX_OK;
-       }
+       struct gre_base_hdr *greh;
 
-       if (!skb->encapsulation) {
-               skb_reset_inner_headers(skb);
-               skb->encapsulation = 1;
-       }
+       skb_push(skb, hdr_len);
 
-       old_iph = ip_hdr(skb);
+       greh = (struct gre_base_hdr *)skb->data;
+       greh->flags = tnl_flags_to_gre_flags(tpi->flags);
+       greh->protocol = tpi->proto;
 
-       if (dev->type == ARPHRD_ETHER)
-               IPCB(skb)->flags = 0;
+       if (tpi->flags&(TUNNEL_KEY|TUNNEL_CSUM|TUNNEL_SEQ)) {
+               __be32 *ptr = (__be32 *)(((u8 *)greh) + hdr_len - 4);
 
-       if (dev->header_ops && dev->type == ARPHRD_IPGRE) {
-               gre_hlen = 0;
-               tiph = (const struct iphdr *)skb->data;
-       } else {
-               gre_hlen = tunnel->hlen;
-               tiph = &tunnel->parms.iph;
-       }
-
-       if ((dst = tiph->daddr) == 0) {
-               /* NBMA tunnel */
-
-               if (skb_dst(skb) == NULL) {
-                       dev->stats.tx_fifo_errors++;
-                       goto tx_error;
+               if (tpi->flags&TUNNEL_SEQ) {
+                       *ptr = tpi->seq;
+                       ptr--;
                }
-
-               if (skb->protocol == htons(ETH_P_IP)) {
-                       rt = skb_rtable(skb);
-                       dst = rt_nexthop(rt, old_iph->daddr);
+               if (tpi->flags&TUNNEL_KEY) {
+                       *ptr = tpi->key;
+                       ptr--;
                }
-#if IS_ENABLED(CONFIG_IPV6)
-               else if (skb->protocol == htons(ETH_P_IPV6)) {
-                       const struct in6_addr *addr6;
-                       struct neighbour *neigh;
-                       bool do_tx_error_icmp;
-                       int addr_type;
-
-                       neigh = dst_neigh_lookup(skb_dst(skb), &ipv6_hdr(skb)->daddr);
-                       if (neigh == NULL)
-                               goto tx_error;
-
-                       addr6 = (const struct in6_addr *)&neigh->primary_key;
-                       addr_type = ipv6_addr_type(addr6);
-
-                       if (addr_type == IPV6_ADDR_ANY) {
-                               addr6 = &ipv6_hdr(skb)->daddr;
-                               addr_type = ipv6_addr_type(addr6);
-                       }
-
-                       if ((addr_type & IPV6_ADDR_COMPATv4) == 0)
-                               do_tx_error_icmp = true;
-                       else {
-                               do_tx_error_icmp = false;
-                               dst = addr6->s6_addr32[3];
-                       }
-                       neigh_release(neigh);
-                       if (do_tx_error_icmp)
-                               goto tx_error_icmp;
+               if (tpi->flags&TUNNEL_CSUM &&
+                   !(skb_shinfo(skb)->gso_type & SKB_GSO_GRE)) {
+                       *(__sum16 *)ptr = 0;
+                       *(__sum16 *)ptr = csum_fold(skb_checksum(skb, 0,
+                                                                skb->len, 0));
                }
-#endif
-               else
-                       goto tx_error;
        }
 
-       ttl = tiph->ttl;
-       tos = tiph->tos;
-       if (tos & 0x1) {
-               tos &= ~0x1;
-               if (skb->protocol == htons(ETH_P_IP))
-                       tos = old_iph->tos;
-               else if (skb->protocol == htons(ETH_P_IPV6))
-                       tos = ipv6_get_dsfield((const struct ipv6hdr *)old_iph);
-       }
+       return skb;
+}
 
-       rt = ip_route_output_gre(dev_net(dev), &fl4, dst, tiph->saddr,
-                                tunnel->parms.o_key, RT_TOS(tos),
-                                tunnel->parms.link);
-       if (IS_ERR(rt)) {
-               dev->stats.tx_carrier_errors++;
-               goto tx_error;
-       }
-       tdev = rt->dst.dev;
+static void __gre_xmit(struct sk_buff *skb, struct net_device *dev,
+                      const struct iphdr *tnl_params,
+                      __be16 proto)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct tnl_ptk_info tpi;
 
-       if (tdev == dev) {
-               ip_rt_put(rt);
-               dev->stats.collisions++;
-               goto tx_error;
+       if (likely(!skb->encapsulation)) {
+               skb_reset_inner_headers(skb);
+               skb->encapsulation = 1;
        }
 
-       df = tiph->frag_off;
-       if (df)
-               mtu = dst_mtu(&rt->dst) - dev->hard_header_len - tunnel->hlen;
-       else
-               mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
-
-       if (skb_dst(skb))
-               skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu);
-
-       if (skb->protocol == htons(ETH_P_IP)) {
-               df |= (old_iph->frag_off&htons(IP_DF));
+       tpi.flags = tunnel->parms.o_flags;
+       tpi.proto = proto;
+       tpi.key = tunnel->parms.o_key;
+       if (tunnel->parms.o_flags & TUNNEL_SEQ)
+               tunnel->o_seqno++;
+       tpi.seq = htonl(tunnel->o_seqno);
 
-               if (!skb_is_gso(skb) &&
-                   (old_iph->frag_off&htons(IP_DF)) &&
-                   mtu < ntohs(old_iph->tot_len)) {
-                       icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
-                       ip_rt_put(rt);
-                       goto tx_error;
-               }
+       /* Push GRE header. */
+       skb = gre_build_header(skb, &tpi, tunnel->hlen);
+       if (unlikely(!skb)) {
+               dev->stats.tx_dropped++;
+               return;
        }
-#if IS_ENABLED(CONFIG_IPV6)
-       else if (skb->protocol == htons(ETH_P_IPV6)) {
-               struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
-
-               if (rt6 && mtu < dst_mtu(skb_dst(skb)) && mtu >= IPV6_MIN_MTU) {
-                       if ((tunnel->parms.iph.daddr &&
-                            !ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
-                           rt6->rt6i_dst.plen == 128) {
-                               rt6->rt6i_flags |= RTF_MODIFIED;
-                               dst_metric_set(skb_dst(skb), RTAX_MTU, mtu);
-                       }
-               }
 
-               if (!skb_is_gso(skb) &&
-                   mtu >= IPV6_MIN_MTU &&
-                   mtu < skb->len - tunnel->hlen + gre_hlen) {
-                       icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
-                       ip_rt_put(rt);
-                       goto tx_error;
-               }
-       }
-#endif
+       ip_tunnel_xmit(skb, dev, tnl_params);
+}
 
-       if (tunnel->err_count > 0) {
-               if (time_before(jiffies,
-                               tunnel->err_time + IPTUNNEL_ERR_TIMEO)) {
-                       tunnel->err_count--;
+static netdev_tx_t ipgre_xmit(struct sk_buff *skb,
+                             struct net_device *dev)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       const struct iphdr *tnl_params;
 
-                       dst_link_failure(skb);
-               } else
-                       tunnel->err_count = 0;
-       }
+       skb = handle_offloads(tunnel, skb);
+       if (IS_ERR(skb))
+               goto out;
 
-       max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen + rt->dst.header_len;
-
-       if (skb_headroom(skb) < max_headroom || skb_shared(skb)||
-           (skb_cloned(skb) && !skb_clone_writable(skb, 0))) {
-               struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom);
-               if (max_headroom > dev->needed_headroom)
-                       dev->needed_headroom = max_headroom;
-               if (!new_skb) {
-                       ip_rt_put(rt);
-                       dev->stats.tx_dropped++;
-                       dev_kfree_skb(skb);
-                       return NETDEV_TX_OK;
-               }
-               if (skb->sk)
-                       skb_set_owner_w(new_skb, skb->sk);
-               dev_kfree_skb(skb);
-               skb = new_skb;
-               old_iph = ip_hdr(skb);
-               /* Warning : tiph value might point to freed memory */
-       }
+       if (dev->header_ops) {
+               /* Need space for new headers */
+               if (skb_cow_head(skb, dev->needed_headroom -
+                                     (tunnel->hlen + sizeof(struct iphdr))));
+                       goto free_skb;
 
-       skb_push(skb, gre_hlen);
-       skb_reset_network_header(skb);
-       skb_set_transport_header(skb, sizeof(*iph));
-       memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
-       IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED |
-                             IPSKB_REROUTED);
-       skb_dst_drop(skb);
-       skb_dst_set(skb, &rt->dst);
-
-       /*
-        *      Push down and install the IPIP header.
-        */
+               tnl_params = (const struct iphdr *)skb->data;
 
-       iph                     =       ip_hdr(skb);
-       iph->version            =       4;
-       iph->ihl                =       sizeof(struct iphdr) >> 2;
-       iph->frag_off           =       df;
-       iph->protocol           =       IPPROTO_GRE;
-       iph->tos                =       ipgre_ecn_encapsulate(tos, old_iph, skb);
-       iph->daddr              =       fl4.daddr;
-       iph->saddr              =       fl4.saddr;
-       iph->ttl                =       ttl;
-
-       tunnel_ip_select_ident(skb, old_iph, &rt->dst);
-
-       if (ttl == 0) {
-               if (skb->protocol == htons(ETH_P_IP))
-                       iph->ttl = old_iph->ttl;
-#if IS_ENABLED(CONFIG_IPV6)
-               else if (skb->protocol == htons(ETH_P_IPV6))
-                       iph->ttl = ((const struct ipv6hdr *)old_iph)->hop_limit;
-#endif
-               else
-                       iph->ttl = ip4_dst_hoplimit(&rt->dst);
-       }
-
-       ((__be16 *)(iph + 1))[0] = tunnel->parms.o_flags;
-       ((__be16 *)(iph + 1))[1] = (dev->type == ARPHRD_ETHER) ?
-                                  htons(ETH_P_TEB) : skb->protocol;
-
-       if (tunnel->parms.o_flags&(GRE_KEY|GRE_CSUM|GRE_SEQ)) {
-               __be32 *ptr = (__be32 *)(((u8 *)iph) + tunnel->hlen - 4);
+               /* Pull skb since ip_tunnel_xmit() needs skb->data pointing
+                * to gre header.
+                */
+               skb_pull(skb, tunnel->hlen + sizeof(struct iphdr));
+       } else {
+               if (skb_cow_head(skb, dev->needed_headroom))
+                       goto free_skb;
 
-               if (tunnel->parms.o_flags&GRE_SEQ) {
-                       ++tunnel->o_seqno;
-                       *ptr = htonl(tunnel->o_seqno);
-                       ptr--;
-               }
-               if (tunnel->parms.o_flags&GRE_KEY) {
-                       *ptr = tunnel->parms.o_key;
-                       ptr--;
-               }
-               /* Skip GRE checksum if skb is getting offloaded. */
-               if (!(skb_shinfo(skb)->gso_type & SKB_GSO_GRE) &&
-                   (tunnel->parms.o_flags&GRE_CSUM)) {
-                       int offset = skb_transport_offset(skb);
-
-                       if (skb_has_shared_frag(skb)) {
-                               err = __skb_linearize(skb);
-                               if (err)
-                                       goto tx_error;
-                       }
-
-                       *ptr = 0;
-                       *(__sum16 *)ptr = csum_fold(skb_checksum(skb, offset,
-                                                                skb->len - offset,
-                                                                0));
-               }
+               tnl_params = &tunnel->parms.iph;
        }
 
-       iptunnel_xmit(skb, dev);
+       __gre_xmit(skb, dev, tnl_params, skb->protocol);
+
        return NETDEV_TX_OK;
 
-#if IS_ENABLED(CONFIG_IPV6)
-tx_error_icmp:
-       dst_link_failure(skb);
-#endif
-tx_error:
-       dev->stats.tx_errors++;
+free_skb:
        dev_kfree_skb(skb);
+out:
+       dev->stats.tx_dropped++;
        return NETDEV_TX_OK;
 }
 
-static int ipgre_tunnel_bind_dev(struct net_device *dev)
+static netdev_tx_t gre_tap_xmit(struct sk_buff *skb,
+                               struct net_device *dev)
 {
-       struct net_device *tdev = NULL;
-       struct ip_tunnel *tunnel;
-       const struct iphdr *iph;
-       int hlen = LL_MAX_HEADER;
-       int mtu = ETH_DATA_LEN;
-       int addend = sizeof(struct iphdr) + 4;
-
-       tunnel = netdev_priv(dev);
-       iph = &tunnel->parms.iph;
-
-       /* Guess output device to choose reasonable mtu and needed_headroom */
-
-       if (iph->daddr) {
-               struct flowi4 fl4;
-               struct rtable *rt;
-
-               rt = ip_route_output_gre(dev_net(dev), &fl4,
-                                        iph->daddr, iph->saddr,
-                                        tunnel->parms.o_key,
-                                        RT_TOS(iph->tos),
-                                        tunnel->parms.link);
-               if (!IS_ERR(rt)) {
-                       tdev = rt->dst.dev;
-                       ip_rt_put(rt);
-               }
-
-               if (dev->type != ARPHRD_ETHER)
-                       dev->flags |= IFF_POINTOPOINT;
-       }
+       struct ip_tunnel *tunnel = netdev_priv(dev);
 
-       if (!tdev && tunnel->parms.link)
-               tdev = __dev_get_by_index(dev_net(dev), tunnel->parms.link);
+       skb = handle_offloads(tunnel, skb);
+       if (IS_ERR(skb))
+               goto out;
 
-       if (tdev) {
-               hlen = tdev->hard_header_len + tdev->needed_headroom;
-               mtu = tdev->mtu;
-       }
-       dev->iflink = tunnel->parms.link;
-
-       /* Precalculate GRE options length */
-       if (tunnel->parms.o_flags&(GRE_CSUM|GRE_KEY|GRE_SEQ)) {
-               if (tunnel->parms.o_flags&GRE_CSUM)
-                       addend += 4;
-               if (tunnel->parms.o_flags&GRE_KEY)
-                       addend += 4;
-               if (tunnel->parms.o_flags&GRE_SEQ)
-                       addend += 4;
-       }
-       dev->needed_headroom = addend + hlen;
-       mtu -= dev->hard_header_len + addend;
+       if (skb_cow_head(skb, dev->needed_headroom))
+               goto free_skb;
 
-       if (mtu < 68)
-               mtu = 68;
+       __gre_xmit(skb, dev, &tunnel->parms.iph, htons(ETH_P_TEB));
 
-       tunnel->hlen = addend;
-       /* TCP offload with GRE SEQ is not supported. */
-       if (!(tunnel->parms.o_flags & GRE_SEQ)) {
-               dev->features           |= NETIF_F_GSO_SOFTWARE;
-               dev->hw_features        |= NETIF_F_GSO_SOFTWARE;
-       }
+       return NETDEV_TX_OK;
 
-       return mtu;
+free_skb:
+       dev_kfree_skb(skb);
+out:
+       dev->stats.tx_dropped++;
+       return NETDEV_TX_OK;
 }
 
-static int
-ipgre_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
+static int ipgre_tunnel_ioctl(struct net_device *dev,
+                             struct ifreq *ifr, int cmd)
 {
        int err = 0;
        struct ip_tunnel_parm p;
-       struct ip_tunnel *t;
-       struct net *net = dev_net(dev);
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-
-       switch (cmd) {
-       case SIOCGETTUNNEL:
-               t = NULL;
-               if (dev == ign->fb_tunnel_dev) {
-                       if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
-                               err = -EFAULT;
-                               break;
-                       }
-                       t = ipgre_tunnel_locate(net, &p, 0);
-               }
-               if (t == NULL)
-                       t = netdev_priv(dev);
-               memcpy(&p, &t->parms, sizeof(p));
-               if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p)))
-                       err = -EFAULT;
-               break;
-
-       case SIOCADDTUNNEL:
-       case SIOCCHGTUNNEL:
-               err = -EPERM;
-               if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
-                       goto done;
-
-               err = -EFAULT;
-               if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
-                       goto done;
-
-               err = -EINVAL;
-               if (p.iph.version != 4 || p.iph.protocol != IPPROTO_GRE ||
-                   p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)) ||
-                   ((p.i_flags|p.o_flags)&(GRE_VERSION|GRE_ROUTING)))
-                       goto done;
-               if (p.iph.ttl)
-                       p.iph.frag_off |= htons(IP_DF);
-
-               if (!(p.i_flags&GRE_KEY))
-                       p.i_key = 0;
-               if (!(p.o_flags&GRE_KEY))
-                       p.o_key = 0;
-
-               t = ipgre_tunnel_locate(net, &p, cmd == SIOCADDTUNNEL);
-
-               if (dev != ign->fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
-                       if (t != NULL) {
-                               if (t->dev != dev) {
-                                       err = -EEXIST;
-                                       break;
-                               }
-                       } else {
-                               unsigned int nflags = 0;
-
-                               t = netdev_priv(dev);
-
-                               if (ipv4_is_multicast(p.iph.daddr))
-                                       nflags = IFF_BROADCAST;
-                               else if (p.iph.daddr)
-                                       nflags = IFF_POINTOPOINT;
-
-                               if ((dev->flags^nflags)&(IFF_POINTOPOINT|IFF_BROADCAST)) {
-                                       err = -EINVAL;
-                                       break;
-                               }
-                               ipgre_tunnel_unlink(ign, t);
-                               synchronize_net();
-                               t->parms.iph.saddr = p.iph.saddr;
-                               t->parms.iph.daddr = p.iph.daddr;
-                               t->parms.i_key = p.i_key;
-                               t->parms.o_key = p.o_key;
-                               memcpy(dev->dev_addr, &p.iph.saddr, 4);
-                               memcpy(dev->broadcast, &p.iph.daddr, 4);
-                               ipgre_tunnel_link(ign, t);
-                               netdev_state_change(dev);
-                       }
-               }
 
-               if (t) {
-                       err = 0;
-                       if (cmd == SIOCCHGTUNNEL) {
-                               t->parms.iph.ttl = p.iph.ttl;
-                               t->parms.iph.tos = p.iph.tos;
-                               t->parms.iph.frag_off = p.iph.frag_off;
-                               if (t->parms.link != p.link) {
-                                       t->parms.link = p.link;
-                                       dev->mtu = ipgre_tunnel_bind_dev(dev);
-                                       netdev_state_change(dev);
-                               }
-                       }
-                       if (copy_to_user(ifr->ifr_ifru.ifru_data, &t->parms, sizeof(p)))
-                               err = -EFAULT;
-               } else
-                       err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT);
-               break;
-
-       case SIOCDELTUNNEL:
-               err = -EPERM;
-               if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
-                       goto done;
-
-               if (dev == ign->fb_tunnel_dev) {
-                       err = -EFAULT;
-                       if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
-                               goto done;
-                       err = -ENOENT;
-                       if ((t = ipgre_tunnel_locate(net, &p, 0)) == NULL)
-                               goto done;
-                       err = -EPERM;
-                       if (t == netdev_priv(ign->fb_tunnel_dev))
-                               goto done;
-                       dev = t->dev;
-               }
-               unregister_netdevice(dev);
-               err = 0;
-               break;
-
-       default:
-               err = -EINVAL;
+       if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
+               return -EFAULT;
+       if (p.iph.version != 4 || p.iph.protocol != IPPROTO_GRE ||
+           p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)) ||
+           ((p.i_flags|p.o_flags)&(GRE_VERSION|GRE_ROUTING))) {
+               return -EINVAL;
        }
+       p.i_flags = gre_flags_to_tnl_flags(p.i_flags);
+       p.o_flags = gre_flags_to_tnl_flags(p.o_flags);
 
-done:
-       return err;
-}
+       err = ip_tunnel_ioctl(dev, &p, cmd);
+       if (err)
+               return err;
 
-static int ipgre_tunnel_change_mtu(struct net_device *dev, int new_mtu)
-{
-       struct ip_tunnel *tunnel = netdev_priv(dev);
-       if (new_mtu < 68 ||
-           new_mtu > 0xFFF8 - dev->hard_header_len - tunnel->hlen)
-               return -EINVAL;
-       dev->mtu = new_mtu;
+       p.i_flags = tnl_flags_to_gre_flags(p.i_flags);
+       p.o_flags = tnl_flags_to_gre_flags(p.o_flags);
+
+       if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p)))
+               return -EFAULT;
        return 0;
 }
 
@@ -1263,25 +546,23 @@ static int ipgre_tunnel_change_mtu(struct net_device *dev, int new_mtu)
    ...
    ftp fec0:6666:6666::193.233.7.65
    ...
-
  */
-
 static int ipgre_header(struct sk_buff *skb, struct net_device *dev,
                        unsigned short type,
                        const void *daddr, const void *saddr, unsigned int len)
 {
        struct ip_tunnel *t = netdev_priv(dev);
-       struct iphdr *iph = (struct iphdr *)skb_push(skb, t->hlen);
-       __be16 *p = (__be16 *)(iph+1);
+       struct iphdr *iph;
+       struct gre_base_hdr *greh;
 
-       memcpy(iph, &t->parms.iph, sizeof(struct iphdr));
-       p[0]            = t->parms.o_flags;
-       p[1]            = htons(type);
+       iph = (struct iphdr *)skb_push(skb, t->hlen + sizeof(*iph));
+       greh = (struct gre_base_hdr *)(iph+1);
+       greh->flags = tnl_flags_to_gre_flags(t->parms.o_flags);
+       greh->protocol = htons(type);
 
-       /*
-        *      Set the source hardware address.
-        */
+       memcpy(iph, &t->parms.iph, sizeof(struct iphdr));
 
+       /* Set the source hardware address. */
        if (saddr)
                memcpy(&iph->saddr, saddr, 4);
        if (daddr)
@@ -1289,7 +570,7 @@ static int ipgre_header(struct sk_buff *skb, struct net_device *dev,
        if (iph->daddr)
                return t->hlen;
 
-       return -t->hlen;
+       return -(t->hlen + sizeof(*iph));
 }
 
 static int ipgre_header_parse(const struct sk_buff *skb, unsigned char *haddr)
@@ -1343,31 +624,21 @@ static int ipgre_close(struct net_device *dev)
        }
        return 0;
 }
-
 #endif
 
 static const struct net_device_ops ipgre_netdev_ops = {
        .ndo_init               = ipgre_tunnel_init,
-       .ndo_uninit             = ipgre_tunnel_uninit,
+       .ndo_uninit             = ip_tunnel_uninit,
 #ifdef CONFIG_NET_IPGRE_BROADCAST
        .ndo_open               = ipgre_open,
        .ndo_stop               = ipgre_close,
 #endif
-       .ndo_start_xmit         = ipgre_tunnel_xmit,
+       .ndo_start_xmit         = ipgre_xmit,
        .ndo_do_ioctl           = ipgre_tunnel_ioctl,
-       .ndo_change_mtu         = ipgre_tunnel_change_mtu,
-       .ndo_get_stats64        = ipgre_get_stats64,
+       .ndo_change_mtu         = ip_tunnel_change_mtu,
+       .ndo_get_stats64        = ip_tunnel_get_stats64,
 };
 
-static void ipgre_dev_free(struct net_device *dev)
-{
-       struct ip_tunnel *tunnel = netdev_priv(dev);
-
-       gro_cells_destroy(&tunnel->gro_cells);
-       free_percpu(dev->tstats);
-       free_netdev(dev);
-}
-
 #define GRE_FEATURES (NETIF_F_SG |             \
                      NETIF_F_FRAGLIST |        \
                      NETIF_F_HIGHDMA |         \
@@ -1376,35 +647,49 @@ static void ipgre_dev_free(struct net_device *dev)
 static void ipgre_tunnel_setup(struct net_device *dev)
 {
        dev->netdev_ops         = &ipgre_netdev_ops;
-       dev->destructor         = ipgre_dev_free;
+       ip_tunnel_setup(dev, ipgre_net_id);
+}
 
-       dev->type               = ARPHRD_IPGRE;
-       dev->needed_headroom    = LL_MAX_HEADER + sizeof(struct iphdr) + 4;
+static void __gre_tunnel_init(struct net_device *dev)
+{
+       struct ip_tunnel *tunnel;
+
+       tunnel = netdev_priv(dev);
+       tunnel->hlen = ip_gre_calc_hlen(tunnel->parms.o_flags);
+       tunnel->parms.iph.protocol = IPPROTO_GRE;
+
+       dev->needed_headroom    = LL_MAX_HEADER + sizeof(struct iphdr) + 4;
        dev->mtu                = ETH_DATA_LEN - sizeof(struct iphdr) - 4;
-       dev->flags              = IFF_NOARP;
        dev->iflink             = 0;
-       dev->addr_len           = 4;
-       dev->features           |= NETIF_F_NETNS_LOCAL;
-       dev->priv_flags         &= ~IFF_XMIT_DST_RELEASE;
 
-       dev->features           |= GRE_FEATURES;
+       dev->features           |= NETIF_F_NETNS_LOCAL | GRE_FEATURES;
        dev->hw_features        |= GRE_FEATURES;
+
+       if (!(tunnel->parms.o_flags & TUNNEL_SEQ)) {
+               /* TCP offload with GRE SEQ is not supported. */
+               dev->features    |= NETIF_F_GSO_SOFTWARE;
+               dev->hw_features |= NETIF_F_GSO_SOFTWARE;
+               /* Can use a lockless transmit, unless we generate
+                * output sequences
+                */
+               dev->features |= NETIF_F_LLTX;
+       }
 }
 
 static int ipgre_tunnel_init(struct net_device *dev)
 {
-       struct ip_tunnel *tunnel;
-       struct iphdr *iph;
-       int err;
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct iphdr *iph = &tunnel->parms.iph;
 
-       tunnel = netdev_priv(dev);
-       iph = &tunnel->parms.iph;
+       __gre_tunnel_init(dev);
 
-       tunnel->dev = dev;
-       strcpy(tunnel->parms.name, dev->name);
+       memcpy(dev->dev_addr, &iph->saddr, 4);
+       memcpy(dev->broadcast, &iph->daddr, 4);
 
-       memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4);
-       memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4);
+       dev->type               = ARPHRD_IPGRE;
+       dev->flags              = IFF_NOARP;
+       dev->priv_flags         &= ~IFF_XMIT_DST_RELEASE;
+       dev->addr_len           = 4;
 
        if (iph->daddr) {
 #ifdef CONFIG_NET_IPGRE_BROADCAST
@@ -1418,106 +703,30 @@ static int ipgre_tunnel_init(struct net_device *dev)
        } else
                dev->header_ops = &ipgre_header_ops;
 
-       dev->tstats = alloc_percpu(struct pcpu_tstats);
-       if (!dev->tstats)
-               return -ENOMEM;
-
-       err = gro_cells_init(&tunnel->gro_cells, dev);
-       if (err) {
-               free_percpu(dev->tstats);
-               return err;
-       }
-
-       return 0;
+       return ip_tunnel_init(dev);
 }
 
-static void ipgre_fb_tunnel_init(struct net_device *dev)
-{
-       struct ip_tunnel *tunnel = netdev_priv(dev);
-       struct iphdr *iph = &tunnel->parms.iph;
-
-       tunnel->dev = dev;
-       strcpy(tunnel->parms.name, dev->name);
-
-       iph->version            = 4;
-       iph->protocol           = IPPROTO_GRE;
-       iph->ihl                = 5;
-       tunnel->hlen            = sizeof(struct iphdr) + 4;
-
-       dev_hold(dev);
-}
-
-
 static const struct gre_protocol ipgre_protocol = {
        .handler     = ipgre_rcv,
        .err_handler = ipgre_err,
 };
 
-static void ipgre_destroy_tunnels(struct ipgre_net *ign, struct list_head *head)
-{
-       int prio;
-
-       for (prio = 0; prio < 4; prio++) {
-               int h;
-               for (h = 0; h < HASH_SIZE; h++) {
-                       struct ip_tunnel *t;
-
-                       t = rtnl_dereference(ign->tunnels[prio][h]);
-
-                       while (t != NULL) {
-                               unregister_netdevice_queue(t->dev, head);
-                               t = rtnl_dereference(t->next);
-                       }
-               }
-       }
-}
-
 static int __net_init ipgre_init_net(struct net *net)
 {
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-       int err;
-
-       ign->fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0",
-                                          ipgre_tunnel_setup);
-       if (!ign->fb_tunnel_dev) {
-               err = -ENOMEM;
-               goto err_alloc_dev;
-       }
-       dev_net_set(ign->fb_tunnel_dev, net);
-
-       ipgre_fb_tunnel_init(ign->fb_tunnel_dev);
-       ign->fb_tunnel_dev->rtnl_link_ops = &ipgre_link_ops;
-
-       if ((err = register_netdev(ign->fb_tunnel_dev)))
-               goto err_reg_dev;
-
-       rcu_assign_pointer(ign->tunnels_wc[0],
-                          netdev_priv(ign->fb_tunnel_dev));
-       return 0;
-
-err_reg_dev:
-       ipgre_dev_free(ign->fb_tunnel_dev);
-err_alloc_dev:
-       return err;
+       return ip_tunnel_init_net(net, ipgre_net_id, &ipgre_link_ops, NULL);
 }
 
 static void __net_exit ipgre_exit_net(struct net *net)
 {
-       struct ipgre_net *ign;
-       LIST_HEAD(list);
-
-       ign = net_generic(net, ipgre_net_id);
-       rtnl_lock();
-       ipgre_destroy_tunnels(ign, &list);
-       unregister_netdevice_many(&list);
-       rtnl_unlock();
+       struct ip_tunnel_net *itn = net_generic(net, ipgre_net_id);
+       ip_tunnel_delete_net(itn);
 }
 
 static struct pernet_operations ipgre_net_ops = {
        .init = ipgre_init_net,
        .exit = ipgre_exit_net,
        .id   = &ipgre_net_id,
-       .size = sizeof(struct ipgre_net),
+       .size = sizeof(struct ip_tunnel_net),
 };
 
 static int ipgre_tunnel_validate(struct nlattr *tb[], struct nlattr *data[])
@@ -1562,8 +771,8 @@ out:
        return ipgre_tunnel_validate(tb, data);
 }
 
-static void ipgre_netlink_parms(struct nlattr *data[],
-                               struct ip_tunnel_parm *parms)
+static void ipgre_netlink_parms(struct nlattr *data[], struct nlattr *tb[],
+                              struct ip_tunnel_parm *parms)
 {
        memset(parms, 0, sizeof(*parms));
 
@@ -1576,10 +785,10 @@ static void ipgre_netlink_parms(struct nlattr *data[],
                parms->link = nla_get_u32(data[IFLA_GRE_LINK]);
 
        if (data[IFLA_GRE_IFLAGS])
-               parms->i_flags = nla_get_be16(data[IFLA_GRE_IFLAGS]);
+               parms->i_flags = gre_flags_to_tnl_flags(nla_get_be16(data[IFLA_GRE_IFLAGS]));
 
        if (data[IFLA_GRE_OFLAGS])
-               parms->o_flags = nla_get_be16(data[IFLA_GRE_OFLAGS]);
+               parms->o_flags = gre_flags_to_tnl_flags(nla_get_be16(data[IFLA_GRE_OFLAGS]));
 
        if (data[IFLA_GRE_IKEY])
                parms->i_key = nla_get_be32(data[IFLA_GRE_IKEY]);
@@ -1603,148 +812,46 @@ static void ipgre_netlink_parms(struct nlattr *data[],
                parms->iph.frag_off = htons(IP_DF);
 }
 
-static int ipgre_tap_init(struct net_device *dev)
+static int gre_tap_init(struct net_device *dev)
 {
-       struct ip_tunnel *tunnel;
-
-       tunnel = netdev_priv(dev);
-
-       tunnel->dev = dev;
-       strcpy(tunnel->parms.name, dev->name);
-
-       ipgre_tunnel_bind_dev(dev);
+       __gre_tunnel_init(dev);
 
-       dev->tstats = alloc_percpu(struct pcpu_tstats);
-       if (!dev->tstats)
-               return -ENOMEM;
-
-       return 0;
+       return ip_tunnel_init(dev);
 }
 
-static const struct net_device_ops ipgre_tap_netdev_ops = {
-       .ndo_init               = ipgre_tap_init,
-       .ndo_uninit             = ipgre_tunnel_uninit,
-       .ndo_start_xmit         = ipgre_tunnel_xmit,
+static const struct net_device_ops gre_tap_netdev_ops = {
+       .ndo_init               = gre_tap_init,
+       .ndo_uninit             = ip_tunnel_uninit,
+       .ndo_start_xmit         = gre_tap_xmit,
        .ndo_set_mac_address    = eth_mac_addr,
        .ndo_validate_addr      = eth_validate_addr,
-       .ndo_change_mtu         = ipgre_tunnel_change_mtu,
-       .ndo_get_stats64        = ipgre_get_stats64,
+       .ndo_change_mtu         = ip_tunnel_change_mtu,
+       .ndo_get_stats64        = ip_tunnel_get_stats64,
 };
 
 static void ipgre_tap_setup(struct net_device *dev)
 {
-
        ether_setup(dev);
-
-       dev->netdev_ops         = &ipgre_tap_netdev_ops;
-       dev->destructor         = ipgre_dev_free;
-
-       dev->iflink             = 0;
-       dev->features           |= NETIF_F_NETNS_LOCAL;
-
-       dev->features           |= GRE_FEATURES;
-       dev->hw_features        |= GRE_FEATURES;
+       dev->netdev_ops         = &gre_tap_netdev_ops;
+       ip_tunnel_setup(dev, gre_tap_net_id);
 }
 
-static int ipgre_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[],
-                        struct nlattr *data[])
+static int ipgre_newlink(struct net *src_net, struct net_device *dev,
+                        struct nlattr *tb[], struct nlattr *data[])
 {
-       struct ip_tunnel *nt;
-       struct net *net = dev_net(dev);
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
-       int mtu;
-       int err;
-
-       nt = netdev_priv(dev);
-       ipgre_netlink_parms(data, &nt->parms);
-
-       if (ipgre_tunnel_find(net, &nt->parms, dev->type))
-               return -EEXIST;
-
-       if (dev->type == ARPHRD_ETHER && !tb[IFLA_ADDRESS])
-               eth_hw_addr_random(dev);
-
-       mtu = ipgre_tunnel_bind_dev(dev);
-       if (!tb[IFLA_MTU])
-               dev->mtu = mtu;
-
-       /* Can use a lockless transmit, unless we generate output sequences */
-       if (!(nt->parms.o_flags & GRE_SEQ))
-               dev->features |= NETIF_F_LLTX;
-
-       err = register_netdevice(dev);
-       if (err)
-               goto out;
-
-       dev_hold(dev);
-       ipgre_tunnel_link(ign, nt);
+       struct ip_tunnel_parm p;
 
-out:
-       return err;
+       ipgre_netlink_parms(data, tb, &p);
+       return ip_tunnel_newlink(dev, tb, &p);
 }
 
 static int ipgre_changelink(struct net_device *dev, struct nlattr *tb[],
                            struct nlattr *data[])
 {
-       struct ip_tunnel *t, *nt;
-       struct net *net = dev_net(dev);
-       struct ipgre_net *ign = net_generic(net, ipgre_net_id);
        struct ip_tunnel_parm p;
-       int mtu;
-
-       if (dev == ign->fb_tunnel_dev)
-               return -EINVAL;
-
-       nt = netdev_priv(dev);
-       ipgre_netlink_parms(data, &p);
-
-       t = ipgre_tunnel_locate(net, &p, 0);
-
-       if (t) {
-               if (t->dev != dev)
-                       return -EEXIST;
-       } else {
-               t = nt;
-
-               if (dev->type != ARPHRD_ETHER) {
-                       unsigned int nflags = 0;
-
-                       if (ipv4_is_multicast(p.iph.daddr))
-                               nflags = IFF_BROADCAST;
-                       else if (p.iph.daddr)
-                               nflags = IFF_POINTOPOINT;
-
-                       if ((dev->flags ^ nflags) &
-                           (IFF_POINTOPOINT | IFF_BROADCAST))
-                               return -EINVAL;
-               }
 
-               ipgre_tunnel_unlink(ign, t);
-               t->parms.iph.saddr = p.iph.saddr;
-               t->parms.iph.daddr = p.iph.daddr;
-               t->parms.i_key = p.i_key;
-               if (dev->type != ARPHRD_ETHER) {
-                       memcpy(dev->dev_addr, &p.iph.saddr, 4);
-                       memcpy(dev->broadcast, &p.iph.daddr, 4);
-               }
-               ipgre_tunnel_link(ign, t);
-               netdev_state_change(dev);
-       }
-
-       t->parms.o_key = p.o_key;
-       t->parms.iph.ttl = p.iph.ttl;
-       t->parms.iph.tos = p.iph.tos;
-       t->parms.iph.frag_off = p.iph.frag_off;
-
-       if (t->parms.link != p.link) {
-               t->parms.link = p.link;
-               mtu = ipgre_tunnel_bind_dev(dev);
-               if (!tb[IFLA_MTU])
-                       dev->mtu = mtu;
-               netdev_state_change(dev);
-       }
-
-       return 0;
+       ipgre_netlink_parms(data, tb, &p);
+       return ip_tunnel_changelink(dev, tb, &p);
 }
 
 static size_t ipgre_get_size(const struct net_device *dev)
@@ -1779,8 +886,8 @@ static int ipgre_fill_info(struct sk_buff *skb, const struct net_device *dev)
        struct ip_tunnel_parm *p = &t->parms;
 
        if (nla_put_u32(skb, IFLA_GRE_LINK, p->link) ||
-           nla_put_be16(skb, IFLA_GRE_IFLAGS, p->i_flags) ||
-           nla_put_be16(skb, IFLA_GRE_OFLAGS, p->o_flags) ||
+           nla_put_be16(skb, IFLA_GRE_IFLAGS, tnl_flags_to_gre_flags(p->i_flags)) ||
+           nla_put_be16(skb, IFLA_GRE_OFLAGS, tnl_flags_to_gre_flags(p->o_flags)) ||
            nla_put_be32(skb, IFLA_GRE_IKEY, p->i_key) ||
            nla_put_be32(skb, IFLA_GRE_OKEY, p->o_key) ||
            nla_put_be32(skb, IFLA_GRE_LOCAL, p->iph.saddr) ||
@@ -1818,6 +925,7 @@ static struct rtnl_link_ops ipgre_link_ops __read_mostly = {
        .validate       = ipgre_tunnel_validate,
        .newlink        = ipgre_newlink,
        .changelink     = ipgre_changelink,
+       .dellink        = ip_tunnel_dellink,
        .get_size       = ipgre_get_size,
        .fill_info      = ipgre_fill_info,
 };
@@ -1831,13 +939,28 @@ static struct rtnl_link_ops ipgre_tap_ops __read_mostly = {
        .validate       = ipgre_tap_validate,
        .newlink        = ipgre_newlink,
        .changelink     = ipgre_changelink,
+       .dellink        = ip_tunnel_dellink,
        .get_size       = ipgre_get_size,
        .fill_info      = ipgre_fill_info,
 };
 
-/*
- *     And now the modules code and kernel interface.
- */
+static int __net_init ipgre_tap_init_net(struct net *net)
+{
+       return ip_tunnel_init_net(net, gre_tap_net_id, &ipgre_tap_ops, NULL);
+}
+
+static void __net_exit ipgre_tap_exit_net(struct net *net)
+{
+       struct ip_tunnel_net *itn = net_generic(net, gre_tap_net_id);
+       ip_tunnel_delete_net(itn);
+}
+
+static struct pernet_operations ipgre_tap_net_ops = {
+       .init = ipgre_tap_init_net,
+       .exit = ipgre_tap_exit_net,
+       .id   = &gre_tap_net_id,
+       .size = sizeof(struct ip_tunnel_net),
+};
 
 static int __init ipgre_init(void)
 {
@@ -1849,6 +972,10 @@ static int __init ipgre_init(void)
        if (err < 0)
                return err;
 
+       err = register_pernet_device(&ipgre_tap_net_ops);
+       if (err < 0)
+               goto pnet_tap_faied;
+
        err = gre_add_protocol(&ipgre_protocol, GREPROTO_CISCO);
        if (err < 0) {
                pr_info("%s: can't add protocol\n", __func__);
@@ -1863,16 +990,17 @@ static int __init ipgre_init(void)
        if (err < 0)
                goto tap_ops_failed;
 
-out:
-       return err;
+       return 0;
 
 tap_ops_failed:
        rtnl_link_unregister(&ipgre_link_ops);
 rtnl_link_failed:
        gre_del_protocol(&ipgre_protocol, GREPROTO_CISCO);
 add_proto_failed:
+       unregister_pernet_device(&ipgre_tap_net_ops);
+pnet_tap_faied:
        unregister_pernet_device(&ipgre_net_ops);
-       goto out;
+       return err;
 }
 
 static void __exit ipgre_fini(void)
@@ -1881,6 +1009,7 @@ static void __exit ipgre_fini(void)
        rtnl_link_unregister(&ipgre_link_ops);
        if (gre_del_protocol(&ipgre_protocol, GREPROTO_CISCO) < 0)
                pr_info("%s: can't remove protocol\n", __func__);
+       unregister_pernet_device(&ipgre_tap_net_ops);
        unregister_pernet_device(&ipgre_net_ops);
 }
 
@@ -1890,3 +1019,4 @@ MODULE_LICENSE("GPL");
 MODULE_ALIAS_RTNL_LINK("gre");
 MODULE_ALIAS_RTNL_LINK("gretap");
 MODULE_ALIAS_NETDEV("gre0");
+MODULE_ALIAS_NETDEV("gretap0");
diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
new file mode 100644 (file)
index 0000000..9d96b68
--- /dev/null
@@ -0,0 +1,1035 @@
+/*
+ * Copyright (c) 2013 Nicira, 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/capability.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/in.h>
+#include <linux/tcp.h>
+#include <linux/udp.h>
+#include <linux/if_arp.h>
+#include <linux/mroute.h>
+#include <linux/init.h>
+#include <linux/in6.h>
+#include <linux/inetdevice.h>
+#include <linux/igmp.h>
+#include <linux/netfilter_ipv4.h>
+#include <linux/etherdevice.h>
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+#include <linux/rculist.h>
+
+#include <net/sock.h>
+#include <net/ip.h>
+#include <net/icmp.h>
+#include <net/protocol.h>
+#include <net/ip_tunnels.h>
+#include <net/arp.h>
+#include <net/checksum.h>
+#include <net/dsfield.h>
+#include <net/inet_ecn.h>
+#include <net/xfrm.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
+#include <net/rtnetlink.h>
+
+#if IS_ENABLED(CONFIG_IPV6)
+#include <net/ipv6.h>
+#include <net/ip6_fib.h>
+#include <net/ip6_route.h>
+#endif
+
+static unsigned int ip_tunnel_hash(struct ip_tunnel_net *itn,
+                                  __be32 key, __be32 remote)
+{
+       return hash_32((__force u32)key ^ (__force u32)remote,
+                        IP_TNL_HASH_BITS);
+}
+
+/* Often modified stats are per cpu, other are shared (netdev->stats) */
+struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev,
+                                               struct rtnl_link_stats64 *tot)
+{
+       int i;
+
+       for_each_possible_cpu(i) {
+               const struct pcpu_tstats *tstats = per_cpu_ptr(dev->tstats, i);
+               u64 rx_packets, rx_bytes, tx_packets, tx_bytes;
+               unsigned int start;
+
+               do {
+                       start = u64_stats_fetch_begin_bh(&tstats->syncp);
+                       rx_packets = tstats->rx_packets;
+                       tx_packets = tstats->tx_packets;
+                       rx_bytes = tstats->rx_bytes;
+                       tx_bytes = tstats->tx_bytes;
+               } while (u64_stats_fetch_retry_bh(&tstats->syncp, start));
+
+               tot->rx_packets += rx_packets;
+               tot->tx_packets += tx_packets;
+               tot->rx_bytes   += rx_bytes;
+               tot->tx_bytes   += tx_bytes;
+       }
+
+       tot->multicast = dev->stats.multicast;
+
+       tot->rx_crc_errors = dev->stats.rx_crc_errors;
+       tot->rx_fifo_errors = dev->stats.rx_fifo_errors;
+       tot->rx_length_errors = dev->stats.rx_length_errors;
+       tot->rx_frame_errors = dev->stats.rx_frame_errors;
+       tot->rx_errors = dev->stats.rx_errors;
+
+       tot->tx_fifo_errors = dev->stats.tx_fifo_errors;
+       tot->tx_carrier_errors = dev->stats.tx_carrier_errors;
+       tot->tx_dropped = dev->stats.tx_dropped;
+       tot->tx_aborted_errors = dev->stats.tx_aborted_errors;
+       tot->tx_errors = dev->stats.tx_errors;
+
+       tot->collisions  = dev->stats.collisions;
+
+       return tot;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_get_stats64);
+
+static bool ip_tunnel_key_match(const struct ip_tunnel_parm *p,
+                               __be16 flags, __be32 key)
+{
+       if (p->i_flags & TUNNEL_KEY) {
+               if (flags & TUNNEL_KEY)
+                       return key == p->i_key;
+               else
+                       /* key expected, none present */
+                       return false;
+       } else
+               return !(flags & TUNNEL_KEY);
+}
+
+/* Fallback tunnel: no source, no destination, no key, no options
+
+   Tunnel hash table:
+   We require exact key match i.e. if a key is present in packet
+   it will match only tunnel with the same key; if it is not present,
+   it will match only keyless tunnel.
+
+   All keysless packets, if not matched configured keyless tunnels
+   will match fallback tunnel.
+   Given src, dst and key, find appropriate for input tunnel.
+*/
+struct ip_tunnel *ip_tunnel_lookup(struct ip_tunnel_net *itn,
+                                  int link, __be16 flags,
+                                  __be32 remote, __be32 local,
+                                  __be32 key)
+{
+       unsigned int hash;
+       struct ip_tunnel *t, *cand = NULL;
+       struct hlist_head *head;
+
+       hash = ip_tunnel_hash(itn, key, remote);
+       head = &itn->tunnels[hash];
+
+       hlist_for_each_entry_rcu(t, head, hash_node) {
+               if (local != t->parms.iph.saddr ||
+                   remote != t->parms.iph.daddr ||
+                   !(t->dev->flags & IFF_UP))
+                       continue;
+
+               if (!ip_tunnel_key_match(&t->parms, flags, key))
+                       continue;
+
+               if (t->parms.link == link)
+                       return t;
+               else
+                       cand = t;
+       }
+
+       hlist_for_each_entry_rcu(t, head, hash_node) {
+               if (remote != t->parms.iph.daddr ||
+                   !(t->dev->flags & IFF_UP))
+                       continue;
+
+               if (!ip_tunnel_key_match(&t->parms, flags, key))
+                       continue;
+
+               if (t->parms.link == link)
+                       return t;
+               else if (!cand)
+                       cand = t;
+       }
+
+       hash = ip_tunnel_hash(itn, key, 0);
+       head = &itn->tunnels[hash];
+
+       hlist_for_each_entry_rcu(t, head, hash_node) {
+               if ((local != t->parms.iph.saddr &&
+                    (local != t->parms.iph.daddr ||
+                     !ipv4_is_multicast(local))) ||
+                   !(t->dev->flags & IFF_UP))
+                       continue;
+
+               if (!ip_tunnel_key_match(&t->parms, flags, key))
+                       continue;
+
+               if (t->parms.link == link)
+                       return t;
+               else if (!cand)
+                       cand = t;
+       }
+
+       if (flags & TUNNEL_NO_KEY)
+               goto skip_key_lookup;
+
+       hlist_for_each_entry_rcu(t, head, hash_node) {
+               if (t->parms.i_key != key ||
+                   !(t->dev->flags & IFF_UP))
+                       continue;
+
+               if (t->parms.link == link)
+                       return t;
+               else if (!cand)
+                       cand = t;
+       }
+
+skip_key_lookup:
+       if (cand)
+               return cand;
+
+       if (itn->fb_tunnel_dev && itn->fb_tunnel_dev->flags & IFF_UP)
+               return netdev_priv(itn->fb_tunnel_dev);
+
+
+       return NULL;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_lookup);
+
+static struct hlist_head *ip_bucket(struct ip_tunnel_net *itn,
+                                   struct ip_tunnel_parm *parms)
+{
+       unsigned int h;
+       __be32 remote;
+
+       if (parms->iph.daddr && !ipv4_is_multicast(parms->iph.daddr))
+               remote = parms->iph.daddr;
+       else
+               remote = 0;
+
+       h = ip_tunnel_hash(itn, parms->i_key, remote);
+       return &itn->tunnels[h];
+}
+
+static void ip_tunnel_add(struct ip_tunnel_net *itn, struct ip_tunnel *t)
+{
+       struct hlist_head *head = ip_bucket(itn, &t->parms);
+
+       hlist_add_head_rcu(&t->hash_node, head);
+}
+
+static void ip_tunnel_del(struct ip_tunnel *t)
+{
+       hlist_del_init_rcu(&t->hash_node);
+}
+
+static struct ip_tunnel *ip_tunnel_find(struct ip_tunnel_net *itn,
+                                       struct ip_tunnel_parm *parms,
+                                       int type)
+{
+       __be32 remote = parms->iph.daddr;
+       __be32 local = parms->iph.saddr;
+       __be32 key = parms->i_key;
+       int link = parms->link;
+       struct ip_tunnel *t = NULL;
+       struct hlist_head *head = ip_bucket(itn, parms);
+
+       hlist_for_each_entry_rcu(t, head, hash_node) {
+               if (local == t->parms.iph.saddr &&
+                   remote == t->parms.iph.daddr &&
+                   key == t->parms.i_key &&
+                   link == t->parms.link &&
+                   type == t->dev->type)
+                       break;
+       }
+       return t;
+}
+
+static struct net_device *__ip_tunnel_create(struct net *net,
+                                            const struct rtnl_link_ops *ops,
+                                            struct ip_tunnel_parm *parms)
+{
+       int err;
+       struct ip_tunnel *tunnel;
+       struct net_device *dev;
+       char name[IFNAMSIZ];
+
+       if (parms->name[0])
+               strlcpy(name, parms->name, IFNAMSIZ);
+       else {
+               if (strlen(ops->kind) + 3 >= IFNAMSIZ) {
+                       err = -E2BIG;
+                       goto failed;
+               }
+               strlcpy(name, ops->kind, IFNAMSIZ);
+               strncat(name, "%d", 2);
+       }
+
+       ASSERT_RTNL();
+       dev = alloc_netdev(ops->priv_size, name, ops->setup);
+       if (!dev) {
+               err = -ENOMEM;
+               goto failed;
+       }
+       dev_net_set(dev, net);
+
+       dev->rtnl_link_ops = ops;
+
+       tunnel = netdev_priv(dev);
+       tunnel->parms = *parms;
+
+       err = register_netdevice(dev);
+       if (err)
+               goto failed_free;
+
+       return dev;
+
+failed_free:
+       free_netdev(dev);
+failed:
+       return ERR_PTR(err);
+}
+
+static inline struct rtable *ip_route_output_tunnel(struct net *net,
+                                                   struct flowi4 *fl4,
+                                                   int proto,
+                                                   __be32 daddr, __be32 saddr,
+                                                   __be32 key, __u8 tos, int oif)
+{
+       memset(fl4, 0, sizeof(*fl4));
+       fl4->flowi4_oif = oif;
+       fl4->daddr = daddr;
+       fl4->saddr = saddr;
+       fl4->flowi4_tos = tos;
+       fl4->flowi4_proto = proto;
+       fl4->fl4_gre_key = key;
+       return ip_route_output_key(net, fl4);
+}
+
+static int ip_tunnel_bind_dev(struct net_device *dev)
+{
+       struct net_device *tdev = NULL;
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       const struct iphdr *iph;
+       int hlen = LL_MAX_HEADER;
+       int mtu = ETH_DATA_LEN;
+       int t_hlen = tunnel->hlen + sizeof(struct iphdr);
+
+       iph = &tunnel->parms.iph;
+
+       /* Guess output device to choose reasonable mtu and needed_headroom */
+       if (iph->daddr) {
+               struct flowi4 fl4;
+               struct rtable *rt;
+
+               rt = ip_route_output_tunnel(dev_net(dev), &fl4,
+                                           tunnel->parms.iph.protocol,
+                                           iph->daddr, iph->saddr,
+                                           tunnel->parms.o_key,
+                                           RT_TOS(iph->tos),
+                                           tunnel->parms.link);
+               if (!IS_ERR(rt)) {
+                       tdev = rt->dst.dev;
+                       ip_rt_put(rt);
+               }
+               if (dev->type != ARPHRD_ETHER)
+                       dev->flags |= IFF_POINTOPOINT;
+       }
+
+       if (!tdev && tunnel->parms.link)
+               tdev = __dev_get_by_index(dev_net(dev), tunnel->parms.link);
+
+       if (tdev) {
+               hlen = tdev->hard_header_len + tdev->needed_headroom;
+               mtu = tdev->mtu;
+       }
+       dev->iflink = tunnel->parms.link;
+
+       dev->needed_headroom = t_hlen + hlen;
+       mtu -= (dev->hard_header_len + t_hlen);
+
+       if (mtu < 68)
+               mtu = 68;
+
+       return mtu;
+}
+
+static struct ip_tunnel *ip_tunnel_create(struct net *net,
+                                         struct ip_tunnel_net *itn,
+                                         struct ip_tunnel_parm *parms)
+{
+       struct ip_tunnel *nt, *fbt;
+       struct net_device *dev;
+
+       BUG_ON(!itn->fb_tunnel_dev);
+       fbt = netdev_priv(itn->fb_tunnel_dev);
+       dev = __ip_tunnel_create(net, itn->fb_tunnel_dev->rtnl_link_ops, parms);
+       if (IS_ERR(dev))
+               return NULL;
+
+       dev->mtu = ip_tunnel_bind_dev(dev);
+
+       nt = netdev_priv(dev);
+       ip_tunnel_add(itn, nt);
+       return nt;
+}
+
+int ip_tunnel_rcv(struct ip_tunnel *tunnel, struct sk_buff *skb,
+                 const struct tnl_ptk_info *tpi, bool log_ecn_error)
+{
+       struct pcpu_tstats *tstats;
+       const struct iphdr *iph = ip_hdr(skb);
+       int err;
+
+       secpath_reset(skb);
+
+       skb->protocol = tpi->proto;
+
+       skb->mac_header = skb->network_header;
+       __pskb_pull(skb, tunnel->hlen);
+       skb_postpull_rcsum(skb, skb_transport_header(skb), tunnel->hlen);
+#ifdef CONFIG_NET_IPGRE_BROADCAST
+       if (ipv4_is_multicast(iph->daddr)) {
+               /* Looped back packet, drop it! */
+               if (rt_is_output_route(skb_rtable(skb)))
+                       goto drop;
+               tunnel->dev->stats.multicast++;
+               skb->pkt_type = PACKET_BROADCAST;
+       }
+#endif
+
+       if ((!(tpi->flags&TUNNEL_CSUM) &&  (tunnel->parms.i_flags&TUNNEL_CSUM)) ||
+            ((tpi->flags&TUNNEL_CSUM) && !(tunnel->parms.i_flags&TUNNEL_CSUM))) {
+               tunnel->dev->stats.rx_crc_errors++;
+               tunnel->dev->stats.rx_errors++;
+               goto drop;
+       }
+
+       if (tunnel->parms.i_flags&TUNNEL_SEQ) {
+               if (!(tpi->flags&TUNNEL_SEQ) ||
+                   (tunnel->i_seqno && (s32)(ntohl(tpi->seq) - tunnel->i_seqno) < 0)) {
+                       tunnel->dev->stats.rx_fifo_errors++;
+                       tunnel->dev->stats.rx_errors++;
+                       goto drop;
+               }
+               tunnel->i_seqno = ntohl(tpi->seq) + 1;
+       }
+
+       /* Warning: All skb pointers will be invalidated! */
+       if (tunnel->dev->type == ARPHRD_ETHER) {
+               if (!pskb_may_pull(skb, ETH_HLEN)) {
+                       tunnel->dev->stats.rx_length_errors++;
+                       tunnel->dev->stats.rx_errors++;
+                       goto drop;
+               }
+
+               iph = ip_hdr(skb);
+               skb->protocol = eth_type_trans(skb, tunnel->dev);
+               skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
+       }
+
+       skb->pkt_type = PACKET_HOST;
+       __skb_tunnel_rx(skb, tunnel->dev);
+
+       skb_reset_network_header(skb);
+       err = IP_ECN_decapsulate(iph, skb);
+       if (unlikely(err)) {
+               if (log_ecn_error)
+                       net_info_ratelimited("non-ECT from %pI4 with TOS=%#x\n",
+                                       &iph->saddr, iph->tos);
+               if (err > 1) {
+                       ++tunnel->dev->stats.rx_frame_errors;
+                       ++tunnel->dev->stats.rx_errors;
+                       goto drop;
+               }
+       }
+
+       tstats = this_cpu_ptr(tunnel->dev->tstats);
+       u64_stats_update_begin(&tstats->syncp);
+       tstats->rx_packets++;
+       tstats->rx_bytes += skb->len;
+       u64_stats_update_end(&tstats->syncp);
+
+       gro_cells_receive(&tunnel->gro_cells, skb);
+       return 0;
+
+drop:
+       kfree_skb(skb);
+       return 0;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_rcv);
+
+void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
+                   const struct iphdr *tnl_params)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       const struct iphdr *inner_iph;
+       struct iphdr *iph;
+       struct flowi4 fl4;
+       u8     tos, ttl;
+       __be16 df;
+       struct rtable *rt;              /* Route to the other host */
+       struct net_device *tdev;        /* Device to other host */
+       unsigned int max_headroom;      /* The extra header space needed */
+       __be32 dst;
+       int mtu;
+
+       inner_iph = (const struct iphdr *)skb_inner_network_header(skb);
+
+       dst = tnl_params->daddr;
+       if (dst == 0) {
+               /* NBMA tunnel */
+
+               if (skb_dst(skb) == NULL) {
+                       dev->stats.tx_fifo_errors++;
+                       goto tx_error;
+               }
+
+               if (skb->protocol == htons(ETH_P_IP)) {
+                       rt = skb_rtable(skb);
+                       dst = rt_nexthop(rt, inner_iph->daddr);
+               }
+#if IS_ENABLED(CONFIG_IPV6)
+               else if (skb->protocol == htons(ETH_P_IPV6)) {
+                       const struct in6_addr *addr6;
+                       struct neighbour *neigh;
+                       bool do_tx_error_icmp;
+                       int addr_type;
+
+                       neigh = dst_neigh_lookup(skb_dst(skb),
+                                                &ipv6_hdr(skb)->daddr);
+                       if (neigh == NULL)
+                               goto tx_error;
+
+                       addr6 = (const struct in6_addr *)&neigh->primary_key;
+                       addr_type = ipv6_addr_type(addr6);
+
+                       if (addr_type == IPV6_ADDR_ANY) {
+                               addr6 = &ipv6_hdr(skb)->daddr;
+                               addr_type = ipv6_addr_type(addr6);
+                       }
+
+                       if ((addr_type & IPV6_ADDR_COMPATv4) == 0)
+                               do_tx_error_icmp = true;
+                       else {
+                               do_tx_error_icmp = false;
+                               dst = addr6->s6_addr32[3];
+                       }
+                       neigh_release(neigh);
+                       if (do_tx_error_icmp)
+                               goto tx_error_icmp;
+               }
+#endif
+               else
+                       goto tx_error;
+       }
+
+       tos = tnl_params->tos;
+       if (tos & 0x1) {
+               tos &= ~0x1;
+               if (skb->protocol == htons(ETH_P_IP))
+                       tos = inner_iph->tos;
+               else if (skb->protocol == htons(ETH_P_IPV6))
+                       tos = ipv6_get_dsfield((const struct ipv6hdr *)inner_iph);
+       }
+
+       rt = ip_route_output_tunnel(dev_net(dev), &fl4,
+                                   tunnel->parms.iph.protocol,
+                                   dst, tnl_params->saddr,
+                                   tunnel->parms.o_key,
+                                   RT_TOS(tos),
+                                   tunnel->parms.link);
+       if (IS_ERR(rt)) {
+               dev->stats.tx_carrier_errors++;
+               goto tx_error;
+       }
+       tdev = rt->dst.dev;
+
+       if (tdev == dev) {
+               ip_rt_put(rt);
+               dev->stats.collisions++;
+               goto tx_error;
+       }
+
+       df = tnl_params->frag_off;
+
+       if (df)
+               mtu = dst_mtu(&rt->dst) - dev->hard_header_len
+                                       - sizeof(struct iphdr);
+       else
+               mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
+
+       if (skb_dst(skb))
+               skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu);
+
+       if (skb->protocol == htons(ETH_P_IP)) {
+               df |= (inner_iph->frag_off&htons(IP_DF));
+
+               if (!skb_is_gso(skb) &&
+                   (inner_iph->frag_off&htons(IP_DF)) &&
+                    mtu < ntohs(inner_iph->tot_len)) {
+                       icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
+                       ip_rt_put(rt);
+                       goto tx_error;
+               }
+       }
+#if IS_ENABLED(CONFIG_IPV6)
+       else if (skb->protocol == htons(ETH_P_IPV6)) {
+               struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
+
+               if (rt6 && mtu < dst_mtu(skb_dst(skb)) &&
+                   mtu >= IPV6_MIN_MTU) {
+                       if ((tunnel->parms.iph.daddr &&
+                           !ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
+                           rt6->rt6i_dst.plen == 128) {
+                               rt6->rt6i_flags |= RTF_MODIFIED;
+                               dst_metric_set(skb_dst(skb), RTAX_MTU, mtu);
+                       }
+               }
+
+               if (!skb_is_gso(skb) && mtu >= IPV6_MIN_MTU &&
+                   mtu < skb->len) {
+                       icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+                       ip_rt_put(rt);
+                       goto tx_error;
+               }
+       }
+#endif
+
+       if (tunnel->err_count > 0) {
+               if (time_before(jiffies,
+                               tunnel->err_time + IPTUNNEL_ERR_TIMEO)) {
+                       tunnel->err_count--;
+
+                       dst_link_failure(skb);
+               } else
+                       tunnel->err_count = 0;
+       }
+
+       ttl = tnl_params->ttl;
+       if (ttl == 0) {
+               if (skb->protocol == htons(ETH_P_IP))
+                       ttl = inner_iph->ttl;
+#if IS_ENABLED(CONFIG_IPV6)
+               else if (skb->protocol == htons(ETH_P_IPV6))
+                       ttl = ((const struct ipv6hdr *)inner_iph)->hop_limit;
+#endif
+               else
+                       ttl = ip4_dst_hoplimit(&rt->dst);
+       }
+
+       max_headroom = LL_RESERVED_SPACE(tdev) + sizeof(struct iphdr)
+                                              + rt->dst.header_len;
+       if (max_headroom > dev->needed_headroom) {
+               dev->needed_headroom = max_headroom;
+               if (skb_cow_head(skb, dev->needed_headroom)) {
+                       dev->stats.tx_dropped++;
+                       dev_kfree_skb(skb);
+                       return;
+               }
+       }
+
+       skb_dst_drop(skb);
+       skb_dst_set(skb, &rt->dst);
+       memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
+
+       /* Push down and install the IP header. */
+       skb_push(skb, sizeof(struct iphdr));
+       skb_reset_network_header(skb);
+
+       iph = ip_hdr(skb);
+       inner_iph = (const struct iphdr *)skb_inner_network_header(skb);
+
+       iph->version    =       4;
+       iph->ihl        =       sizeof(struct iphdr) >> 2;
+       iph->frag_off   =       df;
+       iph->protocol   =       tnl_params->protocol;
+       iph->tos        =       ip_tunnel_ecn_encap(tos, inner_iph, skb);
+       iph->daddr      =       fl4.daddr;
+       iph->saddr      =       fl4.saddr;
+       iph->ttl        =       ttl;
+       tunnel_ip_select_ident(skb, inner_iph, &rt->dst);
+
+       iptunnel_xmit(skb, dev);
+       return;
+
+#if IS_ENABLED(CONFIG_IPV6)
+tx_error_icmp:
+       dst_link_failure(skb);
+#endif
+tx_error:
+       dev->stats.tx_errors++;
+       dev_kfree_skb(skb);
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_xmit);
+
+static void ip_tunnel_update(struct ip_tunnel_net *itn,
+                            struct ip_tunnel *t,
+                            struct net_device *dev,
+                            struct ip_tunnel_parm *p,
+                            bool set_mtu)
+{
+       ip_tunnel_del(t);
+       t->parms.iph.saddr = p->iph.saddr;
+       t->parms.iph.daddr = p->iph.daddr;
+       t->parms.i_key = p->i_key;
+       t->parms.o_key = p->o_key;
+       if (dev->type != ARPHRD_ETHER) {
+               memcpy(dev->dev_addr, &p->iph.saddr, 4);
+               memcpy(dev->broadcast, &p->iph.daddr, 4);
+       }
+       ip_tunnel_add(itn, t);
+
+       t->parms.iph.ttl = p->iph.ttl;
+       t->parms.iph.tos = p->iph.tos;
+       t->parms.iph.frag_off = p->iph.frag_off;
+
+       if (t->parms.link != p->link) {
+               int mtu;
+
+               t->parms.link = p->link;
+               mtu = ip_tunnel_bind_dev(dev);
+               if (set_mtu)
+                       dev->mtu = mtu;
+       }
+       netdev_state_change(dev);
+}
+
+int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd)
+{
+       int err = 0;
+       struct ip_tunnel *t;
+       struct net *net = dev_net(dev);
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct ip_tunnel_net *itn = net_generic(net, tunnel->ip_tnl_net_id);
+
+       BUG_ON(!itn->fb_tunnel_dev);
+       switch (cmd) {
+       case SIOCGETTUNNEL:
+               t = NULL;
+               if (dev == itn->fb_tunnel_dev)
+                       t = ip_tunnel_find(itn, p, itn->fb_tunnel_dev->type);
+               if (t == NULL)
+                       t = netdev_priv(dev);
+               memcpy(p, &t->parms, sizeof(*p));
+               break;
+
+       case SIOCADDTUNNEL:
+       case SIOCCHGTUNNEL:
+               err = -EPERM;
+               if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
+                       goto done;
+               if (p->iph.ttl)
+                       p->iph.frag_off |= htons(IP_DF);
+               if (!(p->i_flags&TUNNEL_KEY))
+                       p->i_key = 0;
+               if (!(p->o_flags&TUNNEL_KEY))
+                       p->o_key = 0;
+
+               t = ip_tunnel_find(itn, p, itn->fb_tunnel_dev->type);
+
+               if (!t && (cmd == SIOCADDTUNNEL))
+                       t = ip_tunnel_create(net, itn, p);
+
+               if (dev != itn->fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
+                       if (t != NULL) {
+                               if (t->dev != dev) {
+                                       err = -EEXIST;
+                                       break;
+                               }
+                       } else {
+                               unsigned int nflags = 0;
+
+                               if (ipv4_is_multicast(p->iph.daddr))
+                                       nflags = IFF_BROADCAST;
+                               else if (p->iph.daddr)
+                                       nflags = IFF_POINTOPOINT;
+
+                               if ((dev->flags^nflags)&(IFF_POINTOPOINT|IFF_BROADCAST)) {
+                                       err = -EINVAL;
+                                       break;
+                               }
+
+                               t = netdev_priv(dev);
+                       }
+               }
+
+               if (t) {
+                       err = 0;
+                       ip_tunnel_update(itn, t, dev, p, true);
+               } else
+                       err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT);
+               break;
+
+       case SIOCDELTUNNEL:
+               err = -EPERM;
+               if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
+                       goto done;
+
+               if (dev == itn->fb_tunnel_dev) {
+                       err = -ENOENT;
+                       t = ip_tunnel_find(itn, p, itn->fb_tunnel_dev->type);
+                       if (t == NULL)
+                               goto done;
+                       err = -EPERM;
+                       if (t == netdev_priv(itn->fb_tunnel_dev))
+                               goto done;
+                       dev = t->dev;
+               }
+               unregister_netdevice(dev);
+               err = 0;
+               break;
+
+       default:
+               err = -EINVAL;
+       }
+
+done:
+       return err;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_ioctl);
+
+int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       int t_hlen = tunnel->hlen + sizeof(struct iphdr);
+
+       if (new_mtu < 68 ||
+           new_mtu > 0xFFF8 - dev->hard_header_len - t_hlen)
+               return -EINVAL;
+       dev->mtu = new_mtu;
+       return 0;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_change_mtu);
+
+static void ip_tunnel_dev_free(struct net_device *dev)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+
+       gro_cells_destroy(&tunnel->gro_cells);
+       free_percpu(dev->tstats);
+       free_netdev(dev);
+}
+
+void ip_tunnel_dellink(struct net_device *dev, struct list_head *head)
+{
+       struct net *net = dev_net(dev);
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct ip_tunnel_net *itn;
+
+       itn = net_generic(net, tunnel->ip_tnl_net_id);
+
+       if (itn->fb_tunnel_dev != dev) {
+               ip_tunnel_del(netdev_priv(dev));
+               unregister_netdevice_queue(dev, head);
+       }
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_dellink);
+
+int __net_init ip_tunnel_init_net(struct net *net, int ip_tnl_net_id,
+                                 struct rtnl_link_ops *ops, char *devname)
+{
+       struct ip_tunnel_net *itn = net_generic(net, ip_tnl_net_id);
+       struct ip_tunnel_parm parms;
+
+       itn->tunnels = kzalloc(IP_TNL_HASH_SIZE * sizeof(struct hlist_head), GFP_KERNEL);
+       if (!itn->tunnels)
+               return -ENOMEM;
+
+       if (!ops) {
+               itn->fb_tunnel_dev = NULL;
+               return 0;
+       }
+       memset(&parms, 0, sizeof(parms));
+       if (devname)
+               strlcpy(parms.name, devname, IFNAMSIZ);
+
+       rtnl_lock();
+       itn->fb_tunnel_dev = __ip_tunnel_create(net, ops, &parms);
+       rtnl_unlock();
+       if (IS_ERR(itn->fb_tunnel_dev)) {
+               kfree(itn->tunnels);
+               return PTR_ERR(itn->fb_tunnel_dev);
+       }
+
+       return 0;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_init_net);
+
+static void ip_tunnel_destroy(struct ip_tunnel_net *itn, struct list_head *head)
+{
+       int h;
+
+       for (h = 0; h < IP_TNL_HASH_SIZE; h++) {
+               struct ip_tunnel *t;
+               struct hlist_node *n;
+               struct hlist_head *thead = &itn->tunnels[h];
+
+               hlist_for_each_entry_safe(t, n, thead, hash_node)
+                       unregister_netdevice_queue(t->dev, head);
+       }
+       if (itn->fb_tunnel_dev)
+               unregister_netdevice_queue(itn->fb_tunnel_dev, head);
+}
+
+void __net_exit ip_tunnel_delete_net(struct ip_tunnel_net *itn)
+{
+       LIST_HEAD(list);
+
+       rtnl_lock();
+       ip_tunnel_destroy(itn, &list);
+       unregister_netdevice_many(&list);
+       rtnl_unlock();
+       kfree(itn->tunnels);
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_delete_net);
+
+int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[],
+                     struct ip_tunnel_parm *p)
+{
+       struct ip_tunnel *nt;
+       struct net *net = dev_net(dev);
+       struct ip_tunnel_net *itn;
+       int mtu;
+       int err;
+
+       nt = netdev_priv(dev);
+       itn = net_generic(net, nt->ip_tnl_net_id);
+
+       if (ip_tunnel_find(itn, p, dev->type))
+               return -EEXIST;
+
+       nt->parms = *p;
+       err = register_netdevice(dev);
+       if (err)
+               goto out;
+
+       if (dev->type == ARPHRD_ETHER && !tb[IFLA_ADDRESS])
+               eth_hw_addr_random(dev);
+
+       mtu = ip_tunnel_bind_dev(dev);
+       if (!tb[IFLA_MTU])
+               dev->mtu = mtu;
+
+       ip_tunnel_add(itn, nt);
+
+out:
+       return err;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_newlink);
+
+int ip_tunnel_changelink(struct net_device *dev, struct nlattr *tb[],
+                        struct ip_tunnel_parm *p)
+{
+       struct ip_tunnel *t, *nt;
+       struct net *net = dev_net(dev);
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct ip_tunnel_net *itn = net_generic(net, tunnel->ip_tnl_net_id);
+
+       if (dev == itn->fb_tunnel_dev)
+               return -EINVAL;
+
+       nt = netdev_priv(dev);
+
+       t = ip_tunnel_find(itn, p, dev->type);
+
+       if (t) {
+               if (t->dev != dev)
+                       return -EEXIST;
+       } else {
+               t = nt;
+
+               if (dev->type != ARPHRD_ETHER) {
+                       unsigned int nflags = 0;
+
+                       if (ipv4_is_multicast(p->iph.daddr))
+                               nflags = IFF_BROADCAST;
+                       else if (p->iph.daddr)
+                               nflags = IFF_POINTOPOINT;
+
+                       if ((dev->flags ^ nflags) &
+                           (IFF_POINTOPOINT | IFF_BROADCAST))
+                               return -EINVAL;
+               }
+       }
+
+       ip_tunnel_update(itn, t, dev, p, !tb[IFLA_MTU]);
+       return 0;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_changelink);
+
+int ip_tunnel_init(struct net_device *dev)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct iphdr *iph = &tunnel->parms.iph;
+       int err;
+
+       dev->destructor = ip_tunnel_dev_free;
+       dev->tstats = alloc_percpu(struct pcpu_tstats);
+       if (!dev->tstats)
+               return -ENOMEM;
+
+       err = gro_cells_init(&tunnel->gro_cells, dev);
+       if (err) {
+               free_percpu(dev->tstats);
+               return err;
+       }
+
+       tunnel->dev = dev;
+       strcpy(tunnel->parms.name, dev->name);
+       iph->version            = 4;
+       iph->ihl                = 5;
+
+       return 0;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_init);
+
+void ip_tunnel_uninit(struct net_device *dev)
+{
+       struct net *net = dev_net(dev);
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       struct ip_tunnel_net *itn;
+
+       itn = net_generic(net, tunnel->ip_tnl_net_id);
+       /* fb_tunnel_dev will be unregisted in net-exit call. */
+       if (itn->fb_tunnel_dev != dev)
+               ip_tunnel_del(netdev_priv(dev));
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_uninit);
+
+/* Do least required initialization, rest of init is done in tunnel_init call */
+void ip_tunnel_setup(struct net_device *dev, int net_id)
+{
+       struct ip_tunnel *tunnel = netdev_priv(dev);
+       tunnel->ip_tnl_net_id = net_id;
+}
+EXPORT_SYMBOL_GPL(ip_tunnel_setup);
+
+MODULE_LICENSE("GPL");
index c3a4233c0ac290394c819882aeff5ba3800378f4..6a628fb3349f1fd3a1d08c229446f792d2accf12 100644 (file)
@@ -38,7 +38,7 @@
 #include <net/sock.h>
 #include <net/ip.h>
 #include <net/icmp.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/inet_ecn.h>
 #include <net/xfrm.h>
 #include <net/net_namespace.h>
index 34e006fe2d870a57607e3d3abc51747e39ec6992..a557d6ab127ae64ab1ee10528425184a639d3464 100644 (file)
 #include <net/sock.h>
 #include <net/ip.h>
 #include <net/icmp.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/inet_ecn.h>
 #include <net/xfrm.h>
 #include <net/net_namespace.h>
index 5f95b3aa579ef03f01ca4a1c5900eb16eecff512..fd61fe16679f631f1ed9664fad5b656b7909aa40 100644 (file)
@@ -61,7 +61,7 @@
 #include <linux/netfilter_ipv4.h>
 #include <linux/compat.h>
 #include <linux/export.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/checksum.h>
 #include <net/netlink.h>
 #include <net/fib_rules.h>
index f56277f159030a272eb3baf3e0e4823f2761b8ed..ab5c7ad482cded367c1180a3a3761615bc2a4d54 100644 (file)
@@ -49,7 +49,6 @@
 #include <net/udp.h>
 #include <net/udplite.h>
 #include <net/tcp.h>
-#include <net/ipip.h>
 #include <net/protocol.h>
 #include <net/inet_common.h>
 #include <net/route.h>
index 6a6ba73ff26582e875d04007534cce1f54ea5a4d..df89ccaaceaa31ead4266dc874735e5302133117 100644 (file)
@@ -38,6 +38,7 @@
 
 #include <net/sock.h>
 #include <net/ip.h>
+#include <net/ip_tunnels.h>
 #include <net/icmp.h>
 #include <net/protocol.h>
 #include <net/addrconf.h>
index bef3fedfdc56e1e075e722f2877c905d3726b763..1e55866cead7425eefdff36c1ddca1aab9504286 100644 (file)
@@ -47,6 +47,7 @@
 
 #include <net/icmp.h>
 #include <net/ip.h>
+#include <net/ip_tunnels.h>
 #include <net/ipv6.h>
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
index 898e671a526bc4ad9ff906f90065487e9e3f72bc..ee4fc570cf2c664cb08d9ed6d418b12f64aa6605 100644 (file)
@@ -49,7 +49,7 @@
 #include <net/ip.h>
 #include <net/udp.h>
 #include <net/icmp.h>
-#include <net/ipip.h>
+#include <net/ip_tunnels.h>
 #include <net/inet_ecn.h>
 #include <net/xfrm.h>
 #include <net/dsfield.h>