datapath: Move generic tunnel functions to lisp module.
authorPravin B Shelar <pshelar@nicira.com>
Tue, 13 Aug 2013 07:19:53 +0000 (00:19 -0700)
committerPravin B Shelar <pshelar@nicira.com>
Tue, 13 Aug 2013 07:19:53 +0000 (00:19 -0700)
Generic tunnel rcv and send function are only used by lisp tunneling
module, so It make sense to move them to lisp module.

CC: Lori Jakab <lojakab@cisco.com>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Lorand Jakab <lojakab@cisco.com>
Acked-by: Jesse Gross <jesse@nicira.com>
datapath/Modules.mk
datapath/datapath.c
datapath/datapath.h
datapath/dp_notify.c
datapath/tunnel.c [deleted file]
datapath/tunnel.h [deleted file]
datapath/vport-gre.c
datapath/vport-lisp.c
datapath/vport-vxlan.c

index 2ce8888..ccf4dfa 100644 (file)
@@ -12,7 +12,6 @@ openvswitch_sources = \
        datapath.c \
        dp_notify.c \
        flow.c \
-       tunnel.c \
        vlan.c \
        vport.c \
        vport-gre.c \
@@ -26,7 +25,6 @@ openvswitch_headers = \
        compat.h \
        datapath.h \
        flow.h \
-       tunnel.h \
        vlan.h \
        vport.h \
        vport-internal_dev.h \
index a6bf70c..48f17c0 100644 (file)
@@ -58,7 +58,6 @@
 #include "datapath.h"
 #include "flow.h"
 #include "vlan.h"
-#include "tunnel.h"
 #include "vport-internal_dev.h"
 #include "vport-netdev.h"
 
index eda87fd..064211d 100644 (file)
@@ -29,7 +29,6 @@
 #include "checksum.h"
 #include "compat.h"
 #include "flow.h"
-#include "tunnel.h"
 #include "vlan.h"
 #include "vport.h"
 
index ec573a5..d530893 100644 (file)
@@ -18,6 +18,8 @@
 
 #include <linux/netdevice.h>
 #include <net/genetlink.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
 
 #include "datapath.h"
 #include "vport-internal_dev.h"
diff --git a/datapath/tunnel.c b/datapath/tunnel.c
deleted file mode 100644 (file)
index c2c3df4..0000000
+++ /dev/null
@@ -1,286 +0,0 @@
-/*
- * Copyright (c) 2007-2012 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/in.h>
-#include <linux/in_route.h>
-#include <linux/inetdevice.h>
-#include <linux/jhash.h>
-#include <linux/list.h>
-#include <linux/kernel.h>
-#include <linux/version.h>
-#include <linux/workqueue.h>
-#include <linux/rculist.h>
-#include <net/route.h>
-#include <net/xfrm.h>
-
-#include "checksum.h"
-#include "compat.h"
-#include "datapath.h"
-#include "tunnel.h"
-#include "vlan.h"
-#include "vport.h"
-
-/**
- *     ovs_tnl_rcv - ingress point for generic tunnel code
- *
- * @vport: port this packet was received on
- * @skb: received packet
- * @tos: ToS from encapsulating IP packet, used to copy ECN bits
- *
- * Must be called with rcu_read_lock.
- *
- * Packets received by this function are in the following state:
- * - skb->data points to the inner Ethernet header.
- * - The inner Ethernet header is in the linear data area.
- * - skb->csum does not include the inner Ethernet header.
- * - The layer pointers are undefined.
- */
-void ovs_tnl_rcv(struct vport *vport, struct sk_buff *skb,
-                struct ovs_key_ipv4_tunnel *tun_key)
-{
-       struct ethhdr *eh;
-
-       skb_reset_mac_header(skb);
-       eh = eth_hdr(skb);
-
-       if (likely(ntohs(eh->h_proto) >= ETH_P_802_3_MIN))
-               skb->protocol = eh->h_proto;
-       else
-               skb->protocol = htons(ETH_P_802_2);
-
-       skb_dst_drop(skb);
-       nf_reset(skb);
-       skb_clear_rxhash(skb);
-       secpath_reset(skb);
-       vlan_set_tci(skb, 0);
-
-       if (unlikely(compute_ip_summed(skb, false))) {
-               kfree_skb(skb);
-               return;
-       }
-
-       ovs_vport_receive(vport, skb, tun_key);
-}
-
-static bool need_linearize(const struct sk_buff *skb)
-{
-       int i;
-
-       if (unlikely(skb_shinfo(skb)->frag_list))
-               return true;
-
-       /*
-        * Generally speaking we should linearize if there are paged frags.
-        * However, if all of the refcounts are 1 we know nobody else can
-        * change them from underneath us and we can skip the linearization.
-        */
-       for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
-               if (unlikely(page_count(skb_frag_page(&skb_shinfo(skb)->frags[i])) > 1))
-                       return true;
-
-       return false;
-}
-
-static struct sk_buff *handle_offloads(struct sk_buff *skb)
-{
-       int err;
-
-       forward_ip_summed(skb, true);
-
-       if (skb_is_gso(skb)) {
-               struct sk_buff *nskb;
-               char cb[sizeof(skb->cb)];
-
-               memcpy(cb, skb->cb, sizeof(cb));
-
-               nskb = __skb_gso_segment(skb, 0, false);
-               if (IS_ERR(nskb)) {
-                       err = PTR_ERR(nskb);
-                       goto error;
-               }
-
-               consume_skb(skb);
-               skb = nskb;
-               while (nskb) {
-                       memcpy(nskb->cb, cb, sizeof(cb));
-                       nskb = nskb->next;
-               }
-       } else if (get_ip_summed(skb) == OVS_CSUM_PARTIAL) {
-               /* Pages aren't locked and could change at any time.
-                * If this happens after we compute the checksum, the
-                * checksum will be wrong.  We linearize now to avoid
-                * this problem.
-                */
-               if (unlikely(need_linearize(skb))) {
-                       err = __skb_linearize(skb);
-                       if (unlikely(err))
-                               goto error;
-               }
-
-               err = skb_checksum_help(skb);
-               if (unlikely(err))
-                       goto error;
-       }
-
-       set_ip_summed(skb, OVS_CSUM_NONE);
-
-       return skb;
-
-error:
-       return ERR_PTR(err);
-}
-
-/* Compute source UDP port for outgoing packet.
- * Currently we use the flow hash.
- */
-u16 ovs_tnl_get_src_port(struct sk_buff *skb)
-{
-       int low;
-       int high;
-       unsigned int range;
-       struct sw_flow_key *pkt_key = OVS_CB(skb)->pkt_key;
-       u32 hash = jhash2((const u32 *)pkt_key,
-                         sizeof(*pkt_key) / sizeof(u32), 0);
-
-       inet_get_local_port_range(&low, &high);
-       range = (high - low) + 1;
-       return (((u64) hash * range) >> 32) + low;
-}
-
-int ovs_tnl_send(struct vport *vport, struct sk_buff *skb,
-                u8 ipproto, int tunnel_hlen,
-                void (*build_header)(const struct vport *,
-                                     struct sk_buff *,
-                                     int tunnel_hlen))
-{
-       int min_headroom;
-       struct rtable *rt;
-       __be32 saddr;
-       int sent_len = 0;
-       int err;
-       struct sk_buff *nskb;
-
-       /* Route lookup */
-       saddr = OVS_CB(skb)->tun_key->ipv4_src;
-       rt = find_route(ovs_dp_get_net(vport->dp),
-                       &saddr,
-                       OVS_CB(skb)->tun_key->ipv4_dst,
-                       ipproto,
-                       OVS_CB(skb)->tun_key->ipv4_tos,
-                       skb_get_mark(skb));
-       if (IS_ERR(rt)) {
-               err = PTR_ERR(rt);
-               goto error;
-       }
-
-       tunnel_hlen += sizeof(struct iphdr);
-
-       min_headroom = LL_RESERVED_SPACE(rt_dst(rt).dev) + rt_dst(rt).header_len
-                       + tunnel_hlen
-                       + (vlan_tx_tag_present(skb) ? VLAN_HLEN : 0);
-
-       if (skb_headroom(skb) < min_headroom || skb_header_cloned(skb)) {
-               int head_delta = SKB_DATA_ALIGN(min_headroom -
-                                               skb_headroom(skb) +
-                                               16);
-
-               err = pskb_expand_head(skb, max_t(int, head_delta, 0),
-                                       0, GFP_ATOMIC);
-               if (unlikely(err))
-                       goto err_free_rt;
-       }
-
-       /* Offloading */
-       nskb = handle_offloads(skb);
-       if (IS_ERR(nskb)) {
-               err = PTR_ERR(nskb);
-               goto err_free_rt;
-       }
-       skb = nskb;
-
-       /* Reset SKB */
-       nf_reset(skb);
-       secpath_reset(skb);
-       skb_dst_drop(skb);
-       skb_clear_rxhash(skb);
-
-       while (skb) {
-               struct sk_buff *next_skb = skb->next;
-               struct iphdr *iph;
-               int frag_len;
-
-               skb->next = NULL;
-
-               if (unlikely(vlan_deaccel_tag(skb)))
-                       goto next;
-
-               frag_len = skb->len;
-               skb_push(skb, tunnel_hlen);
-               skb_reset_network_header(skb);
-               skb_set_transport_header(skb, sizeof(struct iphdr));
-
-               if (next_skb)
-                       skb_dst_set(skb, dst_clone(&rt_dst(rt)));
-               else
-                       skb_dst_set(skb, &rt_dst(rt));
-
-               /* Push Tunnel header. */
-               build_header(vport, skb, tunnel_hlen);
-
-               /* Push IP header. */
-               iph = ip_hdr(skb);
-               iph->version    = 4;
-               iph->ihl        = sizeof(struct iphdr) >> 2;
-               iph->protocol   = ipproto;
-               iph->daddr      = OVS_CB(skb)->tun_key->ipv4_dst;
-               iph->saddr      = saddr;
-               iph->tos        = OVS_CB(skb)->tun_key->ipv4_tos;
-               iph->ttl        = OVS_CB(skb)->tun_key->ipv4_ttl;
-               iph->frag_off   = OVS_CB(skb)->tun_key->tun_flags &
-                                 TUNNEL_DONT_FRAGMENT ?  htons(IP_DF) : 0;
-               /*
-                * Allow our local IP stack to fragment the outer packet even
-                * if the DF bit is set as a last resort.  We also need to
-                * force selection of an IP ID here with __ip_select_ident(),
-                * as ip_select_ident() assumes a proper ID is not needed when
-                * when the DF bit is set.
-                */
-               skb->local_df = 1;
-               __ip_select_ident(iph, skb_dst(skb), 0);
-
-               memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
-
-               err = ip_local_out(skb);
-               if (unlikely(net_xmit_eval(err)))
-                       goto next;
-
-               sent_len += frag_len;
-
-next:
-               skb = next_skb;
-       }
-
-       return sent_len;
-
-err_free_rt:
-       ip_rt_put(rt);
-error:
-       return err;
-}
diff --git a/datapath/tunnel.h b/datapath/tunnel.h
deleted file mode 100644 (file)
index 7625dbe..0000000
+++ /dev/null
@@ -1,40 +0,0 @@
-/*
- * Copyright (c) 2007-2012 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
- */
-
-#ifndef TUNNEL_H
-#define TUNNEL_H 1
-
-#include <linux/version.h>
-#include <net/net_namespace.h>
-#include <net/netns/generic.h>
-
-#include "flow.h"
-#include "vport.h"
-
-u16 ovs_tnl_get_src_port(struct sk_buff *skb);
-
-int ovs_tnl_send(struct vport *vport, struct sk_buff *skb,
-                u8 ipproto, int tunnel_hlen,
-                void (*build_header)(const struct vport *,
-                                     struct sk_buff *,
-                                     int tunnel_hlen));
-
-void ovs_tnl_rcv(struct vport *vport, struct sk_buff *skb,
-                struct ovs_key_ipv4_tunnel *tun_key);
-
-#endif /* TUNNEL_H */
index 9bba42a..5af6dbe 100644 (file)
 #include <linux/kernel.h>
 #include <linux/workqueue.h>
 #include <linux/rculist.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
 #include <net/route.h>
 #include <net/xfrm.h>
 
-
 #include <net/icmp.h>
 #include <net/ip.h>
 #include <net/ip_tunnels.h>
@@ -46,7 +47,6 @@
 #include <net/protocol.h>
 
 #include "datapath.h"
-#include "tunnel.h"
 #include "vport.h"
 
 /* Returns the least-significant 32 bits of a __be64. */
index 6027a16..2f62d11 100644 (file)
 
 #include <net/icmp.h>
 #include <net/ip.h>
+#include <net/route.h>
 #include <net/udp.h>
+#include <net/xfrm.h>
 
 #include "datapath.h"
-#include "tunnel.h"
 #include "vport.h"
 
-
 /*
  *  LISP encapsulation header:
  *
@@ -160,6 +160,23 @@ static __be64 instance_id_to_tunnel_id(__u8 *iid)
 #endif
 }
 
+/* Compute source UDP port for outgoing packet.
+ * Currently we use the flow hash.
+ */
+static u16 ovs_tnl_get_src_port(struct sk_buff *skb)
+{
+       int low;
+       int high;
+       unsigned int range;
+       struct sw_flow_key *pkt_key = OVS_CB(skb)->pkt_key;
+       u32 hash = jhash2((const u32 *)pkt_key,
+                         sizeof(*pkt_key) / sizeof(u32), 0);
+
+       inet_get_local_port_range(&low, &high);
+       range = (high - low) + 1;
+       return (((u64) hash * range) >> 32) + low;
+}
+
 static void lisp_build_header(const struct vport *vport,
                              struct sk_buff *skb,
                              int tunnel_hlen)
@@ -189,6 +206,48 @@ static void lisp_build_header(const struct vport *vport,
        lisph->u2.word2.locator_status_bits = 1;
 }
 
+/**
+ *     ovs_tnl_rcv - ingress point for generic tunnel code
+ *
+ * @vport: port this packet was received on
+ * @skb: received packet
+ * @tos: ToS from encapsulating IP packet, used to copy ECN bits
+ *
+ * Must be called with rcu_read_lock.
+ *
+ * Packets received by this function are in the following state:
+ * - skb->data points to the inner Ethernet header.
+ * - The inner Ethernet header is in the linear data area.
+ * - skb->csum does not include the inner Ethernet header.
+ * - The layer pointers are undefined.
+ */
+static void ovs_tnl_rcv(struct vport *vport, struct sk_buff *skb,
+                       struct ovs_key_ipv4_tunnel *tun_key)
+{
+       struct ethhdr *eh;
+
+       skb_reset_mac_header(skb);
+       eh = eth_hdr(skb);
+
+       if (likely(ntohs(eh->h_proto) >= ETH_P_802_3_MIN))
+               skb->protocol = eh->h_proto;
+       else
+               skb->protocol = htons(ETH_P_802_2);
+
+       skb_dst_drop(skb);
+       nf_reset(skb);
+       skb_clear_rxhash(skb);
+       secpath_reset(skb);
+       vlan_set_tci(skb, 0);
+
+       if (unlikely(compute_ip_summed(skb, false))) {
+               kfree_skb(skb);
+               return;
+       }
+
+       ovs_vport_receive(vport, skb, tun_key);
+}
+
 /* Called with rcu_read_lock and BH disabled. */
 static int lisp_rcv(struct sock *sk, struct sk_buff *skb)
 {
@@ -361,6 +420,196 @@ error:
        return ERR_PTR(err);
 }
 
+static bool need_linearize(const struct sk_buff *skb)
+{
+       int i;
+
+       if (unlikely(skb_shinfo(skb)->frag_list))
+               return true;
+
+       /*
+        * Generally speaking we should linearize if there are paged frags.
+        * However, if all of the refcounts are 1 we know nobody else can
+        * change them from underneath us and we can skip the linearization.
+        */
+       for (i = 0; i < skb_shinfo(skb)->nr_frags; i++)
+               if (unlikely(page_count(skb_frag_page(&skb_shinfo(skb)->frags[i])) > 1))
+                       return true;
+
+       return false;
+}
+
+static struct sk_buff *handle_offloads(struct sk_buff *skb)
+{
+       int err;
+
+       forward_ip_summed(skb, true);
+
+
+       if (skb_is_gso(skb)) {
+               struct sk_buff *nskb;
+               char cb[sizeof(skb->cb)];
+
+               memcpy(cb, skb->cb, sizeof(cb));
+
+               nskb = __skb_gso_segment(skb, 0, false);
+               if (IS_ERR(nskb)) {
+                       err = PTR_ERR(nskb);
+                       goto error;
+               }
+
+               consume_skb(skb);
+               skb = nskb;
+               while (nskb) {
+                       memcpy(nskb->cb, cb, sizeof(cb));
+                       nskb = nskb->next;
+               }
+       } else if (get_ip_summed(skb) == OVS_CSUM_PARTIAL) {
+               /* Pages aren't locked and could change at any time.
+                * If this happens after we compute the checksum, the
+                * checksum will be wrong.  We linearize now to avoid
+                * this problem.
+                */
+               if (unlikely(need_linearize(skb))) {
+                       err = __skb_linearize(skb);
+                       if (unlikely(err))
+                               goto error;
+               }
+
+               err = skb_checksum_help(skb);
+               if (unlikely(err))
+                       goto error;
+       }
+
+       set_ip_summed(skb, OVS_CSUM_NONE);
+
+       return skb;
+
+error:
+       return ERR_PTR(err);
+}
+
+static int ovs_tnl_send(struct vport *vport, struct sk_buff *skb,
+                       u8 ipproto, int tunnel_hlen,
+                       void (*build_header)(const struct vport *,
+                                            struct sk_buff *,
+                                            int tunnel_hlen))
+{
+       int min_headroom;
+       struct rtable *rt;
+       __be32 saddr;
+       int sent_len = 0;
+       int err;
+       struct sk_buff *nskb;
+
+       /* Route lookup */
+       saddr = OVS_CB(skb)->tun_key->ipv4_src;
+       rt = find_route(ovs_dp_get_net(vport->dp),
+                       &saddr,
+                       OVS_CB(skb)->tun_key->ipv4_dst,
+                       ipproto,
+                       OVS_CB(skb)->tun_key->ipv4_tos,
+                       skb_get_mark(skb));
+       if (IS_ERR(rt)) {
+               err = PTR_ERR(rt);
+               goto error;
+       }
+
+       tunnel_hlen += sizeof(struct iphdr);
+
+       min_headroom = LL_RESERVED_SPACE(rt_dst(rt).dev) + rt_dst(rt).header_len
+                       + tunnel_hlen
+                       + (vlan_tx_tag_present(skb) ? VLAN_HLEN : 0);
+
+       if (skb_headroom(skb) < min_headroom || skb_header_cloned(skb)) {
+               int head_delta = SKB_DATA_ALIGN(min_headroom -
+                                               skb_headroom(skb) +
+                                               16);
+
+               err = pskb_expand_head(skb, max_t(int, head_delta, 0),
+                                       0, GFP_ATOMIC);
+               if (unlikely(err))
+                       goto err_free_rt;
+       }
+
+       /* Offloading */
+       nskb = handle_offloads(skb);
+       if (IS_ERR(nskb)) {
+               err = PTR_ERR(nskb);
+               goto err_free_rt;
+       }
+       skb = nskb;
+
+       /* Reset SKB */
+       nf_reset(skb);
+       secpath_reset(skb);
+       skb_dst_drop(skb);
+       skb_clear_rxhash(skb);
+
+       while (skb) {
+               struct sk_buff *next_skb = skb->next;
+               struct iphdr *iph;
+               int frag_len;
+
+               skb->next = NULL;
+
+               if (unlikely(vlan_deaccel_tag(skb)))
+                       goto next;
+
+               frag_len = skb->len;
+               skb_push(skb, tunnel_hlen);
+               skb_reset_network_header(skb);
+               skb_set_transport_header(skb, sizeof(struct iphdr));
+
+               if (next_skb)
+                       skb_dst_set(skb, dst_clone(&rt_dst(rt)));
+               else
+                       skb_dst_set(skb, &rt_dst(rt));
+
+               /* Push Tunnel header. */
+               build_header(vport, skb, tunnel_hlen);
+
+               /* Push IP header. */
+               iph = ip_hdr(skb);
+               iph->version    = 4;
+               iph->ihl        = sizeof(struct iphdr) >> 2;
+               iph->protocol   = ipproto;
+               iph->daddr      = OVS_CB(skb)->tun_key->ipv4_dst;
+               iph->saddr      = saddr;
+               iph->tos        = OVS_CB(skb)->tun_key->ipv4_tos;
+               iph->ttl        = OVS_CB(skb)->tun_key->ipv4_ttl;
+               iph->frag_off   = OVS_CB(skb)->tun_key->tun_flags &
+                                 TUNNEL_DONT_FRAGMENT ?  htons(IP_DF) : 0;
+               /*
+                * Allow our local IP stack to fragment the outer packet even
+                * if the DF bit is set as a last resort.  We also need to
+                * force selection of an IP ID here with __ip_select_ident(),
+                * as ip_select_ident() assumes a proper ID is not needed when
+                * when the DF bit is set.
+                */
+               skb->local_df = 1;
+               __ip_select_ident(iph, skb_dst(skb), 0);
+
+               memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
+
+               err = ip_local_out(skb);
+               if (unlikely(net_xmit_eval(err)))
+                       goto next;
+
+               sent_len += frag_len;
+
+next:
+               skb = next_skb;
+       }
+
+       return sent_len;
+
+err_free_rt:
+       ip_rt_put(rt);
+error:
+       return err;
+}
+
 static int lisp_tnl_send(struct vport *vport, struct sk_buff *skb)
 {
        int tnl_len;
index 8fdec16..f3ef947 100644 (file)
@@ -42,7 +42,6 @@
 #include <net/vxlan.h>
 
 #include "datapath.h"
-#include "tunnel.h"
 #include "vport.h"
 
 #define OVS_VXLAN_RCV_PRIORITY         8