datapath: Add support for lwtunnel

Following patch adds support for lwtunnel to OVS datapath.
With this change OVS datapath detect lwtunnel support and
make use of new APIs if available. On older kernel where the
support is not there the backported tunnel modules are used.
These backported tunnel devices acts as lwtunnel devices.
I tried to keep backported module same as upstream for easier
bug-fix backport. Since STT and LISP are not upstream OVS
always needs to use respective modules from tunnel compat layer.
To make it work on kernel 4.3 I have converted STT and LISP
modules to lwtunnel API model.

lwtunnel make use of skb-dst to pass tunnel information to the
tunnel module. On older kernel this is not possible. So the in
case of old kernel metadata ref is stored in OVS_CB and direct
call to tunnel transmit function is made by respective tunnel
vport modules. Similarly on receive side tunnel recv directly
call netdev-vport-receive to pass the skb to OVS.

Major backported components include:
Geneve, GRE, VXLAN, ip_tunnel, udp-tunnels GRO.

Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Acked-by: Joe Stringer <joe@ovn.org>
Acked-by: Jesse Gross <jesse@kernel.org>

1 files changed, 92 insertions, 124 deletions

diff --git a/datapath/vport-netdev.c b/datapath/vport-netdev.c
index 6c8373740..21431d3c3 100644
--- a/datapath/vport-netdev.c
+++ b/datapath/vport-netdev.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2007-2012 Nicira, Inc.
+ * Copyright (c) 2007-2015 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
@@ -26,32 +26,61 @@
 #include <linux/rtnetlink.h>
 #include <linux/skbuff.h>
 #include <linux/openvswitch.h>
-#include <linux/netdevice.h>
 
-#include <net/llc.h>
+#include <net/ip_tunnels.h>
+#include <net/rtnetlink.h>
 
 #include "datapath.h"
-#include "vlan.h"
+#include "gso.h"
+#include "vport.h"
 #include "vport-internal_dev.h"
 #include "vport-netdev.h"
 
 static struct vport_ops ovs_netdev_vport_ops;
-static void netdev_port_receive(struct vport *vport, struct sk_buff *skb);
+
+/* Must be called with rcu_read_lock. */
+void netdev_port_receive(struct sk_buff *skb, struct ip_tunnel_info *tun_info)
+{
+	struct vport *vport;
+
+	vport = ovs_netdev_get_vport(skb->dev);
+	if (unlikely(!vport))
+		goto error;
+
+	if (unlikely(skb_warn_if_lro(skb)))
+		goto error;
+
+	/* Make our own copy of the packet.  Otherwise we will mangle the
+	 * packet for anyone who came before us (e.g. tcpdump via AF_PACKET).
+	 */
+	skb = skb_share_check(skb, GFP_ATOMIC);
+	if (unlikely(!skb))
+		return;
+
+	skb_push(skb, ETH_HLEN);
+	ovs_skb_postpush_rcsum(skb, skb->data, ETH_HLEN);
+	ovs_vport_receive(vport, skb, tun_info);
+	return;
+error:
+	kfree_skb(skb);
+}
+
+#ifndef HAVE_METADATA_DST
+#define port_receive(skb)  netdev_port_receive(skb, NULL)
+#else
+#define port_receive(skb)  netdev_port_receive(skb, skb_tunnel_info(skb))
+#endif
 
 #if defined HAVE_RX_HANDLER_PSKB  /* 2.6.39 and above or backports */
 /* Called with rcu_read_lock and bottom-halves disabled. */
 static rx_handler_result_t netdev_frame_hook(struct sk_buff **pskb)
 {
 	struct sk_buff *skb = *pskb;
-	struct vport *vport;
 
 	if (unlikely(skb->pkt_type == PACKET_LOOPBACK))
 		return RX_HANDLER_PASS;
 
-	vport = ovs_netdev_get_vport(skb->dev);
-
-	netdev_port_receive(vport, skb);
-
+	port_receive(skb);
 	return RX_HANDLER_CONSUMED;
 }
 #elif LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,36) || \
@@ -59,15 +88,10 @@ static rx_handler_result_t netdev_frame_hook(struct sk_buff **pskb)
 /* Called with rcu_read_lock and bottom-halves disabled. */
 static struct sk_buff *netdev_frame_hook(struct sk_buff *skb)
 {
-	struct vport *vport;
-
 	if (unlikely(skb->pkt_type == PACKET_LOOPBACK))
 		return skb;
 
-	vport = ovs_netdev_get_vport(skb->dev);
-
-	netdev_port_receive(vport, skb);
-
+	port_receive(skb);
 	return NULL;
 }
 #elif LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,32)
@@ -79,7 +103,7 @@ static struct sk_buff *netdev_frame_hook(struct sk_buff *skb)
 static struct sk_buff *netdev_frame_hook(struct net_bridge_port *p,
 					 struct sk_buff *skb)
 {
-	netdev_port_receive((struct vport *)p, skb);
+	port_receive(skb);
 	return NULL;
 }
 #else
@@ -92,167 +116,112 @@ static struct net_device *get_dpdev(const struct datapath *dp)
 
 	local = ovs_vport_ovsl(dp, OVSP_LOCAL);
 	BUG_ON(!local);
-	return netdev_vport_priv(local)->dev;
+	return local->dev;
 }
 
-static struct vport *netdev_create(const struct vport_parms *parms)
+struct vport *ovs_netdev_link(struct vport *vport, const char *name)
 {
-	struct vport *vport;
-	struct netdev_vport *netdev_vport;
 	int err;
 
-	vport = ovs_vport_alloc(sizeof(struct netdev_vport),
-				&ovs_netdev_vport_ops, parms);
-	if (IS_ERR(vport)) {
-		err = PTR_ERR(vport);
-		goto error;
-	}
-
-	netdev_vport = netdev_vport_priv(vport);
-
-	netdev_vport->dev = dev_get_by_name(ovs_dp_get_net(vport->dp), parms->name);
-	if (!netdev_vport->dev) {
+	vport->dev = dev_get_by_name(ovs_dp_get_net(vport->dp), name);
+	if (!vport->dev) {
 		err = -ENODEV;
 		goto error_free_vport;
 	}
 
-	if (netdev_vport->dev->flags & IFF_LOOPBACK ||
-	    netdev_vport->dev->type != ARPHRD_ETHER ||
-	    ovs_is_internal_dev(netdev_vport->dev)) {
+	if (vport->dev->flags & IFF_LOOPBACK ||
+	    vport->dev->type != ARPHRD_ETHER ||
+	    ovs_is_internal_dev(vport->dev)) {
 		err = -EINVAL;
 		goto error_put;
 	}
 
 	rtnl_lock();
-	err = netdev_master_upper_dev_link(netdev_vport->dev,
+	err = netdev_master_upper_dev_link(vport->dev,
 					   get_dpdev(vport->dp));
 	if (err)
 		goto error_unlock;
 
-	err = netdev_rx_handler_register(netdev_vport->dev, netdev_frame_hook,
+	err = netdev_rx_handler_register(vport->dev, netdev_frame_hook,
 					 vport);
 	if (err)
 		goto error_master_upper_dev_unlink;
 
-	dev_disable_lro(netdev_vport->dev);
-	dev_set_promiscuity(netdev_vport->dev, 1);
-	netdev_vport->dev->priv_flags |= IFF_OVS_DATAPATH;
+	dev_disable_lro(vport->dev);
+	dev_set_promiscuity(vport->dev, 1);
+	vport->dev->priv_flags |= IFF_OVS_DATAPATH;
 	rtnl_unlock();
 
 	return vport;
 
 error_master_upper_dev_unlink:
-	netdev_upper_dev_unlink(netdev_vport->dev, get_dpdev(vport->dp));
+	netdev_upper_dev_unlink(vport->dev, get_dpdev(vport->dp));
 error_unlock:
 	rtnl_unlock();
 error_put:
-	dev_put(netdev_vport->dev);
+	dev_put(vport->dev);
 error_free_vport:
 	ovs_vport_free(vport);
-error:
 	return ERR_PTR(err);
 }
+EXPORT_SYMBOL_GPL(ovs_netdev_link);
+
+static struct vport *netdev_create(const struct vport_parms *parms)
+{
+	struct vport *vport;
+
+	vport = ovs_vport_alloc(0, &ovs_netdev_vport_ops, parms);
+	if (IS_ERR(vport))
+		return vport;
+
+	return ovs_netdev_link(vport, parms->name);
+}
 
-static void free_port_rcu(struct rcu_head *rcu)
+static void vport_netdev_free(struct rcu_head *rcu)
 {
-	struct netdev_vport *netdev_vport = container_of(rcu,
-					struct netdev_vport, rcu);
+	struct vport *vport = container_of(rcu, struct vport, rcu);
 
-	dev_put(netdev_vport->dev);
-	ovs_vport_free(vport_from_priv(netdev_vport));
+	if (vport->dev)
+		dev_put(vport->dev);
+	ovs_vport_free(vport);
 }
 
 void ovs_netdev_detach_dev(struct vport *vport)
 {
-	struct netdev_vport *netdev_vport = netdev_vport_priv(vport);
-
 	ASSERT_RTNL();
-	netdev_vport->dev->priv_flags &= ~IFF_OVS_DATAPATH;
-	netdev_rx_handler_unregister(netdev_vport->dev);
-	netdev_upper_dev_unlink(netdev_vport->dev,
-				netdev_master_upper_dev_get(netdev_vport->dev));
-	dev_set_promiscuity(netdev_vport->dev, -1);
+	vport->dev->priv_flags &= ~IFF_OVS_DATAPATH;
+	netdev_rx_handler_unregister(vport->dev);
+	netdev_upper_dev_unlink(vport->dev,
+				netdev_master_upper_dev_get(vport->dev));
+	dev_set_promiscuity(vport->dev, -1);
 }
+EXPORT_SYMBOL_GPL(ovs_netdev_detach_dev);
 
 static void netdev_destroy(struct vport *vport)
 {
-	struct netdev_vport *netdev_vport = netdev_vport_priv(vport);
-
 	rtnl_lock();
-	if (ovs_netdev_get_vport(netdev_vport->dev))
+	if (vport->dev->priv_flags & IFF_OVS_DATAPATH)
 		ovs_netdev_detach_dev(vport);
 	rtnl_unlock();
 
-	call_rcu(&netdev_vport->rcu, free_port_rcu);
-}
-
-const char *ovs_netdev_get_name(const struct vport *vport)
-{
-	const struct netdev_vport *netdev_vport = netdev_vport_priv(vport);
-	return netdev_vport->dev->name;
-}
-
-/* Must be called with rcu_read_lock. */
-static void netdev_port_receive(struct vport *vport, struct sk_buff *skb)
-{
-	if (unlikely(!vport))
-		goto error;
-
-	if (unlikely(skb_warn_if_lro(skb)))
-		goto error;
-
-	/* Make our own copy of the packet.  Otherwise we will mangle the
-	 * packet for anyone who came before us (e.g. tcpdump via AF_PACKET).
-	 * (No one comes after us, since we tell handle_bridge() that we took
-	 * the packet.)
-	 */
-	skb = skb_share_check(skb, GFP_ATOMIC);
-	if (unlikely(!skb))
-		return;
-
-	skb_push(skb, ETH_HLEN);
-	ovs_skb_postpush_rcsum(skb, skb->data, ETH_HLEN);
-
-	ovs_vport_receive(vport, skb, NULL);
-	return;
-
-error:
-	kfree_skb(skb);
+	call_rcu(&vport->rcu, vport_netdev_free);
 }
 
-static unsigned int packet_length(const struct sk_buff *skb)
+void ovs_netdev_tunnel_destroy(struct vport *vport)
 {
-	unsigned int length = skb->len - ETH_HLEN;
-
-	if (skb->protocol == htons(ETH_P_8021Q))
-		length -= VLAN_HLEN;
-
-	return length;
-}
-
-static int netdev_send(struct vport *vport, struct sk_buff *skb)
-{
-	struct netdev_vport *netdev_vport = netdev_vport_priv(vport);
-	int mtu = netdev_vport->dev->mtu;
-	int len;
-
-	if (unlikely(packet_length(skb) > mtu && !skb_is_gso(skb))) {
-		net_warn_ratelimited("%s: dropped over-mtu packet: %d > %d\n",
-				     netdev_vport->dev->name,
-				     packet_length(skb), mtu);
-		goto drop;
-	}
-
-	skb->dev = netdev_vport->dev;
-	len = skb->len;
-	dev_queue_xmit(skb);
+	rtnl_lock();
+	if (vport->dev->priv_flags & IFF_OVS_DATAPATH)
+		ovs_netdev_detach_dev(vport);
 
-	return len;
+	/* Early release so we can unregister the device */
+	dev_put(vport->dev);
+	rtnl_delete_link(vport->dev);
+	vport->dev = NULL;
+	rtnl_unlock();
 
-drop:
-	kfree_skb(skb);
-	return 0;
+	call_rcu(&vport->rcu, vport_netdev_free);
 }
+EXPORT_SYMBOL_GPL(ovs_netdev_tunnel_destroy);
 
 /* Returns null if this device is not attached to a datapath. */
 struct vport *ovs_netdev_get_vport(struct net_device *dev)
@@ -285,8 +254,7 @@ static struct vport_ops ovs_netdev_vport_ops = {
 	.type		= OVS_VPORT_TYPE_NETDEV,
 	.create		= netdev_create,
 	.destroy	= netdev_destroy,
-	.get_name	= ovs_netdev_get_name,
-	.send		= netdev_send,
+	.send		= dev_queue_xmit,
 };
 
 int __init ovs_netdev_init(void)