Merge branch 'l3mdev_saddr_op'

David Ahern says:

====================
net: Add saddr op to l3mdev and vrf

First 2 patches are re-sends of patches that got lost in the ethosphere
Tuesday; they were part of the first round of l3mdev conversions.
Next 3 handle the source address lookup for raw and datagram sockets
bound to a VRF device.

The conversion to the get_saddr op also fixes locally originated TCP
packets showing up at the VRF device. The use of the FLOWI_FLAG_L3MDEV_SRC
flag in ip_route_connect_init was causing locally generated packets
to skip the VRF device.

v2
- rebased to top of net-next per device delete fix and hash based
  multipath patches
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 41 insertions, 4 deletions

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index b9450784ae06..b3374402c1ea 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1261,6 +1261,7 @@ struct net_device_ops {
  * @IFF_L3MDEV_MASTER: device is an L3 master device
  * @IFF_NO_QUEUE: device can run without qdisc attached
  * @IFF_OPENVSWITCH: device is a Open vSwitch master
+ * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device
  */
 enum netdev_priv_flags {
 	IFF_802_1Q_VLAN			= 1<<0,
@@ -1286,6 +1287,7 @@ enum netdev_priv_flags {
 	IFF_L3MDEV_MASTER		= 1<<20,
 	IFF_NO_QUEUE			= 1<<21,
 	IFF_OPENVSWITCH			= 1<<22,
+	IFF_L3MDEV_SLAVE		= 1<<23,
 };
 
 #define IFF_802_1Q_VLAN			IFF_802_1Q_VLAN
@@ -3830,6 +3832,11 @@ static inline bool netif_is_l3_master(const struct net_device *dev)
 	return dev->priv_flags & IFF_L3MDEV_MASTER;
 }
 
+static inline bool netif_is_l3_slave(const struct net_device *dev)
+{
+	return dev->priv_flags & IFF_L3MDEV_SLAVE;
+}
+
 static inline bool netif_is_bridge_master(const struct net_device *dev)
 {
 	return dev->priv_flags & IFF_EBRIDGE;
diff --git a/include/net/flow.h b/include/net/flow.h
index 9b85db85f13c..83969eebebf3 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -34,7 +34,7 @@ struct flowi_common {
 	__u8	flowic_flags;
 #define FLOWI_FLAG_ANYSRC		0x01
 #define FLOWI_FLAG_KNOWN_NH		0x02
-#define FLOWI_FLAG_VRFSRC		0x04
+#define FLOWI_FLAG_L3MDEV_SRC		0x04
 #define FLOWI_FLAG_SKIP_NH_OIF		0x08
 	__u32	flowic_secid;
 	struct flowi_tunnel flowic_tun_key;
diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h
index 7a51fd8d99e4..ac5c6e80586a 100644
--- a/include/net/ip_fib.h
+++ b/include/net/ip_fib.h
@@ -329,6 +329,8 @@ static inline int fib_multipath_hash(__be32 saddr, __be32 daddr)
 }
 
 void fib_select_multipath(struct fib_result *res, int hash);
+void fib_select_path(struct net *net, struct fib_result *res,
+		     struct flowi4 *fl4, int mp_hash);
 
 /* Exported by fib_trie.c */
 void fib_trie_init(void);
diff --git a/include/net/l3mdev.h b/include/net/l3mdev.h
index 87cee05a0a17..44a19a171104 100644
--- a/include/net/l3mdev.h
+++ b/include/net/l3mdev.h
@@ -17,12 +17,16 @@
  * @l3mdev_fib_table: Get FIB table id to use for lookups
  *
  * @l3mdev_get_rtable: Get cached IPv4 rtable (dst_entry) for device
+ *
+ * @l3mdev_get_saddr: Get source address for a flow
  */
 
 struct l3mdev_ops {
 	u32		(*l3mdev_fib_table)(const struct net_device *dev);
 	struct rtable *	(*l3mdev_get_rtable)(const struct net_device *dev,
 					     const struct flowi4 *fl4);
+	void		(*l3mdev_get_saddr)(struct net_device *dev,
+					    struct flowi4 *fl4);
 };
 
 #ifdef CONFIG_NET_L3_MASTER_DEV
@@ -100,6 +104,25 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex)
 	return rc;
 }
 
+static inline void l3mdev_get_saddr(struct net *net, int ifindex,
+				    struct flowi4 *fl4)
+{
+	struct net_device *dev;
+
+	if (ifindex) {
+
+		rcu_read_lock();
+
+		dev = dev_get_by_index_rcu(net, ifindex);
+		if (dev && netif_is_l3_master(dev) &&
+		    dev->l3mdev_ops->l3mdev_get_saddr) {
+			dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
+		}
+
+		rcu_read_unlock();
+	}
+}
+
 #else
 
 static inline int l3mdev_master_ifindex_rcu(struct net_device *dev)
@@ -144,6 +167,10 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex)
 	return false;
 }
 
+static inline void l3mdev_get_saddr(struct net *net, int ifindex,
+				    struct flowi4 *fl4)
+{
+}
 #endif
 
 #endif /* _NET_L3MDEV_H_ */
diff --git a/include/net/route.h b/include/net/route.h
index d32cb76f5302..ee81307863d5 100644
--- a/include/net/route.h
+++ b/include/net/route.h
@@ -266,9 +266,6 @@ static inline void ip_route_connect_init(struct flowi4 *fl4, __be32 dst, __be32
 	if (inet_sk(sk)->transparent)
 		flow_flags |= FLOWI_FLAG_ANYSRC;
 
-	if (netif_index_is_l3_master(sock_net(sk), oif))
-		flow_flags |= FLOWI_FLAG_VRFSRC | FLOWI_FLAG_SKIP_NH_OIF;
-
 	flowi4_init_output(fl4, oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE,
 			   protocol, flow_flags, dst, src, dport, sport);
 }
@@ -285,6 +282,10 @@ static inline struct rtable *ip_route_connect(struct flowi4 *fl4,
 	ip_route_connect_init(fl4, dst, src, tos, oif, protocol,
 			      sport, dport, sk);
 
+	if (!src && oif) {
+		l3mdev_get_saddr(net, oif, fl4);
+		src = fl4->saddr;
+	}
 	if (!dst || !src) {
 		rt = __ip_route_output_key(net, fl4);
 		if (IS_ERR(rt))