[PATCH net-next v2 0/2] net: l3mdev: Allow send on enslaved interface

[PATCH net-next v2 0/2] net: l3mdev: Allow send on enslaved interface

[David Ahern]

First patch preps for the second. The second is required for several use
cases such as ping on an interface and BFD that need to send packets on
a specific interface, including ones enslaved to a VRF device.

v2
- fixed brackets on both patches per comment from DaveM

David Ahern (2):
  net: l3mdev: Move get_saddr and rt6_dst
  net: l3mdev: Allow send on enslaved interface

 drivers/net/vrf.c    |  2 ++
 include/net/l3mdev.h | 56 +++-------------------------------------------
 net/ipv4/route.c     |  4 ++++
 net/ipv6/route.c     |  2 +-
 net/l3mdev/l3mdev.c  | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 5 files changed, 73 insertions(+), 54 deletions(-)

-- 
2.1.4

[PATCH net-next 1/2] net: l3mdev: Move get_saddr and rt6_dst

[David Ahern]

Move l3mdev_rt6_dst_by_oif and l3mdev_get_saddr to l3mdev.c. Collapse
l3mdev_get_rt6_dst into l3mdev_rt6_dst_by_oif since it is the only
user and keep the l3mdev_get_rt6_dst name for consistency with other
hooks.

A follow-on patch adds more code to these functions making them long
for inlined functions.

Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 include/net/l3mdev.h | 56 +++-------------------------------------------------
 net/ipv6/route.c     |  2 +-
 net/l3mdev/l3mdev.c  | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 58 insertions(+), 54 deletions(-)

diff --git a/include/net/l3mdev.h b/include/net/l3mdev.h
index c43a9c73de5e..78872bd1dc2c 100644
--- a/include/net/l3mdev.h
+++ b/include/net/l3mdev.h
@@ -130,52 +130,9 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex)
 	return rc;
 }
 
-static inline int l3mdev_get_saddr(struct net *net, int ifindex,
-				   struct flowi4 *fl4)
-{
-	struct net_device *dev;
-	int rc = 0;
-
-	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) {
-			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
-		}
-
-		rcu_read_unlock();
-	}
-
-	return rc;
-}
+int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4);
 
-static inline struct dst_entry *l3mdev_get_rt6_dst(const struct net_device *dev,
-						   const struct flowi6 *fl6)
-{
-	if (netif_is_l3_master(dev) && dev->l3mdev_ops->l3mdev_get_rt6_dst)
-		return dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6);
-
-	return NULL;
-}
-
-static inline
-struct dst_entry *l3mdev_rt6_dst_by_oif(struct net *net,
-					const struct flowi6 *fl6)
-{
-	struct dst_entry *dst = NULL;
-	struct net_device *dev;
-
-	dev = dev_get_by_index(net, fl6->flowi6_oif);
-	if (dev) {
-		dst = l3mdev_get_rt6_dst(dev, fl6);
-		dev_put(dev);
-	}
-
-	return dst;
-}
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net, const struct flowi6 *fl6);
 
 #else
 
@@ -233,14 +190,7 @@ static inline int l3mdev_get_saddr(struct net *net, int ifindex,
 }
 
 static inline
-struct dst_entry *l3mdev_get_rt6_dst(const struct net_device *dev,
-				     const struct flowi6 *fl6)
-{
-	return NULL;
-}
-static inline
-struct dst_entry *l3mdev_rt6_dst_by_oif(struct net *net,
-					const struct flowi6 *fl6)
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net, const struct flowi6 *fl6)
 {
 	return NULL;
 }
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index af46e19205f5..c42fa1deb152 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1190,7 +1190,7 @@ struct dst_entry *ip6_route_output_flags(struct net *net, const struct sock *sk,
 	struct dst_entry *dst;
 	bool any_src;
 
-	dst = l3mdev_rt6_dst_by_oif(net, fl6);
+	dst = l3mdev_get_rt6_dst(net, fl6);
 	if (dst)
 		return dst;
 
diff --git a/net/l3mdev/l3mdev.c b/net/l3mdev/l3mdev.c
index e925037fa0df..898d01e0f87b 100644
--- a/net/l3mdev/l3mdev.c
+++ b/net/l3mdev/l3mdev.c
@@ -97,3 +97,57 @@ u32 l3mdev_fib_table_by_index(struct net *net, int ifindex)
 	return tb_id;
 }
 EXPORT_SYMBOL_GPL(l3mdev_fib_table_by_index);
+
+/**
+ *	l3mdev_get_rt6_dst - IPv6 route lookup based on flow. Returns
+ *			     cached route for L3 master device if relevant
+ *			     to flow
+ *	@net: network namespace for device index lookup
+ *	@fl6: IPv6 flow struct for lookup
+ */
+
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net,
+				     const struct flowi6 *fl6)
+{
+	struct dst_entry *dst = NULL;
+	struct net_device *dev;
+
+	dev = dev_get_by_index(net, fl6->flowi6_oif);
+	if (dev) {
+		if (netif_is_l3_master(dev) &&
+		    dev->l3mdev_ops->l3mdev_get_rt6_dst)
+			dst = dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6);
+		dev_put(dev);
+	}
+
+	return dst;
+}
+EXPORT_SYMBOL_GPL(l3mdev_get_rt6_dst);
+
+/**
+ *	l3mdev_get_saddr - get source address for a flow based on an interface
+ *			   enslaved to an L3 master device
+ *	@net: network namespace for device index lookup
+ *	@ifindex: Interface index
+ *	@fl4: IPv4 flow struct
+ */
+
+int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4)
+{
+	struct net_device *dev;
+	int rc = 0;
+
+	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)
+			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
+
+		rcu_read_unlock();
+	}
+
+	return rc;
+}
+EXPORT_SYMBOL_GPL(l3mdev_get_saddr);
-- 
2.1.4

[PATCH net-next 2/2] net: l3mdev: Allow send on enslaved interface

[David Ahern]

Allow udp and raw sockets to send by oif that is an enslaved interface
versus the l3mdev/VRF device. For example, this allows BFD to use ifindex
from IP_PKTINFO on a receive to send a response without the need to
convert to the VRF index. It also allows ping and ping6 to work when
specifying an enslaved interface (e.g., ping -I swp1 <ip>) which is
a natural use case.

Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 drivers/net/vrf.c   |  2 ++
 net/ipv4/route.c    |  4 ++++
 net/l3mdev/l3mdev.c | 17 +++++++++++++----
 3 files changed, 19 insertions(+), 4 deletions(-)

diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c
index 4b2461ae5d3b..c8db55aa8280 100644
--- a/drivers/net/vrf.c
+++ b/drivers/net/vrf.c
@@ -648,6 +648,8 @@ static int vrf_get_saddr(struct net_device *dev, struct flowi4 *fl4)
 
 	fl4->flowi4_flags |= FLOWI_FLAG_SKIP_NH_OIF;
 	fl4->flowi4_iif = LOOPBACK_IFINDEX;
+	/* make sure oif is set to VRF device for lookup */
+	fl4->flowi4_oif = dev->ifindex;
 	fl4->flowi4_tos = tos & IPTOS_RT_MASK;
 	fl4->flowi4_scope = ((tos & RTO_ONLINK) ?
 			     RT_SCOPE_LINK : RT_SCOPE_UNIVERSE);
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 8c8c655bb2c4..a1f2830d8110 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -2146,6 +2146,7 @@ struct rtable *__ip_route_output_key_hash(struct net *net, struct flowi4 *fl4,
 	unsigned int flags = 0;
 	struct fib_result res;
 	struct rtable *rth;
+	int master_idx;
 	int orig_oif;
 	int err = -ENETUNREACH;
 
@@ -2155,6 +2156,9 @@ struct rtable *__ip_route_output_key_hash(struct net *net, struct flowi4 *fl4,
 
 	orig_oif = fl4->flowi4_oif;
 
+	master_idx = l3mdev_master_ifindex_by_index(net, fl4->flowi4_oif);
+	if (master_idx)
+		fl4->flowi4_oif = master_idx;
 	fl4->flowi4_iif = LOOPBACK_IFINDEX;
 	fl4->flowi4_tos = tos & IPTOS_RT_MASK;
 	fl4->flowi4_scope = ((tos & RTO_ONLINK) ?
diff --git a/net/l3mdev/l3mdev.c b/net/l3mdev/l3mdev.c
index 898d01e0f87b..6651a78e100c 100644
--- a/net/l3mdev/l3mdev.c
+++ b/net/l3mdev/l3mdev.c
@@ -112,12 +112,18 @@ struct dst_entry *l3mdev_get_rt6_dst(struct net *net,
 	struct dst_entry *dst = NULL;
 	struct net_device *dev;
 
-	dev = dev_get_by_index(net, fl6->flowi6_oif);
-	if (dev) {
-		if (netif_is_l3_master(dev) &&
+	if (fl6->flowi6_oif) {
+		rcu_read_lock();
+
+		dev = dev_get_by_index_rcu(net, fl6->flowi6_oif);
+		if (dev && netif_is_l3_slave(dev))
+			dev = netdev_master_upper_dev_get_rcu(dev);
+
+		if (dev && netif_is_l3_master(dev) &&
 		    dev->l3mdev_ops->l3mdev_get_rt6_dst)
 			dst = dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6);
-		dev_put(dev);
+
+		rcu_read_unlock();
 	}
 
 	return dst;
@@ -141,6 +147,9 @@ int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4)
 		rcu_read_lock();
 
 		dev = dev_get_by_index_rcu(net, ifindex);
+		if (dev && netif_is_l3_slave(dev))
+			dev = netdev_master_upper_dev_get_rcu(dev);
+
 		if (dev && netif_is_l3_master(dev) &&
 		    dev->l3mdev_ops->l3mdev_get_saddr)
 			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
-- 
2.1.4

Re: [PATCH net-next v2 0/2] net: l3mdev: Allow send on enslaved interface

[David Miller]

From: David Ahern <dsa@cumulusnetworks.com>
Date: Sat,  7 May 2016 16:48:58 -0700

> First patch preps for the second. The second is required for several use
> cases such as ping on an interface and BFD that need to send packets on
> a specific interface, including ones enslaved to a VRF device.
> 
> v2
> - fixed brackets on both patches per comment from DaveM

Series applied.

Re: [PATCH net-next 1/2] net: l3mdev: Move get_saddr and rt6_dst

[David Miller]

From: David Ahern <dsa@cumulusnetworks.com>
Date: Wed,  4 May 2016 21:54:09 -0700

> +		if (dev && netif_is_l3_master(dev) &&
> +		    dev->l3mdev_ops->l3mdev_get_saddr) {
> +			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
> +		}

Please do not use braces for single statement basic blocks.

The same issue exists in patch #2 so please fix it there as well.

Thanks.

[PATCH net-next 1/2] net: l3mdev: Move get_saddr and rt6_dst

[David Ahern]

Move l3mdev_rt6_dst_by_oif and l3mdev_get_saddr to l3mdev.c. Collapse
l3mdev_get_rt6_dst into l3mdev_rt6_dst_by_oif since it is the only
user and keep the l3mdev_get_rt6_dst name for consistency with other
hooks.

A follow-on patch adds more code to these functions making them long
for inlined functions.

Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 include/net/l3mdev.h | 56 +++-------------------------------------------------
 net/ipv6/route.c     |  2 +-
 net/l3mdev/l3mdev.c  | 55 +++++++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 59 insertions(+), 54 deletions(-)

diff --git a/include/net/l3mdev.h b/include/net/l3mdev.h
index c43a9c73de5e..78872bd1dc2c 100644
--- a/include/net/l3mdev.h
+++ b/include/net/l3mdev.h
@@ -130,52 +130,9 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex)
 	return rc;
 }
 
-static inline int l3mdev_get_saddr(struct net *net, int ifindex,
-				   struct flowi4 *fl4)
-{
-	struct net_device *dev;
-	int rc = 0;
-
-	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) {
-			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
-		}
-
-		rcu_read_unlock();
-	}
-
-	return rc;
-}
+int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4);
 
-static inline struct dst_entry *l3mdev_get_rt6_dst(const struct net_device *dev,
-						   const struct flowi6 *fl6)
-{
-	if (netif_is_l3_master(dev) && dev->l3mdev_ops->l3mdev_get_rt6_dst)
-		return dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6);
-
-	return NULL;
-}
-
-static inline
-struct dst_entry *l3mdev_rt6_dst_by_oif(struct net *net,
-					const struct flowi6 *fl6)
-{
-	struct dst_entry *dst = NULL;
-	struct net_device *dev;
-
-	dev = dev_get_by_index(net, fl6->flowi6_oif);
-	if (dev) {
-		dst = l3mdev_get_rt6_dst(dev, fl6);
-		dev_put(dev);
-	}
-
-	return dst;
-}
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net, const struct flowi6 *fl6);
 
 #else
 
@@ -233,14 +190,7 @@ static inline int l3mdev_get_saddr(struct net *net, int ifindex,
 }
 
 static inline
-struct dst_entry *l3mdev_get_rt6_dst(const struct net_device *dev,
-				     const struct flowi6 *fl6)
-{
-	return NULL;
-}
-static inline
-struct dst_entry *l3mdev_rt6_dst_by_oif(struct net *net,
-					const struct flowi6 *fl6)
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net, const struct flowi6 *fl6)
 {
 	return NULL;
 }
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index af46e19205f5..c42fa1deb152 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1190,7 +1190,7 @@ struct dst_entry *ip6_route_output_flags(struct net *net, const struct sock *sk,
 	struct dst_entry *dst;
 	bool any_src;
 
-	dst = l3mdev_rt6_dst_by_oif(net, fl6);
+	dst = l3mdev_get_rt6_dst(net, fl6);
 	if (dst)
 		return dst;
 
diff --git a/net/l3mdev/l3mdev.c b/net/l3mdev/l3mdev.c
index e925037fa0df..0fe4211e646f 100644
--- a/net/l3mdev/l3mdev.c
+++ b/net/l3mdev/l3mdev.c
@@ -97,3 +97,58 @@ u32 l3mdev_fib_table_by_index(struct net *net, int ifindex)
 	return tb_id;
 }
 EXPORT_SYMBOL_GPL(l3mdev_fib_table_by_index);
+
+/**
+ *	l3mdev_get_rt6_dst - IPv6 route lookup based on flow. Returns
+ *			     cached route for L3 master device if relevant
+ *			     to flow
+ *	@net: network namespace for device index lookup
+ *	@fl6: IPv6 flow struct for lookup
+ */
+
+struct dst_entry *l3mdev_get_rt6_dst(struct net *net,
+				     const struct flowi6 *fl6)
+{
+	struct dst_entry *dst = NULL;
+	struct net_device *dev;
+
+	dev = dev_get_by_index(net, fl6->flowi6_oif);
+	if (dev) {
+		if (netif_is_l3_master(dev) &&
+		    dev->l3mdev_ops->l3mdev_get_rt6_dst)
+			dst = dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6);
+		dev_put(dev);
+	}
+
+	return dst;
+}
+EXPORT_SYMBOL_GPL(l3mdev_get_rt6_dst);
+
+/**
+ *	l3mdev_get_saddr - get source address for a flow based on an interface
+ *			   enslaved to an L3 master device
+ *	@net: network namespace for device index lookup
+ *	@ifindex: Interface index
+ *	@fl4: IPv4 flow struct
+ */
+
+int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4)
+{
+	struct net_device *dev;
+	int rc = 0;
+
+	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) {
+			rc = dev->l3mdev_ops->l3mdev_get_saddr(dev, fl4);
+		}
+
+		rcu_read_unlock();
+	}
+
+	return rc;
+}
+EXPORT_SYMBOL_GPL(l3mdev_get_saddr);
-- 
2.1.4