提交 4c1feac5 编写于 作者: D David Ahern 提交者: David S. Miller

net: vrf: Flip IPv6 output path from FIB lookup hook to out hook

Flip the IPv6 output path to use the l3mdev tx out hook. The VRF dst
is not returned on the first FIB lookup. Instead, the dst on the
skb is switched at the beginning of the IPv6 output processing to
send the packet to the VRF driver on xmit.

Link scope addresses (linklocal and multicast) need special handling:
specifically the oif the flow struct can not be changed because we
want the lookup tied to the enslaved interface. ie., the source address
and the returned route MUST point to the interface scope passed in.
Convert the existing vrf_get_rt6_dst to handle only link scope addresses.
Signed-off-by: NDavid Ahern <dsa@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
上级 ebfc102c
...@@ -137,6 +137,20 @@ static int vrf_local_xmit(struct sk_buff *skb, struct net_device *dev, ...@@ -137,6 +137,20 @@ static int vrf_local_xmit(struct sk_buff *skb, struct net_device *dev,
} }
#if IS_ENABLED(CONFIG_IPV6) #if IS_ENABLED(CONFIG_IPV6)
static int vrf_ip6_local_out(struct net *net, struct sock *sk,
struct sk_buff *skb)
{
int err;
err = nf_hook(NFPROTO_IPV6, NF_INET_LOCAL_OUT, net,
sk, skb, NULL, skb_dst(skb)->dev, dst_output);
if (likely(err == 1))
err = dst_output(net, sk, skb);
return err;
}
static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb, static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb,
struct net_device *dev) struct net_device *dev)
{ {
...@@ -207,7 +221,7 @@ static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb, ...@@ -207,7 +221,7 @@ static netdev_tx_t vrf_process_v6_outbound(struct sk_buff *skb,
/* strip the ethernet header added for pass through VRF device */ /* strip the ethernet header added for pass through VRF device */
__skb_pull(skb, skb_network_offset(skb)); __skb_pull(skb, skb_network_offset(skb));
ret = ip6_local_out(net, skb->sk, skb); ret = vrf_ip6_local_out(net, skb->sk, skb);
if (unlikely(net_xmit_eval(ret))) if (unlikely(net_xmit_eval(ret)))
dev->stats.tx_errors++; dev->stats.tx_errors++;
else else
...@@ -391,6 +405,43 @@ static int vrf_output6(struct net *net, struct sock *sk, struct sk_buff *skb) ...@@ -391,6 +405,43 @@ static int vrf_output6(struct net *net, struct sock *sk, struct sk_buff *skb)
!(IP6CB(skb)->flags & IP6SKB_REROUTED)); !(IP6CB(skb)->flags & IP6SKB_REROUTED));
} }
/* set dst on skb to send packet to us via dev_xmit path. Allows
* packet to go through device based features such as qdisc, netfilter
* hooks and packet sockets with skb->dev set to vrf device.
*/
static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev,
struct sock *sk,
struct sk_buff *skb)
{
struct net_vrf *vrf = netdev_priv(vrf_dev);
struct dst_entry *dst = NULL;
struct rt6_info *rt6;
/* don't divert link scope packets */
if (rt6_need_strict(&ipv6_hdr(skb)->daddr))
return skb;
rcu_read_lock();
rt6 = rcu_dereference(vrf->rt6);
if (likely(rt6)) {
dst = &rt6->dst;
dst_hold(dst);
}
rcu_read_unlock();
if (unlikely(!dst)) {
vrf_tx_error(vrf_dev, skb);
return NULL;
}
skb_dst_drop(skb);
skb_dst_set(skb, dst);
return skb;
}
/* holding rtnl */ /* holding rtnl */
static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf) static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf)
{ {
...@@ -477,6 +528,13 @@ static int vrf_rt6_create(struct net_device *dev) ...@@ -477,6 +528,13 @@ static int vrf_rt6_create(struct net_device *dev)
return rc; return rc;
} }
#else #else
static struct sk_buff *vrf_ip6_out(struct net_device *vrf_dev,
struct sock *sk,
struct sk_buff *skb)
{
return skb;
}
static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf) static void vrf_rt6_release(struct net_device *dev, struct net_vrf *vrf)
{ {
} }
...@@ -587,6 +645,8 @@ static struct sk_buff *vrf_l3_out(struct net_device *vrf_dev, ...@@ -587,6 +645,8 @@ static struct sk_buff *vrf_l3_out(struct net_device *vrf_dev,
switch (proto) { switch (proto) {
case AF_INET: case AF_INET:
return vrf_ip_out(vrf_dev, sk, skb); return vrf_ip_out(vrf_dev, sk, skb);
case AF_INET6:
return vrf_ip6_out(vrf_dev, sk, skb);
} }
return skb; return skb;
...@@ -1031,53 +1091,33 @@ static struct sk_buff *vrf_l3_rcv(struct net_device *vrf_dev, ...@@ -1031,53 +1091,33 @@ static struct sk_buff *vrf_l3_rcv(struct net_device *vrf_dev,
} }
#if IS_ENABLED(CONFIG_IPV6) #if IS_ENABLED(CONFIG_IPV6)
static struct dst_entry *vrf_get_rt6_dst(const struct net_device *dev, /* send to link-local or multicast address via interface enslaved to
struct flowi6 *fl6) * VRF device. Force lookup to VRF table without changing flow struct
*/
static struct dst_entry *vrf_link_scope_lookup(const struct net_device *dev,
struct flowi6 *fl6)
{ {
bool need_strict = rt6_need_strict(&fl6->daddr);
struct net_vrf *vrf = netdev_priv(dev);
struct net *net = dev_net(dev); struct net *net = dev_net(dev);
int flags = RT6_LOOKUP_F_IFACE;
struct dst_entry *dst = NULL; struct dst_entry *dst = NULL;
struct rt6_info *rt; struct rt6_info *rt;
/* send to link-local or multicast address */ /* VRF device does not have a link-local address and
if (need_strict) { * sending packets to link-local or mcast addresses over
int flags = RT6_LOOKUP_F_IFACE; * a VRF device does not make sense
*/
/* VRF device does not have a link-local address and if (fl6->flowi6_oif == dev->ifindex) {
* sending packets to link-local or mcast addresses over dst = &net->ipv6.ip6_null_entry->dst;
* a VRF device does not make sense dst_hold(dst);
*/ return dst;
if (fl6->flowi6_oif == dev->ifindex) {
struct dst_entry *dst = &net->ipv6.ip6_null_entry->dst;
dst_hold(dst);
return dst;
}
if (!ipv6_addr_any(&fl6->saddr))
flags |= RT6_LOOKUP_F_HAS_SADDR;
rt = vrf_ip6_route_lookup(net, dev, fl6, fl6->flowi6_oif, flags);
if (rt)
dst = &rt->dst;
} else if (!(fl6->flowi6_flags & FLOWI_FLAG_L3MDEV_SRC)) {
rcu_read_lock();
rt = rcu_dereference(vrf->rt6);
if (likely(rt)) {
dst = &rt->dst;
dst_hold(dst);
}
rcu_read_unlock();
} }
/* make sure oif is set to VRF device for lookup */ if (!ipv6_addr_any(&fl6->saddr))
if (!need_strict) flags |= RT6_LOOKUP_F_HAS_SADDR;
fl6->flowi6_oif = dev->ifindex;
rt = vrf_ip6_route_lookup(net, dev, fl6, fl6->flowi6_oif, flags);
if (rt)
dst = &rt->dst;
return dst; return dst;
} }
...@@ -1130,7 +1170,7 @@ static const struct l3mdev_ops vrf_l3mdev_ops = { ...@@ -1130,7 +1170,7 @@ static const struct l3mdev_ops vrf_l3mdev_ops = {
.l3mdev_l3_rcv = vrf_l3_rcv, .l3mdev_l3_rcv = vrf_l3_rcv,
.l3mdev_l3_out = vrf_l3_out, .l3mdev_l3_out = vrf_l3_out,
#if IS_ENABLED(CONFIG_IPV6) #if IS_ENABLED(CONFIG_IPV6)
.l3mdev_get_rt6_dst = vrf_get_rt6_dst, .l3mdev_link_scope_lookup = vrf_link_scope_lookup,
.l3mdev_get_saddr6 = vrf_get_saddr6, .l3mdev_get_saddr6 = vrf_get_saddr6,
#endif #endif
}; };
......
...@@ -27,7 +27,7 @@ ...@@ -27,7 +27,7 @@
* *
* @l3mdev_get_saddr: Get source address for a flow * @l3mdev_get_saddr: Get source address for a flow
* *
* @l3mdev_get_rt6_dst: Get cached IPv6 rt6_info (dst_entry) for device * @l3mdev_link_scope_lookup: IPv6 lookup for linklocal and mcast destinations
*/ */
struct l3mdev_ops { struct l3mdev_ops {
...@@ -45,7 +45,7 @@ struct l3mdev_ops { ...@@ -45,7 +45,7 @@ struct l3mdev_ops {
struct flowi4 *fl4); struct flowi4 *fl4);
/* IPv6 ops */ /* IPv6 ops */
struct dst_entry * (*l3mdev_get_rt6_dst)(const struct net_device *dev, struct dst_entry * (*l3mdev_link_scope_lookup)(const struct net_device *dev,
struct flowi6 *fl6); struct flowi6 *fl6);
int (*l3mdev_get_saddr6)(struct net_device *dev, int (*l3mdev_get_saddr6)(struct net_device *dev,
const struct sock *sk, const struct sock *sk,
...@@ -177,7 +177,7 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex) ...@@ -177,7 +177,7 @@ static inline bool netif_index_is_l3_master(struct net *net, int ifindex)
int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4); int l3mdev_get_saddr(struct net *net, int ifindex, struct flowi4 *fl4);
struct dst_entry *l3mdev_get_rt6_dst(struct net *net, struct flowi6 *fl6); struct dst_entry *l3mdev_link_scope_lookup(struct net *net, struct flowi6 *fl6);
int l3mdev_get_saddr6(struct net *net, const struct sock *sk, int l3mdev_get_saddr6(struct net *net, const struct sock *sk,
struct flowi6 *fl6); struct flowi6 *fl6);
...@@ -299,7 +299,7 @@ static inline int l3mdev_get_saddr(struct net *net, int ifindex, ...@@ -299,7 +299,7 @@ static inline int l3mdev_get_saddr(struct net *net, int ifindex,
} }
static inline static inline
struct dst_entry *l3mdev_get_rt6_dst(struct net *net, struct flowi6 *fl6) struct dst_entry *l3mdev_link_scope_lookup(struct net *net, struct flowi6 *fl6)
{ {
return NULL; return NULL;
} }
......
...@@ -1188,12 +1188,15 @@ static struct rt6_info *ip6_pol_route_output(struct net *net, struct fib6_table ...@@ -1188,12 +1188,15 @@ static struct rt6_info *ip6_pol_route_output(struct net *net, struct fib6_table
struct dst_entry *ip6_route_output_flags(struct net *net, const struct sock *sk, struct dst_entry *ip6_route_output_flags(struct net *net, const struct sock *sk,
struct flowi6 *fl6, int flags) struct flowi6 *fl6, int flags)
{ {
struct dst_entry *dst;
bool any_src; bool any_src;
dst = l3mdev_get_rt6_dst(net, fl6); if (rt6_need_strict(&fl6->daddr)) {
if (dst) struct dst_entry *dst;
return dst;
dst = l3mdev_link_scope_lookup(net, fl6);
if (dst)
return dst;
}
fl6->flowi6_iif = LOOPBACK_IFINDEX; fl6->flowi6_iif = LOOPBACK_IFINDEX;
......
...@@ -100,15 +100,14 @@ u32 l3mdev_fib_table_by_index(struct net *net, int ifindex) ...@@ -100,15 +100,14 @@ u32 l3mdev_fib_table_by_index(struct net *net, int ifindex)
EXPORT_SYMBOL_GPL(l3mdev_fib_table_by_index); EXPORT_SYMBOL_GPL(l3mdev_fib_table_by_index);
/** /**
* l3mdev_get_rt6_dst - IPv6 route lookup based on flow. Returns * l3mdev_link_scope_lookup - IPv6 route lookup based on flow for link
* cached route for L3 master device if relevant * local and multicast addresses
* to flow
* @net: network namespace for device index lookup * @net: network namespace for device index lookup
* @fl6: IPv6 flow struct for lookup * @fl6: IPv6 flow struct for lookup
*/ */
struct dst_entry *l3mdev_get_rt6_dst(struct net *net, struct dst_entry *l3mdev_link_scope_lookup(struct net *net,
struct flowi6 *fl6) struct flowi6 *fl6)
{ {
struct dst_entry *dst = NULL; struct dst_entry *dst = NULL;
struct net_device *dev; struct net_device *dev;
...@@ -121,15 +120,15 @@ struct dst_entry *l3mdev_get_rt6_dst(struct net *net, ...@@ -121,15 +120,15 @@ struct dst_entry *l3mdev_get_rt6_dst(struct net *net,
dev = netdev_master_upper_dev_get_rcu(dev); dev = netdev_master_upper_dev_get_rcu(dev);
if (dev && netif_is_l3_master(dev) && if (dev && netif_is_l3_master(dev) &&
dev->l3mdev_ops->l3mdev_get_rt6_dst) dev->l3mdev_ops->l3mdev_link_scope_lookup)
dst = dev->l3mdev_ops->l3mdev_get_rt6_dst(dev, fl6); dst = dev->l3mdev_ops->l3mdev_link_scope_lookup(dev, fl6);
rcu_read_unlock(); rcu_read_unlock();
} }
return dst; return dst;
} }
EXPORT_SYMBOL_GPL(l3mdev_get_rt6_dst); EXPORT_SYMBOL_GPL(l3mdev_link_scope_lookup);
/** /**
* l3mdev_get_saddr - get source address for a flow based on an interface * l3mdev_get_saddr - get source address for a flow based on an interface
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册