ip_tunnel: Add tnl_update_pmtu in ip_md_tunnel_xmit

Add tnl_update_pmtu in ip_md_tunnel_xmit to dynamic modify
the pmtu which packet send through collect_metadata mode
ip tunnel

Signed-off-by: wenxu <wenxu@ucloud.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
wenxu 2019-01-22 18:39:50 +08:00 committed by David S. Miller
parent f46fe4f8d7
commit c8b34e680a
3 changed files with 29 additions and 12 deletions

View File

@ -267,7 +267,7 @@ void ip_tunnel_delete_nets(struct list_head *list_net, unsigned int id,
void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
const struct iphdr *tnl_params, const u8 protocol); const struct iphdr *tnl_params, const u8 protocol);
void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
const u8 proto); const u8 proto, int tunnel_hlen);
int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd); int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd);
int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict); int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict);
int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu); int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu);

View File

@ -501,15 +501,19 @@ EXPORT_SYMBOL_GPL(ip_tunnel_encap_setup);
static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb, static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
struct rtable *rt, __be16 df, struct rtable *rt, __be16 df,
const struct iphdr *inner_iph) const struct iphdr *inner_iph,
int tunnel_hlen, __be32 dst, bool md)
{ {
struct ip_tunnel *tunnel = netdev_priv(dev); struct ip_tunnel *tunnel = netdev_priv(dev);
int pkt_size = skb->len - tunnel->hlen - dev->hard_header_len; int pkt_size;
int mtu; int mtu;
tunnel_hlen = md ? tunnel_hlen : tunnel->hlen;
pkt_size = skb->len - tunnel_hlen - dev->hard_header_len;
if (df) if (df)
mtu = dst_mtu(&rt->dst) - dev->hard_header_len mtu = dst_mtu(&rt->dst) - dev->hard_header_len
- sizeof(struct iphdr) - tunnel->hlen; - sizeof(struct iphdr) - tunnel_hlen;
else else
mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu; mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
@ -527,11 +531,13 @@ static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
#if IS_ENABLED(CONFIG_IPV6) #if IS_ENABLED(CONFIG_IPV6)
else if (skb->protocol == htons(ETH_P_IPV6)) { else if (skb->protocol == htons(ETH_P_IPV6)) {
struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb); struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
__be32 daddr;
daddr = md ? dst : tunnel->parms.iph.daddr;
if (rt6 && mtu < dst_mtu(skb_dst(skb)) && if (rt6 && mtu < dst_mtu(skb_dst(skb)) &&
mtu >= IPV6_MIN_MTU) { mtu >= IPV6_MIN_MTU) {
if ((tunnel->parms.iph.daddr && if ((daddr && !ipv4_is_multicast(daddr)) ||
!ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
rt6->rt6i_dst.plen == 128) { rt6->rt6i_dst.plen == 128) {
rt6->rt6i_flags |= RTF_MODIFIED; rt6->rt6i_flags |= RTF_MODIFIED;
dst_metric_set(skb_dst(skb), RTAX_MTU, mtu); dst_metric_set(skb_dst(skb), RTAX_MTU, mtu);
@ -548,7 +554,8 @@ static int tnl_update_pmtu(struct net_device *dev, struct sk_buff *skb,
return 0; return 0;
} }
void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, u8 proto) void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
u8 proto, int tunnel_hlen)
{ {
struct ip_tunnel *tunnel = netdev_priv(dev); struct ip_tunnel *tunnel = netdev_priv(dev);
u32 headroom = sizeof(struct iphdr); u32 headroom = sizeof(struct iphdr);
@ -598,6 +605,15 @@ void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, u8 proto)
dev->stats.collisions++; dev->stats.collisions++;
goto tx_error; goto tx_error;
} }
if (key->tun_flags & TUNNEL_DONT_FRAGMENT)
df = htons(IP_DF);
if (tnl_update_pmtu(dev, skb, rt, df, inner_iph, tunnel_hlen,
key->u.ipv4.dst, true)) {
ip_rt_put(rt);
goto tx_error;
}
tos = ip_tunnel_ecn_encap(tos, inner_iph, skb); tos = ip_tunnel_ecn_encap(tos, inner_iph, skb);
ttl = key->ttl; ttl = key->ttl;
if (ttl == 0) { if (ttl == 0) {
@ -608,10 +624,10 @@ void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, u8 proto)
else else
ttl = ip4_dst_hoplimit(&rt->dst); ttl = ip4_dst_hoplimit(&rt->dst);
} }
if (key->tun_flags & TUNNEL_DONT_FRAGMENT)
df = htons(IP_DF); if (!df && skb->protocol == htons(ETH_P_IP))
else if (skb->protocol == htons(ETH_P_IP))
df = inner_iph->frag_off & htons(IP_DF); df = inner_iph->frag_off & htons(IP_DF);
headroom += LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len; headroom += LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len;
if (headroom > dev->needed_headroom) if (headroom > dev->needed_headroom)
dev->needed_headroom = headroom; dev->needed_headroom = headroom;
@ -741,7 +757,8 @@ void ip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev,
goto tx_error; goto tx_error;
} }
if (tnl_update_pmtu(dev, skb, rt, tnl_params->frag_off, inner_iph)) { if (tnl_update_pmtu(dev, skb, rt, tnl_params->frag_off, inner_iph,
0, 0, false)) {
ip_rt_put(rt); ip_rt_put(rt);
goto tx_error; goto tx_error;
} }

View File

@ -302,7 +302,7 @@ static netdev_tx_t ipip_tunnel_xmit(struct sk_buff *skb,
skb_set_inner_ipproto(skb, ipproto); skb_set_inner_ipproto(skb, ipproto);
if (tunnel->collect_md) if (tunnel->collect_md)
ip_md_tunnel_xmit(skb, dev, ipproto); ip_md_tunnel_xmit(skb, dev, ipproto, 0);
else else
ip_tunnel_xmit(skb, dev, tiph, ipproto); ip_tunnel_xmit(skb, dev, tiph, ipproto);
return NETDEV_TX_OK; return NETDEV_TX_OK;