diff options
| author | Denis V. Lunev <den@openvz.org> | 2008-01-22 23:50:25 -0800 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2008-01-28 15:11:13 -0800 |
| commit | b5921910a1de4ba82add59154976c3dc7352c8c2 (patch) | |
| tree | c1de2c3734219548c2883777597247caa5450dc5 /net | |
| parent | eee80592c3c1f7381c04913d9d3eb6e3c3c87628 (diff) | |
| download | linux-3.10-b5921910a1de4ba82add59154976c3dc7352c8c2.tar.gz linux-3.10-b5921910a1de4ba82add59154976c3dc7352c8c2.tar.bz2 linux-3.10-b5921910a1de4ba82add59154976c3dc7352c8c2.zip | |
[NETNS]: Routing cache virtualization.
Basically, this piece looks relatively easy. Namespace is already
available on the dst entry via device and the device is safe to
dereferrence. Compare it with one of a searcher and skip entry if
appropriate.
The only exception is ip_rt_frag_needed. So, add namespace parameter to it.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net')
| -rw-r--r-- | net/ipv4/icmp.c | 2 | ||||
| -rw-r--r-- | net/ipv4/route.c | 21 |
2 files changed, 17 insertions, 6 deletions
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index a142f19fec4..63ffc7d86f9 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -696,7 +696,7 @@ static void icmp_unreach(struct sk_buff *skb) "and DF set.\n", NIPQUAD(iph->daddr)); } else { - info = ip_rt_frag_needed(iph, + info = ip_rt_frag_needed(&init_net, iph, ntohs(icmph->un.frag.mtu)); if (!info) goto out; diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 39a40342142..896c768e41a 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -648,6 +648,11 @@ static inline int compare_keys(struct flowi *fl1, struct flowi *fl2) (fl1->iif ^ fl2->iif)) == 0; } +static inline int compare_netns(struct rtable *rt1, struct rtable *rt2) +{ + return rt1->u.dst.dev->nd_net == rt2->u.dst.dev->nd_net; +} + /* * Perform a full scan of hash table and free all entries. * Can be called by a softirq or a process. @@ -961,7 +966,7 @@ restart: spin_lock_bh(rt_hash_lock_addr(hash)); while ((rth = *rthp) != NULL) { - if (compare_keys(&rth->fl, &rt->fl)) { + if (compare_keys(&rth->fl, &rt->fl) && compare_netns(rth, rt)) { /* Put it first */ *rthp = rth->u.dst.rt_next; /* @@ -1415,7 +1420,8 @@ static __inline__ unsigned short guess_mtu(unsigned short old_mtu) return 68; } -unsigned short ip_rt_frag_needed(struct iphdr *iph, unsigned short new_mtu) +unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, + unsigned short new_mtu) { int i; unsigned short old_mtu = ntohs(iph->tot_len); @@ -1438,7 +1444,8 @@ unsigned short ip_rt_frag_needed(struct iphdr *iph, unsigned short new_mtu) rth->rt_dst == daddr && rth->rt_src == iph->saddr && rth->fl.iif == 0 && - !(dst_metric_locked(&rth->u.dst, RTAX_MTU))) { + !(dst_metric_locked(&rth->u.dst, RTAX_MTU)) && + rth->u.dst.dev->nd_net == net) { unsigned short mtu = new_mtu; if (new_mtu < 68 || new_mtu >= old_mtu) { @@ -2049,7 +2056,9 @@ int ip_route_input(struct sk_buff *skb, __be32 daddr, __be32 saddr, struct rtable * rth; unsigned hash; int iif = dev->ifindex; + struct net *net; + net = skb->dev->nd_net; tos &= IPTOS_RT_MASK; hash = rt_hash(daddr, saddr, iif); @@ -2061,7 +2070,8 @@ int ip_route_input(struct sk_buff *skb, __be32 daddr, __be32 saddr, rth->fl.iif == iif && rth->fl.oif == 0 && rth->fl.mark == skb->mark && - rth->fl.fl4_tos == tos) { + rth->fl.fl4_tos == tos && + rth->u.dst.dev->nd_net == net) { dst_use(&rth->u.dst, jiffies); RT_CACHE_STAT_INC(in_hit); rcu_read_unlock(); @@ -2460,7 +2470,8 @@ int __ip_route_output_key(struct net *net, struct rtable **rp, rth->fl.oif == flp->oif && rth->fl.mark == flp->mark && !((rth->fl.fl4_tos ^ flp->fl4_tos) & - (IPTOS_RT_MASK | RTO_ONLINK))) { + (IPTOS_RT_MASK | RTO_ONLINK)) && + rth->u.dst.dev->nd_net == net) { dst_use(&rth->u.dst, jiffies); RT_CACHE_STAT_INC(out_hit); rcu_read_unlock_bh(); |