]> git.kernelconcepts.de Git - karo-tx-linux.git/blobdiff - net/ipv6/route.c
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
[karo-tx-linux.git] / net / ipv6 / route.c
index cc88a73b9ce076c5951d296a25471930d6e0079b..c8bc9b4ac32835e7328025a8912f3627b365e38a 100644 (file)
@@ -61,6 +61,7 @@
 #include <net/nexthop.h>
 #include <net/lwtunnel.h>
 #include <net/ip_tunnels.h>
+#include <net/l3mdev.h>
 
 #include <asm/uaccess.h>
 
@@ -86,9 +87,9 @@ static void           ip6_dst_ifdown(struct dst_entry *,
 static int              ip6_dst_gc(struct dst_ops *ops);
 
 static int             ip6_pkt_discard(struct sk_buff *skb);
-static int             ip6_pkt_discard_out(struct sock *sk, struct sk_buff *skb);
+static int             ip6_pkt_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb);
 static int             ip6_pkt_prohibit(struct sk_buff *skb);
-static int             ip6_pkt_prohibit_out(struct sock *sk, struct sk_buff *skb);
+static int             ip6_pkt_prohibit_out(struct net *net, struct sock *sk, struct sk_buff *skb);
 static void            ip6_link_failure(struct sk_buff *skb);
 static void            ip6_rt_update_pmtu(struct dst_entry *dst, struct sock *sk,
                                           struct sk_buff *skb, u32 mtu);
@@ -303,7 +304,7 @@ static const struct rt6_info ip6_blk_hole_entry_template = {
                .obsolete       = DST_OBSOLETE_FORCE_CHK,
                .error          = -EINVAL,
                .input          = dst_discard,
-               .output         = dst_discard_sk,
+               .output         = dst_discard_out,
        },
        .rt6i_flags     = (RTF_REJECT | RTF_NONEXTHOP),
        .rt6i_protocol  = RTPROT_KERNEL,
@@ -421,31 +422,7 @@ static bool rt6_check_expired(const struct rt6_info *rt)
 static int rt6_info_hash_nhsfn(unsigned int candidate_count,
                               const struct flowi6 *fl6)
 {
-       unsigned int val = fl6->flowi6_proto;
-
-       val ^= ipv6_addr_hash(&fl6->daddr);
-       val ^= ipv6_addr_hash(&fl6->saddr);
-
-       /* Work only if this not encapsulated */
-       switch (fl6->flowi6_proto) {
-       case IPPROTO_UDP:
-       case IPPROTO_TCP:
-       case IPPROTO_SCTP:
-               val ^= (__force u16)fl6->fl6_sport;
-               val ^= (__force u16)fl6->fl6_dport;
-               break;
-
-       case IPPROTO_ICMPV6:
-               val ^= (__force u16)fl6->fl6_icmp_type;
-               val ^= (__force u16)fl6->fl6_icmp_code;
-               break;
-       }
-       /* RFC6438 recommands to use flowlabel */
-       val ^= (__force u32)fl6->flowlabel;
-
-       /* Perhaps, we need to tune, this function? */
-       val = val ^ (val >> 7) ^ (val >> 12);
-       return val % candidate_count;
+       return get_hash_from_flowi6(fl6) % candidate_count;
 }
 
 static struct rt6_info *rt6_multipath_select(struct rt6_info *match,
@@ -498,10 +475,10 @@ static inline struct rt6_info *rt6_device_match(struct net *net,
                        if (dev->flags & IFF_LOOPBACK) {
                                if (!sprt->rt6i_idev ||
                                    sprt->rt6i_idev->dev->ifindex != oif) {
-                                       if (flags & RT6_LOOKUP_F_IFACE && oif)
+                                       if (flags & RT6_LOOKUP_F_IFACE)
                                                continue;
-                                       if (local && (!oif ||
-                                                     local->rt6i_idev->dev->ifindex == oif))
+                                       if (local &&
+                                           local->rt6i_idev->dev->ifindex == oif)
                                                continue;
                                }
                                local = sprt;
@@ -538,7 +515,7 @@ static void rt6_probe_deferred(struct work_struct *w)
                container_of(w, struct __rt6_probe_work, work);
 
        addrconf_addr_solict_mult(&work->target, &mcaddr);
-       ndisc_send_ns(work->dev, NULL, &work->target, &mcaddr, NULL, NULL);
+       ndisc_send_ns(work->dev, &work->target, &mcaddr, NULL, NULL);
        dev_put(work->dev);
        kfree(work);
 }
@@ -1168,7 +1145,7 @@ void ip6_route_input(struct sk_buff *skb)
        int flags = RT6_LOOKUP_F_HAS_SADDR;
        struct ip_tunnel_info *tun_info;
        struct flowi6 fl6 = {
-               .flowi6_iif = skb->dev->ifindex,
+               .flowi6_iif = l3mdev_fib_oif(skb->dev),
                .daddr = iph->daddr,
                .saddr = iph->saddr,
                .flowlabel = ip6_flowinfo(iph),
@@ -1192,9 +1169,14 @@ static struct rt6_info *ip6_pol_route_output(struct net *net, struct fib6_table
 struct dst_entry *ip6_route_output(struct net *net, const struct sock *sk,
                                    struct flowi6 *fl6)
 {
+       struct dst_entry *dst;
        int flags = 0;
        bool any_src;
 
+       dst = l3mdev_rt6_dst_by_oif(net, fl6);
+       if (dst)
+               return dst;
+
        fl6->flowi6_iif = LOOPBACK_IFINDEX;
 
        any_src = ipv6_addr_any(&fl6->saddr);
@@ -1223,7 +1205,7 @@ struct dst_entry *ip6_blackhole_route(struct net *net, struct dst_entry *dst_ori
                new = &rt->dst;
                new->__use = 1;
                new->input = dst_discard;
-               new->output = dst_discard_sk;
+               new->output = dst_discard_out;
 
                dst_copy_metrics(new, &ort->dst);
                rt->rt6i_idev = ort->rt6i_idev;
@@ -1749,21 +1731,21 @@ static int ip6_convert_metrics(struct mx6_config *mxc,
        return -EINVAL;
 }
 
-int ip6_route_info_create(struct fib6_config *cfg, struct rt6_info **rt_ret)
+static struct rt6_info *ip6_route_info_create(struct fib6_config *cfg)
 {
-       int err;
        struct net *net = cfg->fc_nlinfo.nl_net;
        struct rt6_info *rt = NULL;
        struct net_device *dev = NULL;
        struct inet6_dev *idev = NULL;
        struct fib6_table *table;
        int addr_type;
+       int err = -EINVAL;
 
        if (cfg->fc_dst_len > 128 || cfg->fc_src_len > 128)
-               return -EINVAL;
+               goto out;
 #ifndef CONFIG_IPV6_SUBTREES
        if (cfg->fc_src_len)
-               return -EINVAL;
+               goto out;
 #endif
        if (cfg->fc_ifindex) {
                err = -ENODEV;
@@ -1878,7 +1860,7 @@ int ip6_route_info_create(struct fib6_config *cfg, struct rt6_info **rt_ret)
                switch (cfg->fc_type) {
                case RTN_BLACKHOLE:
                        rt->dst.error = -EINVAL;
-                       rt->dst.output = dst_discard_sk;
+                       rt->dst.output = dst_discard_out;
                        rt->dst.input = dst_discard;
                        break;
                case RTN_PROHIBIT:
@@ -1983,9 +1965,7 @@ install_route:
 
        cfg->fc_nlinfo.nl_net = dev_net(dev);
 
-       *rt_ret = rt;
-
-       return 0;
+       return rt;
 out:
        if (dev)
                dev_put(dev);
@@ -1994,20 +1974,21 @@ out:
        if (rt)
                dst_free(&rt->dst);
 
-       *rt_ret = NULL;
-
-       return err;
+       return ERR_PTR(err);
 }
 
 int ip6_route_add(struct fib6_config *cfg)
 {
        struct mx6_config mxc = { .mx = NULL, };
-       struct rt6_info *rt = NULL;
+       struct rt6_info *rt;
        int err;
 
-       err = ip6_route_info_create(cfg, &rt);
-       if (err)
+       rt = ip6_route_info_create(cfg);
+       if (IS_ERR(rt)) {
+               err = PTR_ERR(rt);
+               rt = NULL;
                goto out;
+       }
 
        err = ip6_convert_metrics(&mxc, cfg);
        if (err)
@@ -2288,7 +2269,6 @@ static struct rt6_info *rt6_add_route_info(struct net *net,
                                           unsigned int pref)
 {
        struct fib6_config cfg = {
-               .fc_table       = RT6_TABLE_INFO,
                .fc_metric      = IP6_RT_PRIO_USER,
                .fc_ifindex     = ifindex,
                .fc_dst_len     = prefixlen,
@@ -2299,6 +2279,7 @@ static struct rt6_info *rt6_add_route_info(struct net *net,
                .fc_nlinfo.nl_net = net,
        };
 
+       cfg.fc_table = l3mdev_fib_table_by_index(net, ifindex) ? : RT6_TABLE_INFO;
        cfg.fc_dst = *prefix;
        cfg.fc_gateway = *gwaddr;
 
@@ -2339,7 +2320,7 @@ struct rt6_info *rt6_add_dflt_router(const struct in6_addr *gwaddr,
                                     unsigned int pref)
 {
        struct fib6_config cfg = {
-               .fc_table       = RT6_TABLE_DFLT,
+               .fc_table       = l3mdev_fib_table(dev) ? : RT6_TABLE_DFLT,
                .fc_metric      = IP6_RT_PRIO_USER,
                .fc_ifindex     = dev->ifindex,
                .fc_flags       = RTF_GATEWAY | RTF_ADDRCONF | RTF_DEFAULT |
@@ -2386,7 +2367,8 @@ static void rtmsg_to_fib6_config(struct net *net,
 {
        memset(cfg, 0, sizeof(*cfg));
 
-       cfg->fc_table = RT6_TABLE_MAIN;
+       cfg->fc_table = l3mdev_fib_table_by_index(net, rtmsg->rtmsg_ifindex) ?
+                        : RT6_TABLE_MAIN;
        cfg->fc_ifindex = rtmsg->rtmsg_ifindex;
        cfg->fc_metric = rtmsg->rtmsg_metric;
        cfg->fc_expires = rtmsg->rtmsg_info;
@@ -2470,7 +2452,7 @@ static int ip6_pkt_discard(struct sk_buff *skb)
        return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_INNOROUTES);
 }
 
-static int ip6_pkt_discard_out(struct sock *sk, struct sk_buff *skb)
+static int ip6_pkt_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb)
 {
        skb->dev = skb_dst(skb)->dev;
        return ip6_pkt_drop(skb, ICMPV6_NOROUTE, IPSTATS_MIB_OUTNOROUTES);
@@ -2481,7 +2463,7 @@ static int ip6_pkt_prohibit(struct sk_buff *skb)
        return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_INNOROUTES);
 }
 
-static int ip6_pkt_prohibit_out(struct sock *sk, struct sk_buff *skb)
+static int ip6_pkt_prohibit_out(struct net *net, struct sock *sk, struct sk_buff *skb)
 {
        skb->dev = skb_dst(skb)->dev;
        return ip6_pkt_drop(skb, ICMPV6_ADM_PROHIBITED, IPSTATS_MIB_OUTNOROUTES);
@@ -2495,6 +2477,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
                                    const struct in6_addr *addr,
                                    bool anycast)
 {
+       u32 tb_id;
        struct net *net = dev_net(idev->dev);
        struct rt6_info *rt = ip6_dst_alloc(net, net->loopback_dev,
                                            DST_NOCOUNT);
@@ -2517,7 +2500,8 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
        rt->rt6i_gateway  = *addr;
        rt->rt6i_dst.addr = *addr;
        rt->rt6i_dst.plen = 128;
-       rt->rt6i_table = fib6_get_table(net, RT6_TABLE_LOCAL);
+       tb_id = l3mdev_fib_table(idev->dev) ? : RT6_TABLE_LOCAL;
+       rt->rt6i_table = fib6_get_table(net, tb_id);
        rt->dst.flags |= DST_NOCACHE;
 
        atomic_set(&rt->dst.__refcnt, 1);
@@ -2896,9 +2880,12 @@ static int ip6_route_multipath_add(struct fib6_config *cfg)
                                r_cfg.fc_encap_type = nla_get_u16(nla);
                }
 
-               err = ip6_route_info_create(&r_cfg, &rt);
-               if (err)
+               rt = ip6_route_info_create(&r_cfg);
+               if (IS_ERR(rt)) {
+                       err = PTR_ERR(rt);
+                       rt = NULL;
                        goto cleanup;
+               }
 
                err = ip6_route_info_append(&rt6_nh_list, rt, &r_cfg);
                if (err) {
@@ -3277,6 +3264,11 @@ static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh)
        } else {
                fl6.flowi6_oif = oif;
 
+               if (netif_index_is_l3_master(net, oif)) {
+                       fl6.flowi6_flags = FLOWI_FLAG_L3MDEV_SRC |
+                                          FLOWI_FLAG_SKIP_NH_OIF;
+               }
+
                rt = (struct rt6_info *)ip6_route_output(net, NULL, &fl6);
        }