ipv6: Move gateway checks to a fib6_nh setting

The gateway setting is not per fib6_info entry but per-fib6_nh. Add a new
fib_nh_has_gw flag to fib6_nh and convert references to RTF_GATEWAY to
the new flag. For IPv6 address the flag is cheaper than checking that
nh_gw is non-0 like IPv4 does.

While this increases fib6_nh by 8-bytes, the effective allocation size of
a fib6_info is unchanged. The 8 bytes is recovered later with a
fib_nh_common change.

Signed-off-by: David Ahern <dsahern@gmail.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 52fed8c..663d819 100644 (file)
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -4913,7 +4913,7 @@ static void mlxsw_sp_rt6_destroy(struct mlxsw_sp_rt6 *mlxsw_sp_rt6)
 static bool mlxsw_sp_fib6_rt_can_mp(const struct fib6_info *rt)
 {
        /* RTF_CACHE routes are ignored */
-       return (rt->fib6_flags & (RTF_GATEWAY | RTF_ADDRCONF)) == RTF_GATEWAY;
+       return !(rt->fib6_flags & RTF_ADDRCONF) && rt->fib6_nh.fib_nh_has_gw;
 }
 
 static struct fib6_info *
@@ -5053,7 +5053,7 @@ static void mlxsw_sp_nexthop6_fini(struct mlxsw_sp *mlxsw_sp,
 static bool mlxsw_sp_rt6_is_gateway(const struct mlxsw_sp *mlxsw_sp,
                                    const struct fib6_info *rt)
 {
-       return rt->fib6_flags & RTF_GATEWAY ||
+       return rt->fib6_nh.fib_nh_has_gw ||
               mlxsw_sp_nexthop6_ipip_type(mlxsw_sp, rt, NULL);
 }
 

diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index 2d2a468..3b04b31 100644 (file)
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -126,6 +126,7 @@ struct rt6_exception {
 
 struct fib6_nh {
        struct in6_addr         nh_gw;
+       bool                    fib_nh_has_gw;
        struct net_device       *nh_dev;
        struct lwtunnel_state   *nh_lwtstate;
 

diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 7ab1199..95cd8a2 100644 (file)
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -68,8 +68,8 @@ static inline bool rt6_need_strict(const struct in6_addr *daddr)
 
 static inline bool rt6_qualify_for_ecmp(const struct fib6_info *f6i)
 {
-       return (f6i->fib6_flags & (RTF_GATEWAY|RTF_ADDRCONF|RTF_DYNAMIC)) ==
-              RTF_GATEWAY;
+       return !(f6i->fib6_flags & (RTF_ADDRCONF|RTF_DYNAMIC)) &&
+               f6i->fib6_nh.fib_nh_has_gw;
 }
 
 void ip6_route_input(struct sk_buff *skb);

diff --git a/net/core/filter.c b/net/core/filter.c
index 22eb2ed..e778476 100644 (file)
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -4751,7 +4751,7 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params,
        if (f6i->fib6_nh.nh_lwtstate)
                return BPF_FIB_LKUP_RET_UNSUPP_LWT;
 
-       if (f6i->fib6_flags & RTF_GATEWAY)
+       if (f6i->fib6_nh.fib_nh_has_gw)
                *dst = f6i->fib6_nh.nh_gw;
 
        dev = f6i->fib6_nh.nh_dev;

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 4ae17a9..c5ac08f 100644 (file)
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -173,7 +173,8 @@ static int addrconf_ifdown(struct net_device *dev, int how);
 static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
                                                  int plen,
                                                  const struct net_device *dev,
-                                                 u32 flags, u32 noflags);
+                                                 u32 flags, u32 noflags,
+                                                 bool no_gw);
 
 static void addrconf_dad_start(struct inet6_ifaddr *ifp);
 static void addrconf_dad_work(struct work_struct *w);
@@ -1230,10 +1231,8 @@ cleanup_prefix_route(struct inet6_ifaddr *ifp, unsigned long expires, bool del_r
 {
        struct fib6_info *f6i;
 
-       f6i = addrconf_get_prefix_route(&ifp->addr,
-                                      ifp->prefix_len,
-                                      ifp->idev->dev,
-                                      0, RTF_GATEWAY | RTF_DEFAULT);
+       f6i = addrconf_get_prefix_route(&ifp->addr, ifp->prefix_len,
+                                       ifp->idev->dev, 0, RTF_DEFAULT, true);
        if (f6i) {
                if (del_rt)
                        ip6_del_rt(dev_net(ifp->idev->dev), f6i);
@@ -2402,7 +2401,8 @@ addrconf_prefix_route(struct in6_addr *pfx, int plen, u32 metric,
 static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
                                                  int plen,
                                                  const struct net_device *dev,
-                                                 u32 flags, u32 noflags)
+                                                 u32 flags, u32 noflags,
+                                                 bool no_gw)
 {
        struct fib6_node *fn;
        struct fib6_info *rt = NULL;
@@ -2421,6 +2421,8 @@ static struct fib6_info *addrconf_get_prefix_route(const struct in6_addr *pfx,
        for_each_fib6_node_rt_rcu(fn) {
                if (rt->fib6_nh.nh_dev->ifindex != dev->ifindex)
                        continue;
+               if (no_gw && rt->fib6_nh.fib_nh_has_gw)
+                       continue;
                if ((rt->fib6_flags & flags) != flags)
                        continue;
                if ((rt->fib6_flags & noflags) != 0)
@@ -2717,7 +2719,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len, bool sllao)
                                               pinfo->prefix_len,
                                               dev,
                                               RTF_ADDRCONF | RTF_PREFIX_RT,
-                                              RTF_GATEWAY | RTF_DEFAULT);
+                                              RTF_DEFAULT, true);
 
                if (rt) {
                        /* Autoconf prefix route */
@@ -4588,10 +4590,8 @@ static int modify_prefix_route(struct inet6_ifaddr *ifp,
        struct fib6_info *f6i;
        u32 prio;
 
-       f6i = addrconf_get_prefix_route(&ifp->addr,
-                                       ifp->prefix_len,
-                                       ifp->idev->dev,
-                                       0, RTF_GATEWAY | RTF_DEFAULT);
+       f6i = addrconf_get_prefix_route(&ifp->addr, ifp->prefix_len,
+                                       ifp->idev->dev, 0, RTF_DEFAULT, true);
        if (!f6i)
                return -ENOENT;
 
@@ -5972,7 +5972,8 @@ static void __ipv6_ifa_notify(int event, struct inet6_ifaddr *ifp)
                        struct fib6_info *rt;
 
                        rt = addrconf_get_prefix_route(&ifp->peer_addr, 128,
-                                                      ifp->idev->dev, 0, 0);
+                                                      ifp->idev->dev, 0, 0,
+                                                      false);
                        if (rt)
                                ip6_del_rt(net, rt);
                }

diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index db88608..91ce84e 100644 (file)
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -2294,6 +2294,7 @@ static int ipv6_route_seq_show(struct seq_file *seq, void *v)
 {
        struct fib6_info *rt = v;
        struct ipv6_route_iter *iter = seq->private;
+       unsigned int flags = rt->fib6_flags;
        const struct net_device *dev;
 
        seq_printf(seq, "%pi6 %02x ", &rt->fib6_dst.addr, rt->fib6_dst.plen);
@@ -2303,15 +2304,17 @@ static int ipv6_route_seq_show(struct seq_file *seq, void *v)
 #else
        seq_puts(seq, "00000000000000000000000000000000 00 ");
 #endif
-       if (rt->fib6_flags & RTF_GATEWAY)
+       if (rt->fib6_nh.fib_nh_has_gw) {
+               flags |= RTF_GATEWAY;
                seq_printf(seq, "%pi6", &rt->fib6_nh.nh_gw);
-       else
+       } else {
                seq_puts(seq, "00000000000000000000000000000000");
+       }
 
        dev = rt->fib6_nh.nh_dev;
        seq_printf(seq, " %08x %08x %08x %08x %8s\n",
                   rt->fib6_metric, atomic_read(&rt->fib6_ref), 0,
-                  rt->fib6_flags, dev ? dev->name : "");
+                  flags, dev ? dev->name : "");
        iter->w.leaf = NULL;
        return 0;
 }

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 5f453c7..69c96cf 100644 (file)
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -533,7 +533,7 @@ static void rt6_probe(struct fib6_info *rt)
         * Router Reachability Probe MUST be rate-limited
         * to no more than one per minute.
         */
-       if (!rt || !(rt->fib6_flags & RTF_GATEWAY))
+       if (!rt || !rt->fib6_nh.fib_nh_has_gw)
                return;
 
        nh_gw = &rt->fib6_nh.nh_gw;
@@ -595,7 +595,7 @@ static inline enum rt6_nud_state rt6_check_neigh(struct fib6_info *rt)
        struct neighbour *neigh;
 
        if (rt->fib6_flags & RTF_NONEXTHOP ||
-           !(rt->fib6_flags & RTF_GATEWAY))
+           !rt->fib6_nh.fib_nh_has_gw)
                return RT6_NUD_SUCCEED;
 
        rcu_read_lock_bh();
@@ -784,7 +784,7 @@ static struct fib6_info *rt6_select(struct net *net, struct fib6_node *fn,
 
 static bool rt6_is_gw_or_nonexthop(const struct fib6_info *rt)
 {
-       return (rt->fib6_flags & (RTF_NONEXTHOP | RTF_GATEWAY));
+       return (rt->fib6_flags & RTF_NONEXTHOP) || rt->fib6_nh.fib_nh_has_gw;
 }
 
 #ifdef CONFIG_IPV6_ROUTE_INFO
@@ -989,8 +989,11 @@ static void ip6_rt_copy_init(struct rt6_info *rt, struct fib6_info *ort)
 
        rt->rt6i_dst = ort->fib6_dst;
        rt->rt6i_idev = dev ? in6_dev_get(dev) : NULL;
-       rt->rt6i_gateway = ort->fib6_nh.nh_gw;
        rt->rt6i_flags = ort->fib6_flags;
+       if (ort->fib6_nh.fib_nh_has_gw) {
+               rt->rt6i_gateway = ort->fib6_nh.nh_gw;
+               rt->rt6i_flags |= RTF_GATEWAY;
+       }
        rt6_set_from(rt, ort);
 #ifdef CONFIG_IPV6_SUBTREES
        rt->rt6i_src = ort->fib6_src;
@@ -1872,7 +1875,7 @@ struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table,
                rcu_read_unlock();
                return rt;
        } else if (unlikely((fl6->flowi6_flags & FLOWI_FLAG_KNOWN_NH) &&
-                           !(f6i->fib6_flags & RTF_GATEWAY))) {
+                           !f6i->fib6_nh.fib_nh_has_gw)) {
                /* Create a RTF_CACHE clone which will not be
                 * owned by the fib6 tree.  It is for the special case where
                 * the daddr in the skb during the neighbor look-up is different
@@ -2442,7 +2445,7 @@ restart:
                        continue;
                if (rt->fib6_flags & RTF_REJECT)
                        break;
-               if (!(rt->fib6_flags & RTF_GATEWAY))
+               if (!rt->fib6_nh.fib_nh_has_gw)
                        continue;
                if (fl6->flowi6_oif != rt->fib6_nh.nh_dev->ifindex)
                        continue;
@@ -2986,6 +2989,7 @@ int fib6_nh_init(struct net *net, struct fib6_nh *fib6_nh,
                        goto out;
 
                fib6_nh->nh_gw = cfg->fc_gateway;
+               fib6_nh->fib_nh_has_gw = 1;
        }
 
        err = -ENODEV;
@@ -3121,7 +3125,7 @@ static struct fib6_info *ip6_route_info_create(struct fib6_config *cfg,
        rt->fib6_table = table;
        rt->fib6_metric = cfg->fc_metric;
        rt->fib6_type = cfg->fc_type;
-       rt->fib6_flags = cfg->fc_flags;
+       rt->fib6_flags = cfg->fc_flags & ~RTF_GATEWAY;
 
        ipv6_addr_prefix(&rt->fib6_dst.addr, &cfg->fc_dst, cfg->fc_dst_len);
        rt->fib6_dst.plen = cfg->fc_dst_len;
@@ -3490,7 +3494,8 @@ static struct fib6_info *rt6_get_route_info(struct net *net,
        for_each_fib6_node_rt_rcu(fn) {
                if (rt->fib6_nh.nh_dev->ifindex != ifindex)
                        continue;
-               if ((rt->fib6_flags & (RTF_ROUTEINFO|RTF_GATEWAY)) != (RTF_ROUTEINFO|RTF_GATEWAY))
+               if (!(rt->fib6_flags & RTF_ROUTEINFO) ||
+                   !rt->fib6_nh.fib_nh_has_gw)
                        continue;
                if (!ipv6_addr_equal(&rt->fib6_nh.nh_gw, gwaddr))
                        continue;
@@ -3811,7 +3816,7 @@ void rt6_remove_prefsrc(struct inet6_ifaddr *ifp)
        fib6_clean_all(net, fib6_remove_prefsrc, &adni);
 }
 
-#define RTF_RA_ROUTER          (RTF_ADDRCONF | RTF_DEFAULT | RTF_GATEWAY)
+#define RTF_RA_ROUTER          (RTF_ADDRCONF | RTF_DEFAULT)
 
 /* Remove routers and update dst entries when gateway turn into host. */
 static int fib6_clean_tohost(struct fib6_info *rt, void *arg)
@@ -3819,6 +3824,7 @@ static int fib6_clean_tohost(struct fib6_info *rt, void *arg)
        struct in6_addr *gateway = (struct in6_addr *)arg;
 
        if (((rt->fib6_flags & RTF_RA_ROUTER) == RTF_RA_ROUTER) &&
+           rt->fib6_nh.fib_nh_has_gw &&
            ipv6_addr_equal(gateway, &rt->fib6_nh.nh_gw)) {
                return -1;
        }
@@ -4607,7 +4613,7 @@ static int rt6_nexthop_info(struct sk_buff *skb, struct fib6_info *rt,
                rcu_read_unlock();
        }
 
-       if (rt->fib6_flags & RTF_GATEWAY) {
+       if (rt->fib6_nh.fib_nh_has_gw) {
                if (nla_put_in6_addr(skb, RTA_GATEWAY, &rt->fib6_nh.nh_gw) < 0)
                        goto nla_put_failure;
        }