diff --git a/include/linux/filter.h b/include/linux/filter.h index dda418f10..7c7501971 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -601,6 +601,14 @@ struct bpf_skb_data_end { void *data_end; }; +struct bpf_nh_params { + u32 nh_family; + union { + u32 ipv4_nh; + struct in6_addr ipv6_nh; + }; +}; + struct bpf_redirect_info { u32 flags; u32 tgt_index; @@ -608,6 +616,7 @@ struct bpf_redirect_info { struct bpf_map *map; struct bpf_map *map_to_flush; u32 kern_flags; + struct bpf_nh_params nh; }; DECLARE_PER_CPU(struct bpf_redirect_info, bpf_redirect_info); diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 4a9f25dd5..a69cf8a19 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -3383,13 +3383,20 @@ union bpf_attr { * Return * The id is returned or 0 in case the id could not be retrieved. * - * long bpf_redirect_neigh(u32 ifindex, u64 flags) + * long bpf_redirect_neigh(u32 ifindex, struct bpf_redir_neigh *params, int plen, u64 flags) * Description * Redirect the packet to another net device of index *ifindex* * and fill in L2 addresses from neighboring subsystem. This helper * is somewhat similar to **bpf_redirect**\ (), except that it - * fills in e.g. MAC addresses based on the L3 information from - * the packet. This helper is supported for IPv4 and IPv6 protocols. + * populates L2 addresses as well, meaning, internally, the helper + * relies on the neighbor lookup for the L2 address of the nexthop. + * + * The helper will perform a FIB lookup based on the skb's + * networking header to get the address of the next hop, unless + * this is supplied by the caller in the *params* argument. The + * *plen* argument indicates the len of *params* and should be set + * to 0 if *params* is NULL. + * * The *flags* argument is reserved and must be 0. The helper is * currently only supported for tc BPF program types. * Return @@ -4336,6 +4343,16 @@ struct bpf_fib_lookup { __u8 dmac[6]; /* ETH_ALEN */ }; +struct bpf_redir_neigh { + /* network family for lookup (AF_INET, AF_INET6) */ + __u32 nh_family; + /* network address of nexthop; skips fib lookup to find gateway */ + union { + __be32 ipv4_nh; + __u32 ipv6_nh[4]; /* in6_addr; network order */ + }; +}; + enum bpf_task_fd_type { BPF_FD_TYPE_RAW_TRACEPOINT, /* tp name */ BPF_FD_TYPE_TRACEPOINT, /* tp name */ diff --git a/net/core/filter.c b/net/core/filter.c index fa683b312..093cfc23b 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2115,12 +2115,12 @@ static int __bpf_redirect(struct sk_buff *skb, struct net_device *dev, } #if IS_ENABLED(CONFIG_IPV6) -static int bpf_out_neigh_v6(struct net *net, struct sk_buff *skb) +static int bpf_out_neigh_v6(struct net *net, struct sk_buff *skb, + struct net_device *dev, struct bpf_nh_params *nh) { - struct dst_entry *dst = skb_dst(skb); - struct net_device *dev = dst->dev; u32 hh_len = LL_RESERVED_SPACE(dev); const struct in6_addr *nexthop; + struct dst_entry *dst = NULL; struct neighbour *neigh; if (dev_xmit_recursion()) { @@ -2146,8 +2146,13 @@ static int bpf_out_neigh_v6(struct net *net, struct sk_buff *skb) } rcu_read_lock_bh(); - nexthop = rt6_nexthop(container_of(dst, struct rt6_info, dst), - &ipv6_hdr(skb)->daddr); + if (!nh) { + dst = skb_dst(skb); + nexthop = rt6_nexthop(container_of(dst, struct rt6_info, dst), + &ipv6_hdr(skb)->daddr); + } else { + nexthop = &nh->ipv6_nh; + } neigh = ip_neigh_gw6(dev, nexthop); if (likely(!IS_ERR(neigh))) { int ret; @@ -2160,36 +2165,43 @@ static int bpf_out_neigh_v6(struct net *net, struct sk_buff *skb) return ret; } rcu_read_unlock_bh(); - IP6_INC_STATS(dev_net(dst->dev), - ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES); + if (dst) + IP6_INC_STATS(dev_net(dst->dev), + ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES); out_drop: kfree_skb(skb); return -ENETDOWN; } -static int __bpf_redirect_neigh_v6(struct sk_buff *skb, struct net_device *dev) +static int __bpf_redirect_neigh_v6(struct sk_buff *skb, struct net_device *dev, + struct bpf_nh_params *nh) { const struct ipv6hdr *ip6h = ipv6_hdr(skb); struct net *net = dev_net(dev); int err, ret = NET_XMIT_DROP; - struct dst_entry *dst; - struct flowi6 fl6 = { - .flowi6_flags = FLOWI_FLAG_ANYSRC, - .flowi6_mark = skb->mark, - .flowlabel = ip6_flowinfo(ip6h), - .flowi6_oif = dev->ifindex, - .flowi6_proto = ip6h->nexthdr, - .daddr = ip6h->daddr, - .saddr = ip6h->saddr, - }; - dst = ipv6_stub->ipv6_dst_lookup_flow(net, NULL, &fl6, NULL); - if (IS_ERR(dst)) + if (!nh) { + struct dst_entry *dst; + struct flowi6 fl6 = { + .flowi6_flags = FLOWI_FLAG_ANYSRC, + .flowi6_mark = skb->mark, + .flowlabel = ip6_flowinfo(ip6h), + .flowi6_oif = dev->ifindex, + .flowi6_proto = ip6h->nexthdr, + .daddr = ip6h->daddr, + .saddr = ip6h->saddr, + }; + + dst = ipv6_stub->ipv6_dst_lookup_flow(net, NULL, &fl6, NULL); + if (IS_ERR(dst)) + goto out_drop; + + skb_dst_set(skb, dst); + } else if (nh->nh_family != AF_INET6) { goto out_drop; + } - skb_dst_set(skb, dst); - - err = bpf_out_neigh_v6(net, skb); + err = bpf_out_neigh_v6(net, skb, dev, nh); if (unlikely(net_xmit_eval(err))) dev->stats.tx_errors++; else @@ -2202,7 +2214,8 @@ out_xmit: return ret; } #else -static int __bpf_redirect_neigh_v6(struct sk_buff *skb, struct net_device *dev) +static int __bpf_redirect_neigh_v6(struct sk_buff *skb, struct net_device *dev, + struct bpf_nh_params *nh) { kfree_skb(skb); return NET_XMIT_DROP; @@ -2210,11 +2223,9 @@ static int __bpf_redirect_neigh_v6(struct sk_buff *skb, struct net_device *dev) #endif /* CONFIG_IPV6 */ #if IS_ENABLED(CONFIG_INET) -static int bpf_out_neigh_v4(struct net *net, struct sk_buff *skb) +static int bpf_out_neigh_v4(struct net *net, struct sk_buff *skb, + struct net_device *dev, struct bpf_nh_params *nh) { - struct dst_entry *dst = skb_dst(skb); - struct rtable *rt = container_of(dst, struct rtable, dst); - struct net_device *dev = dst->dev; u32 hh_len = LL_RESERVED_SPACE(dev); struct neighbour *neigh; bool is_v6gw = false; @@ -2242,7 +2253,21 @@ static int bpf_out_neigh_v4(struct net *net, struct sk_buff *skb) } rcu_read_lock_bh(); - neigh = ip_neigh_for_gw(rt, skb, &is_v6gw); + if (!nh) { + struct dst_entry *dst = skb_dst(skb); + struct rtable *rt = container_of(dst, struct rtable, dst); + + neigh = ip_neigh_for_gw(rt, skb, &is_v6gw); + } else if (nh->nh_family == AF_INET6) { + neigh = ip_neigh_gw6(dev, &nh->ipv6_nh); + is_v6gw = true; + } else if (nh->nh_family == AF_INET) { + neigh = ip_neigh_gw4(dev, nh->ipv4_nh); + } else { + rcu_read_unlock_bh(); + goto out_drop; + } + if (likely(!IS_ERR(neigh))) { int ret; @@ -2259,33 +2284,37 @@ out_drop: return -ENETDOWN; } -static int __bpf_redirect_neigh_v4(struct sk_buff *skb, struct net_device *dev) +static int __bpf_redirect_neigh_v4(struct sk_buff *skb, struct net_device *dev, + struct bpf_nh_params *nh) { const struct iphdr *ip4h = ip_hdr(skb); struct net *net = dev_net(dev); int err, ret = NET_XMIT_DROP; - struct rtable *rt; - struct flowi4 fl4 = { - .flowi4_flags = FLOWI_FLAG_ANYSRC, - .flowi4_mark = skb->mark, - .flowi4_tos = RT_TOS(ip4h->tos), - .flowi4_oif = dev->ifindex, - .flowi4_proto = ip4h->protocol, - .daddr = ip4h->daddr, - .saddr = ip4h->saddr, - }; - rt = ip_route_output_flow(net, &fl4, NULL); - if (IS_ERR(rt)) - goto out_drop; - if (rt->rt_type != RTN_UNICAST && rt->rt_type != RTN_LOCAL) { - ip_rt_put(rt); - goto out_drop; + if (!nh) { + struct flowi4 fl4 = { + .flowi4_flags = FLOWI_FLAG_ANYSRC, + .flowi4_mark = skb->mark, + .flowi4_tos = RT_TOS(ip4h->tos), + .flowi4_oif = dev->ifindex, + .flowi4_proto = ip4h->protocol, + .daddr = ip4h->daddr, + .saddr = ip4h->saddr, + }; + struct rtable *rt; + + rt = ip_route_output_flow(net, &fl4, NULL); + if (IS_ERR(rt)) + goto out_drop; + if (rt->rt_type != RTN_UNICAST && rt->rt_type != RTN_LOCAL) { + ip_rt_put(rt); + goto out_drop; + } + + skb_dst_set(skb, &rt->dst); } - skb_dst_set(skb, &rt->dst); - - err = bpf_out_neigh_v4(net, skb); + err = bpf_out_neigh_v4(net, skb, dev, nh); if (unlikely(net_xmit_eval(err))) dev->stats.tx_errors++; else @@ -2298,14 +2327,16 @@ out_xmit: return ret; } #else -static int __bpf_redirect_neigh_v4(struct sk_buff *skb, struct net_device *dev) +static int __bpf_redirect_neigh_v4(struct sk_buff *skb, struct net_device *dev, + struct bpf_nh_params *nh) { kfree_skb(skb); return NET_XMIT_DROP; } #endif /* CONFIG_INET */ -static int __bpf_redirect_neigh(struct sk_buff *skb, struct net_device *dev) +static int __bpf_redirect_neigh(struct sk_buff *skb, struct net_device *dev, + struct bpf_nh_params *nh) { struct ethhdr *ethh = eth_hdr(skb); @@ -2320,9 +2351,9 @@ static int __bpf_redirect_neigh(struct sk_buff *skb, struct net_device *dev) skb_reset_network_header(skb); if (skb->protocol == htons(ETH_P_IP)) - return __bpf_redirect_neigh_v4(skb, dev); + return __bpf_redirect_neigh_v4(skb, dev, nh); else if (skb->protocol == htons(ETH_P_IPV6)) - return __bpf_redirect_neigh_v6(skb, dev); + return __bpf_redirect_neigh_v6(skb, dev, nh); out: kfree_skb(skb); return -ENOTSUPP; @@ -2331,7 +2362,8 @@ out: /* Internal, non-exposed redirect flags. */ enum { BPF_F_NEIGH = (1ULL << 1), -#define BPF_F_REDIRECT_INTERNAL (BPF_F_NEIGH) + BPF_F_NEXTHOP = (1ULL << 3), +#define BPF_F_REDIRECT_INTERNAL (BPF_F_NEIGH | BPF_F_PEER | BPF_F_NEXTHOP) }; BPF_CALL_3(bpf_clone_redirect, struct sk_buff *, skb, u32, ifindex, u64, flags) @@ -2391,7 +2423,8 @@ int skb_do_redirect(struct sk_buff *skb) } return flags & BPF_F_NEIGH ? - __bpf_redirect_neigh(skb, dev) : + __bpf_redirect_neigh(skb, dev, flags & BPF_F_NEXTHOP ? + &ri->nh : NULL) : __bpf_redirect(skb, dev, flags); } @@ -2416,16 +2449,21 @@ static const struct bpf_func_proto bpf_redirect_proto = { .arg2_type = ARG_ANYTHING, }; -BPF_CALL_2(bpf_redirect_neigh, u32, ifindex, u64, flags) +BPF_CALL_4(bpf_redirect_neigh, u32, ifindex, struct bpf_redir_neigh *, params, + int, plen, u64, flags) { struct bpf_redirect_info *ri = this_cpu_ptr(&bpf_redirect_info); - if (unlikely(flags)) + if (unlikely((plen && plen < sizeof(*params)) || flags)) return TC_ACT_SHOT; - ri->flags = BPF_F_NEIGH; + ri->flags = BPF_F_NEIGH | (plen ? BPF_F_NEXTHOP : 0); ri->tgt_index = ifindex; + BUILD_BUG_ON(sizeof(struct bpf_redir_neigh) != sizeof(struct bpf_nh_params)); + if (plen) + memcpy(&ri->nh, params, sizeof(ri->nh)); + return TC_ACT_REDIRECT; } @@ -2434,7 +2472,9 @@ static const struct bpf_func_proto bpf_redirect_neigh_proto = { .gpl_only = false, .ret_type = RET_INTEGER, .arg1_type = ARG_ANYTHING, - .arg2_type = ARG_ANYTHING, + .arg2_type = ARG_PTR_TO_MEM_OR_NULL, + .arg3_type = ARG_CONST_SIZE_OR_ZERO, + .arg4_type = ARG_ANYTHING, }; BPF_CALL_2(bpf_msg_apply_bytes, struct sk_msg *, msg, u32, bytes)