diff options
Diffstat (limited to 'net/ipv4/udp.c')
| -rw-r--r-- | net/ipv4/udp.c | 1069 |
1 files changed, 738 insertions, 331 deletions
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 0794a2c46a56..ffe074cb5865 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -68,7 +68,7 @@ * YOSHIFUJI Hideaki @USAGI and: Support IPV6_V6ONLY socket option, which * Alexey Kuznetsov: allow both IPv4 and IPv6 sockets to bind * a single port at the same time. - * Derek Atkins <derek@ihtfp.com>: Add Encapulation Support + * Derek Atkins <derek@ihtfp.com>: Add Encapsulation Support * James Chapman : Add L2TP encapsulation type. */ @@ -93,6 +93,7 @@ #include <linux/inet.h> #include <linux/netdevice.h> #include <linux/slab.h> +#include <linux/sock_diag.h> #include <net/tcp_states.h> #include <linux/skbuff.h> #include <linux/proc_fs.h> @@ -100,6 +101,7 @@ #include <net/net_namespace.h> #include <net/icmp.h> #include <net/inet_hashtables.h> +#include <net/ip.h> #include <net/ip_tunnels.h> #include <net/route.h> #include <net/checksum.h> @@ -118,15 +120,13 @@ #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6_stubs.h> #endif +#include <net/rps.h> struct udp_table udp_table __read_mostly; -EXPORT_SYMBOL(udp_table); long sysctl_udp_mem[3] __read_mostly; -EXPORT_SYMBOL(sysctl_udp_mem); +EXPORT_IPV6_MOD(sysctl_udp_mem); -atomic_long_t udp_memory_allocated ____cacheline_aligned_in_smp; -EXPORT_SYMBOL(udp_memory_allocated); DEFINE_PER_CPU(int, udp_memory_per_cpu_fw_alloc); EXPORT_PER_CPU_SYMBOL_GPL(udp_memory_per_cpu_fw_alloc); @@ -143,8 +143,8 @@ static int udp_lib_lport_inuse(struct net *net, __u16 num, unsigned long *bitmap, struct sock *sk, unsigned int log) { + kuid_t uid = sk_uid(sk); struct sock *sk2; - kuid_t uid = sock_i_uid(sk); sk_for_each(sk2, &hslot->head) { if (net_eq(sock_net(sk2), net) && @@ -156,7 +156,7 @@ static int udp_lib_lport_inuse(struct net *net, __u16 num, inet_rcv_saddr_equal(sk, sk2, true)) { if (sk2->sk_reuseport && sk->sk_reuseport && !rcu_access_pointer(sk->sk_reuseport_cb) && - uid_eq(uid, sock_i_uid(sk2))) { + uid_eq(uid, sk_uid(sk2))) { if (!bitmap) return 0; } else { @@ -178,8 +178,8 @@ static int udp_lib_lport_inuse2(struct net *net, __u16 num, struct udp_hslot *hslot2, struct sock *sk) { + kuid_t uid = sk_uid(sk); struct sock *sk2; - kuid_t uid = sock_i_uid(sk); int res = 0; spin_lock(&hslot2->lock); @@ -193,7 +193,7 @@ static int udp_lib_lport_inuse2(struct net *net, __u16 num, inet_rcv_saddr_equal(sk, sk2, true)) { if (sk2->sk_reuseport && sk->sk_reuseport && !rcu_access_pointer(sk->sk_reuseport_cb) && - uid_eq(uid, sock_i_uid(sk2))) { + uid_eq(uid, sk_uid(sk2))) { res = 0; } else { res = 1; @@ -208,7 +208,7 @@ static int udp_lib_lport_inuse2(struct net *net, __u16 num, static int udp_reuseport_add_sock(struct sock *sk, struct udp_hslot *hslot) { struct net *net = sock_net(sk); - kuid_t uid = sock_i_uid(sk); + kuid_t uid = sk_uid(sk); struct sock *sk2; sk_for_each(sk2, &hslot->head) { @@ -218,7 +218,7 @@ static int udp_reuseport_add_sock(struct sock *sk, struct udp_hslot *hslot) ipv6_only_sock(sk2) == ipv6_only_sock(sk) && (udp_sk(sk2)->udp_port_hash == udp_sk(sk)->udp_port_hash) && (sk2->sk_bound_dev_if == sk->sk_bound_dev_if) && - sk2->sk_reuseport && uid_eq(uid, sock_i_uid(sk2)) && + sk2->sk_reuseport && uid_eq(uid, sk_uid(sk2)) && inet_rcv_saddr_equal(sk, sk2, false)) { return reuseport_add_sock(sk, sk2, inet_rcv_saddr_any(sk)); @@ -326,6 +326,8 @@ found: goto fail_unlock; } + sock_set_flag(sk, SOCK_RCU_FREE); + sk_add_node_rcu(sk, &hslot->head); hslot->count++; sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); @@ -342,14 +344,14 @@ found: hslot2->count++; spin_unlock(&hslot2->lock); } - sock_set_flag(sk, SOCK_RCU_FREE); + error = 0; fail_unlock: spin_unlock_bh(&hslot->lock); fail: return error; } -EXPORT_SYMBOL(udp_lib_get_port); +EXPORT_IPV6_MOD(udp_lib_get_port); int udp_v4_get_port(struct sock *sk, unsigned short snum) { @@ -363,7 +365,7 @@ int udp_v4_get_port(struct sock *sk, unsigned short snum) return udp_lib_get_port(sk, snum, hash2_nulladdr); } -static int compute_score(struct sock *sk, struct net *net, +static int compute_score(struct sock *sk, const struct net *net, __be32 saddr, __be16 sport, __be32 daddr, unsigned short hnum, int dif, int sdif) @@ -407,20 +409,61 @@ static int compute_score(struct sock *sk, struct net *net, return score; } -INDIRECT_CALLABLE_SCOPE u32 udp_ehashfn(const struct net *net, const __be32 laddr, const __u16 lport, const __be32 faddr, const __be16 fport) { - static u32 udp_ehash_secret __read_mostly; - net_get_random_once(&udp_ehash_secret, sizeof(udp_ehash_secret)); return __inet_ehashfn(laddr, lport, faddr, fport, udp_ehash_secret + net_hash_mix(net)); } +EXPORT_IPV6_MOD(udp_ehashfn); + +/** + * udp4_lib_lookup1() - Simplified lookup using primary hash (destination port) + * @net: Network namespace + * @saddr: Source address, network order + * @sport: Source port, network order + * @daddr: Destination address, network order + * @hnum: Destination port, host order + * @dif: Destination interface index + * @sdif: Destination bridge port index, if relevant + * @udptable: Set of UDP hash tables + * + * Simplified lookup to be used as fallback if no sockets are found due to a + * potential race between (receive) address change, and lookup happening before + * the rehash operation. This function ignores SO_REUSEPORT groups while scoring + * result sockets, because if we have one, we don't need the fallback at all. + * + * Called under rcu_read_lock(). + * + * Return: socket with highest matching score if any, NULL if none + */ +static struct sock *udp4_lib_lookup1(const struct net *net, + __be32 saddr, __be16 sport, + __be32 daddr, unsigned int hnum, + int dif, int sdif, + const struct udp_table *udptable) +{ + unsigned int slot = udp_hashfn(net, hnum, udptable->mask); + struct udp_hslot *hslot = &udptable->hash[slot]; + struct sock *sk, *result = NULL; + int score, badness = 0; + + sk_for_each_rcu(sk, &hslot->head) { + score = compute_score(sk, net, + saddr, sport, daddr, hnum, dif, sdif); + if (score > badness) { + result = sk; + badness = score; + } + } + + return result; +} /* called with rcu_read_lock() */ -static struct sock *udp4_lib_lookup2(struct net *net, +static struct sock *udp4_lib_lookup2(const struct net *net, __be32 saddr, __be16 sport, __be32 daddr, unsigned int hnum, int dif, int sdif, @@ -429,15 +472,21 @@ static struct sock *udp4_lib_lookup2(struct net *net, { struct sock *sk, *result; int score, badness; + bool need_rescore; result = NULL; badness = 0; udp_portaddr_for_each_entry_rcu(sk, &hslot2->head) { - score = compute_score(sk, net, saddr, sport, - daddr, hnum, dif, sdif); + need_rescore = false; +rescore: + score = compute_score(need_rescore ? result : sk, net, saddr, + sport, daddr, hnum, dif, sdif); if (score > badness) { badness = score; + if (need_rescore) + continue; + if (sk->sk_state == TCP_ESTABLISHED) { result = sk; continue; @@ -458,29 +507,191 @@ static struct sock *udp4_lib_lookup2(struct net *net, if (IS_ERR(result)) continue; - badness = compute_score(result, net, saddr, sport, - daddr, hnum, dif, sdif); - + /* compute_score is too long of a function to be + * inlined, and calling it again here yields + * measurable overhead for some + * workloads. Work around it by jumping + * backwards to rescore 'result'. + */ + need_rescore = true; + goto rescore; } } return result; } +#if IS_ENABLED(CONFIG_BASE_SMALL) +static struct sock *udp4_lib_lookup4(const struct net *net, + __be32 saddr, __be16 sport, + __be32 daddr, unsigned int hnum, + int dif, int sdif, + struct udp_table *udptable) +{ + return NULL; +} + +static void udp_rehash4(struct udp_table *udptable, struct sock *sk, + u16 newhash4) +{ +} + +static void udp_unhash4(struct udp_table *udptable, struct sock *sk) +{ +} +#else /* !CONFIG_BASE_SMALL */ +static struct sock *udp4_lib_lookup4(const struct net *net, + __be32 saddr, __be16 sport, + __be32 daddr, unsigned int hnum, + int dif, int sdif, + struct udp_table *udptable) +{ + const __portpair ports = INET_COMBINED_PORTS(sport, hnum); + const struct hlist_nulls_node *node; + struct udp_hslot *hslot4; + unsigned int hash4, slot; + struct udp_sock *up; + struct sock *sk; + + hash4 = udp_ehashfn(net, daddr, hnum, saddr, sport); + slot = hash4 & udptable->mask; + hslot4 = &udptable->hash4[slot]; + INET_ADDR_COOKIE(acookie, saddr, daddr); + +begin: + /* SLAB_TYPESAFE_BY_RCU not used, so we don't need to touch sk_refcnt */ + udp_lrpa_for_each_entry_rcu(up, node, &hslot4->nulls_head) { + sk = (struct sock *)up; + if (inet_match(net, sk, acookie, ports, dif, sdif)) + return sk; + } + + /* if the nulls value we got at the end of this lookup is not the + * expected one, we must restart lookup. We probably met an item that + * was moved to another chain due to rehash. + */ + if (get_nulls_value(node) != slot) + goto begin; + + return NULL; +} + +/* udp_rehash4() only checks hslot4, and hash4_cnt is not processed. */ +static void udp_rehash4(struct udp_table *udptable, struct sock *sk, + u16 newhash4) +{ + struct udp_hslot *hslot4, *nhslot4; + + hslot4 = udp_hashslot4(udptable, udp_sk(sk)->udp_lrpa_hash); + nhslot4 = udp_hashslot4(udptable, newhash4); + udp_sk(sk)->udp_lrpa_hash = newhash4; + + if (hslot4 != nhslot4) { + spin_lock_bh(&hslot4->lock); + hlist_nulls_del_init_rcu(&udp_sk(sk)->udp_lrpa_node); + hslot4->count--; + spin_unlock_bh(&hslot4->lock); + + spin_lock_bh(&nhslot4->lock); + hlist_nulls_add_head_rcu(&udp_sk(sk)->udp_lrpa_node, + &nhslot4->nulls_head); + nhslot4->count++; + spin_unlock_bh(&nhslot4->lock); + } +} + +static void udp_unhash4(struct udp_table *udptable, struct sock *sk) +{ + struct udp_hslot *hslot2, *hslot4; + + if (udp_hashed4(sk)) { + hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); + hslot4 = udp_hashslot4(udptable, udp_sk(sk)->udp_lrpa_hash); + + spin_lock(&hslot4->lock); + hlist_nulls_del_init_rcu(&udp_sk(sk)->udp_lrpa_node); + hslot4->count--; + spin_unlock(&hslot4->lock); + + spin_lock(&hslot2->lock); + udp_hash4_dec(hslot2); + spin_unlock(&hslot2->lock); + } +} + +void udp_lib_hash4(struct sock *sk, u16 hash) +{ + struct udp_hslot *hslot, *hslot2, *hslot4; + struct net *net = sock_net(sk); + struct udp_table *udptable; + + /* Connected udp socket can re-connect to another remote address, which + * will be handled by rehash. Thus no need to redo hash4 here. + */ + if (udp_hashed4(sk)) + return; + + udptable = net->ipv4.udp_table; + hslot = udp_hashslot(udptable, net, udp_sk(sk)->udp_port_hash); + hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); + hslot4 = udp_hashslot4(udptable, hash); + udp_sk(sk)->udp_lrpa_hash = hash; + + spin_lock_bh(&hslot->lock); + if (rcu_access_pointer(sk->sk_reuseport_cb)) + reuseport_detach_sock(sk); + + spin_lock(&hslot4->lock); + hlist_nulls_add_head_rcu(&udp_sk(sk)->udp_lrpa_node, + &hslot4->nulls_head); + hslot4->count++; + spin_unlock(&hslot4->lock); + + spin_lock(&hslot2->lock); + udp_hash4_inc(hslot2); + spin_unlock(&hslot2->lock); + + spin_unlock_bh(&hslot->lock); +} +EXPORT_IPV6_MOD(udp_lib_hash4); + +/* call with sock lock */ +void udp4_hash4(struct sock *sk) +{ + struct net *net = sock_net(sk); + unsigned int hash; + + if (sk_unhashed(sk) || sk->sk_rcv_saddr == htonl(INADDR_ANY)) + return; + + hash = udp_ehashfn(net, sk->sk_rcv_saddr, sk->sk_num, + sk->sk_daddr, sk->sk_dport); + + udp_lib_hash4(sk, hash); +} +EXPORT_IPV6_MOD(udp4_hash4); +#endif /* CONFIG_BASE_SMALL */ + /* UDP is nearly always wildcards out the wazoo, it makes no sense to try * harder than this. -DaveM */ -struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr, +struct sock *__udp4_lib_lookup(const struct net *net, __be32 saddr, __be16 sport, __be32 daddr, __be16 dport, int dif, int sdif, struct udp_table *udptable, struct sk_buff *skb) { unsigned short hnum = ntohs(dport); - unsigned int hash2, slot2; struct udp_hslot *hslot2; struct sock *result, *sk; + unsigned int hash2; hash2 = ipv4_portaddr_hash(net, daddr, hnum); - slot2 = hash2 & udptable->mask; - hslot2 = &udptable->hash2[slot2]; + hslot2 = udp_hashslot2(udptable, hash2); + + if (udp_has_hash4(hslot2)) { + result = udp4_lib_lookup4(net, saddr, sport, daddr, hnum, + dif, sdif, udptable); + if (result) /* udp4_lib_lookup4 return sk or NULL */ + return result; + } /* Lookup connected or non-wildcard socket */ result = udp4_lib_lookup2(net, saddr, sport, @@ -507,12 +718,24 @@ struct sock *__udp4_lib_lookup(struct net *net, __be32 saddr, /* Lookup wildcard sockets */ hash2 = ipv4_portaddr_hash(net, htonl(INADDR_ANY), hnum); - slot2 = hash2 & udptable->mask; - hslot2 = &udptable->hash2[slot2]; + hslot2 = udp_hashslot2(udptable, hash2); result = udp4_lib_lookup2(net, saddr, sport, htonl(INADDR_ANY), hnum, dif, sdif, hslot2, skb); + if (!IS_ERR_OR_NULL(result)) + goto done; + + /* Primary hash (destination port) lookup as fallback for this race: + * 1. __ip4_datagram_connect() sets sk_rcv_saddr + * 2. lookup (this function): new sk_rcv_saddr, hashes not updated yet + * 3. rehash operation updating _secondary and four-tuple_ hashes + * The primary hash doesn't need an update after 1., so, thanks to this + * further step, 1. and 3. don't need to be atomic against the lookup. + */ + result = udp4_lib_lookup1(net, saddr, sport, daddr, hnum, dif, sdif, + udptable); + done: if (IS_ERR(result)) return NULL; @@ -534,7 +757,8 @@ static inline struct sock *__udp4_lib_lookup_skb(struct sk_buff *skb, struct sock *udp4_lib_lookup_skb(const struct sk_buff *skb, __be16 sport, __be16 dport) { - const struct iphdr *iph = ip_hdr(skb); + const u16 offset = NAPI_GRO_CB(skb)->network_offsets[skb->encapsulation]; + const struct iphdr *iph = (struct iphdr *)(skb->data + offset); struct net *net = dev_net(skb->dev); int iif, sdif; @@ -549,7 +773,7 @@ struct sock *udp4_lib_lookup_skb(const struct sk_buff *skb, * Does increment socket refcount. */ #if IS_ENABLED(CONFIG_NF_TPROXY_IPV4) || IS_ENABLED(CONFIG_NF_SOCKET_IPV4) -struct sock *udp4_lib_lookup(struct net *net, __be32 saddr, __be16 sport, +struct sock *udp4_lib_lookup(const struct net *net, __be32 saddr, __be16 sport, __be32 daddr, __be16 dport, int dif) { struct sock *sk; @@ -584,6 +808,13 @@ static inline bool __udp_is_mcast_sock(struct net *net, const struct sock *sk, } DEFINE_STATIC_KEY_FALSE(udp_encap_needed_key); +EXPORT_IPV6_MOD(udp_encap_needed_key); + +#if IS_ENABLED(CONFIG_IPV6) +DEFINE_STATIC_KEY_FALSE(udpv6_encap_needed_key); +EXPORT_IPV6_MOD(udpv6_encap_needed_key); +#endif + void udp_encap_enable(void) { static_branch_inc(&udp_encap_needed_key); @@ -714,7 +945,7 @@ int __udp4_lib_err(struct sk_buff *skb, u32 info, struct udp_table *udptable) iph->saddr, uh->source, skb->dev->ifindex, inet_sdif(skb), udptable, NULL); - if (!sk || udp_sk(sk)->encap_type) { + if (!sk || READ_ONCE(udp_sk(sk)->encap_type)) { /* No socket for error: try tunnels before discarding */ if (static_branch_unlikely(&udp_encap_needed_key)) { sk = __udp4_lib_err_encap(net, iph, uh, udptable, sk, skb, @@ -750,7 +981,7 @@ int __udp4_lib_err(struct sk_buff *skb, u32 info, struct udp_table *udptable) case ICMP_DEST_UNREACH: if (code == ICMP_FRAG_NEEDED) { /* Path MTU discovery */ ipv4_sk_update_pmtu(skb, sk, info); - if (inet->pmtudisc != IP_PMTUDISC_DONT) { + if (READ_ONCE(inet->pmtudisc) != IP_PMTUDISC_DONT) { err = EMSGSIZE; harderr = 1; break; @@ -805,11 +1036,11 @@ void udp_flush_pending_frames(struct sock *sk) if (up->pending) { up->len = 0; - up->pending = 0; + WRITE_ONCE(up->pending, 0); ip_flush_pending_frames(sk); } } -EXPORT_SYMBOL(udp_flush_pending_frames); +EXPORT_IPV6_MOD(udp_flush_pending_frames); /** * udp4_hwcsum - handle outgoing HW checksumming @@ -909,9 +1140,9 @@ static int udp_send_skb(struct sk_buff *skb, struct flowi4 *fl4, const int hlen = skb_network_header_len(skb) + sizeof(struct udphdr); - if (hlen + cork->gso_size > cork->fragsize) { + if (hlen + min(datalen, cork->gso_size) > cork->fragsize) { kfree_skb(skb); - return -EINVAL; + return -EMSGSIZE; } if (datalen > cork->gso_size * UDP_MAX_SEGMENTS) { kfree_skb(skb); @@ -921,8 +1152,7 @@ static int udp_send_skb(struct sk_buff *skb, struct flowi4 *fl4, kfree_skb(skb); return -EINVAL; } - if (skb->ip_summed != CHECKSUM_PARTIAL || is_udplite || - dst_xfrm(skb_dst(skb))) { + if (is_udplite || dst_xfrm(skb_dst(skb))) { kfree_skb(skb); return -EIO; } @@ -932,8 +1162,10 @@ static int udp_send_skb(struct sk_buff *skb, struct flowi4 *fl4, skb_shinfo(skb)->gso_type = SKB_GSO_UDP_L4; skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(datalen, cork->gso_size); + + /* Don't checksum the payload, skb will get segmented */ + goto csum_partial; } - goto csum_partial; } if (is_udplite) /* UDP-Lite */ @@ -993,10 +1225,10 @@ int udp_push_pending_frames(struct sock *sk) out: up->len = 0; - up->pending = 0; + WRITE_ONCE(up->pending, 0); return err; } -EXPORT_SYMBOL(udp_push_pending_frames); +EXPORT_IPV6_MOD(udp_push_pending_frames); static int __udp_cmsg_send(struct cmsghdr *cmsg, u16 *gso_size) { @@ -1033,7 +1265,7 @@ int udp_cmsg_send(struct sock *sk, struct msghdr *msg, u16 *gso_size) return need_ip; } -EXPORT_SYMBOL_GPL(udp_cmsg_send); +EXPORT_IPV6_MOD_GPL(udp_cmsg_send); int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) { @@ -1048,13 +1280,14 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) int free = 0; int connected = 0; __be32 daddr, faddr, saddr; - u8 tos, scope; + u8 scope; __be16 dport; int err, is_udplite = IS_UDPLITE(sk); - int corkreq = READ_ONCE(up->corkflag) || msg->msg_flags&MSG_MORE; + int corkreq = udp_test_bit(CORK, sk) || msg->msg_flags & MSG_MORE; int (*getfrag)(void *, char *, int, int, int, struct sk_buff *); struct sk_buff *skb; struct ip_options_data opt_copy; + int uc_index; if (len > 0xFFFF) return -EMSGSIZE; @@ -1069,7 +1302,7 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) getfrag = is_udplite ? udplite_getfrag : ip_generic_getfrag; fl4 = &inet->cork.fl.u.ip4; - if (up->pending) { + if (READ_ONCE(up->pending)) { /* * There are pending frames. * The socket lock must be held while it's corked. @@ -1117,16 +1350,17 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) if (msg->msg_controllen) { err = udp_cmsg_send(sk, msg, &ipc.gso_size); - if (err > 0) + if (err > 0) { err = ip_cmsg_send(sk, msg, &ipc, sk->sk_family == AF_INET6); + connected = 0; + } if (unlikely(err < 0)) { kfree(ipc.opt); return err; } if (ipc.opt) free = 1; - connected = 0; } if (!ipc.opt) { struct ip_options_rcu *inet_opt; @@ -1143,7 +1377,9 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) if (cgroup_bpf_enabled(CGROUP_UDP4_SENDMSG) && !connected) { err = BPF_CGROUP_RUN_PROG_UDP4_SENDMSG_LOCK(sk, - (struct sockaddr *)usin, &ipc.addr); + (struct sockaddr *)usin, + &msg->msg_namelen, + &ipc.addr); if (err) goto out_free; if (usin) { @@ -1168,35 +1404,35 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) faddr = ipc.opt->opt.faddr; connected = 0; } - tos = get_rttos(&ipc, inet); scope = ip_sendmsg_scope(inet, &ipc, msg); if (scope == RT_SCOPE_LINK) connected = 0; + uc_index = READ_ONCE(inet->uc_index); if (ipv4_is_multicast(daddr)) { if (!ipc.oif || netif_index_is_l3_master(sock_net(sk), ipc.oif)) - ipc.oif = inet->mc_index; + ipc.oif = READ_ONCE(inet->mc_index); if (!saddr) - saddr = inet->mc_addr; + saddr = READ_ONCE(inet->mc_addr); connected = 0; } else if (!ipc.oif) { - ipc.oif = inet->uc_index; - } else if (ipv4_is_lbcast(daddr) && inet->uc_index) { + ipc.oif = uc_index; + } else if (ipv4_is_lbcast(daddr) && uc_index) { /* oif is set, packet is to local broadcast and * uc_index is set. oif is most likely set * by sk_bound_dev_if. If uc_index != oif check if the * oif is an L3 master and uc_index is an L3 slave. * If so, we want to allow the send using the uc_index. */ - if (ipc.oif != inet->uc_index && + if (ipc.oif != uc_index && ipc.oif == l3mdev_master_ifindex_by_index(sock_net(sk), - inet->uc_index)) { - ipc.oif = inet->uc_index; + uc_index)) { + ipc.oif = uc_index; } } if (connected) - rt = (struct rtable *)sk_dst_check(sk, 0); + rt = dst_rtable(sk_dst_check(sk, 0)); if (!rt) { struct net *net = sock_net(sk); @@ -1204,9 +1440,11 @@ int udp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) fl4 = &fl4_stack; - flowi4_init_output(fl4, ipc.oif, ipc.sockc.mark, tos, scope, + flowi4_init_output(fl4, ipc.oif, ipc.sockc.mark, + ipc.tos & INET_DSCP_MASK, scope, sk->sk_protocol, flow_flags, faddr, saddr, - dport, inet->inet_sport, sk->sk_uid); + dport, inet->inet_sport, + sk_uid(sk)); security_sk_classify_flow(sk, flowi4_to_flowi_common(fl4)); rt = ip_route_output_flow(net, fl4, sk); @@ -1265,7 +1503,7 @@ back_from_confirm: fl4->saddr = saddr; fl4->fl4_dport = dport; fl4->fl4_sport = inet->inet_sport; - up->pending = AF_INET; + WRITE_ONCE(up->pending, AF_INET); do_append_data: up->len += ulen; @@ -1277,7 +1515,7 @@ do_append_data: else if (!corkreq) err = udp_push_pending_frames(sk); else if (unlikely(skb_queue_empty(&sk->sk_write_queue))) - up->pending = 0; + WRITE_ONCE(up->pending, 0); release_sock(sk); out: @@ -1315,15 +1553,15 @@ void udp_splice_eof(struct socket *sock) struct sock *sk = sock->sk; struct udp_sock *up = udp_sk(sk); - if (!up->pending || READ_ONCE(up->corkflag)) + if (!READ_ONCE(up->pending) || udp_test_bit(CORK, sk)) return; lock_sock(sk); - if (up->pending && !READ_ONCE(up->corkflag)) + if (up->pending && !udp_test_bit(CORK, sk)) udp_push_pending_frames(sk); release_sock(sk); } -EXPORT_SYMBOL_GPL(udp_splice_eof); +EXPORT_IPV6_MOD_GPL(udp_splice_eof); #define UDP_SKB_IS_STATELESS 0x80000000 @@ -1388,12 +1626,12 @@ static bool udp_skb_has_head_state(struct sk_buff *skb) } /* fully reclaim rmem/fwd memory allocated for skb */ -static void udp_rmem_release(struct sock *sk, int size, int partial, - bool rx_queue_lock_held) +static void udp_rmem_release(struct sock *sk, unsigned int size, + int partial, bool rx_queue_lock_held) { struct udp_sock *up = udp_sk(sk); struct sk_buff_head *sk_queue; - int amt; + unsigned int amt; if (likely(partial)) { up->forward_deficit += size; @@ -1413,10 +1651,8 @@ static void udp_rmem_release(struct sock *sk, int size, int partial, if (!rx_queue_lock_held) spin_lock(&sk_queue->lock); - - sk->sk_forward_alloc += size; - amt = (sk->sk_forward_alloc - partial) & ~(PAGE_SIZE - 1); - sk->sk_forward_alloc -= amt; + amt = (size + sk->sk_forward_alloc - partial) & ~(PAGE_SIZE - 1); + sk_forward_alloc_add(sk, size - amt); if (amt) __sk_mem_reduce_allocated(sk, amt >> PAGE_SHIFT); @@ -1440,7 +1676,7 @@ void udp_skb_destructor(struct sock *sk, struct sk_buff *skb) prefetch(&skb->data); udp_rmem_release(sk, udp_skb_truesize(skb), 1, false); } -EXPORT_SYMBOL(udp_skb_destructor); +EXPORT_IPV6_MOD(udp_skb_destructor); /* as above, but the caller held the rx queue lock, too */ static void udp_skb_dtor_locked(struct sock *sk, struct sk_buff *skb) @@ -1449,31 +1685,6 @@ static void udp_skb_dtor_locked(struct sock *sk, struct sk_buff *skb) udp_rmem_release(sk, udp_skb_truesize(skb), 1, true); } -/* Idea of busylocks is to let producers grab an extra spinlock - * to relieve pressure on the receive_queue spinlock shared by consumer. - * Under flood, this means that only one producer can be in line - * trying to acquire the receive_queue spinlock. - * These busylock can be allocated on a per cpu manner, instead of a - * per socket one (that would consume a cache line per socket) - */ -static int udp_busylocks_log __read_mostly; -static spinlock_t *udp_busylocks __read_mostly; - -static spinlock_t *busylock_acquire(void *ptr) -{ - spinlock_t *busy; - - busy = udp_busylocks + hash_ptr(ptr, udp_busylocks_log); - spin_lock(busy); - return busy; -} - -static void busylock_release(spinlock_t *busy) -{ - if (busy) - spin_unlock(busy); -} - static int udp_rmem_schedule(struct sock *sk, int size) { int delta; @@ -1488,16 +1699,35 @@ static int udp_rmem_schedule(struct sock *sk, int size) int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb) { struct sk_buff_head *list = &sk->sk_receive_queue; - int rmem, err = -ENOMEM; - spinlock_t *busy = NULL; - int size; + struct udp_prod_queue *udp_prod_queue; + struct sk_buff *next, *to_drop = NULL; + struct llist_node *ll_list; + unsigned int rmem, rcvbuf; + int size, err = -ENOMEM; + int total_size = 0; + int q_size = 0; + int dropcount; + int nb = 0; - /* try to avoid the costly atomic add/sub pair when the receive - * queue is full; always allow at least a packet - */ rmem = atomic_read(&sk->sk_rmem_alloc); - if (rmem > sk->sk_rcvbuf) - goto drop; + rcvbuf = READ_ONCE(sk->sk_rcvbuf); + size = skb->truesize; + + udp_prod_queue = &udp_sk(sk)->udp_prod_queue[numa_node_id()]; + + rmem += atomic_read(&udp_prod_queue->rmem_alloc); + + /* Immediately drop when the receive queue is full. + * Cast to unsigned int performs the boundary check for INT_MAX. + */ + if (rmem + size > rcvbuf) { + if (rcvbuf > INT_MAX >> 1) + goto drop; + + /* Accept the packet if queue is empty. */ + if (rmem) + goto drop; + } /* Under mem pressure, it might be helpful to help udp_recvmsg() * having linear skbs : @@ -1505,53 +1735,83 @@ int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb) * - Less cache line misses at copyout() time * - Less work at consume_skb() (less alien page frag freeing) */ - if (rmem > (sk->sk_rcvbuf >> 1)) { + if (rmem > (rcvbuf >> 1)) { skb_condense(skb); - - busy = busylock_acquire(sk); + size = skb->truesize; } - size = skb->truesize; + udp_set_dev_scratch(skb); - /* we drop only if the receive buf is full and the receive - * queue contains some other skb - */ - rmem = atomic_add_return(size, &sk->sk_rmem_alloc); - if (rmem > (size + (unsigned int)sk->sk_rcvbuf)) - goto uncharge_drop; + atomic_add(size, &udp_prod_queue->rmem_alloc); + + if (!llist_add(&skb->ll_node, &udp_prod_queue->ll_root)) + return 0; + + dropcount = sock_flag(sk, SOCK_RXQ_OVFL) ? sk_drops_read(sk) : 0; spin_lock(&list->lock); - err = udp_rmem_schedule(sk, size); - if (err) { - spin_unlock(&list->lock); - goto uncharge_drop; - } - sk->sk_forward_alloc -= size; + ll_list = llist_del_all(&udp_prod_queue->ll_root); - /* no need to setup a destructor, we will explicitly release the - * forward allocated memory on dequeue - */ - sock_skb_set_dropcount(sk, skb); + ll_list = llist_reverse_order(ll_list); + + llist_for_each_entry_safe(skb, next, ll_list, ll_node) { + size = udp_skb_truesize(skb); + total_size += size; + err = udp_rmem_schedule(sk, size); + if (unlikely(err)) { + /* Free the skbs outside of locked section. */ + skb->next = to_drop; + to_drop = skb; + continue; + } + + q_size += size; + sk_forward_alloc_add(sk, -size); + + /* no need to setup a destructor, we will explicitly release the + * forward allocated memory on dequeue + */ + SOCK_SKB_CB(skb)->dropcount = dropcount; + nb++; + __skb_queue_tail(list, skb); + } + + atomic_add(q_size, &sk->sk_rmem_alloc); - __skb_queue_tail(list, skb); spin_unlock(&list->lock); - if (!sock_flag(sk, SOCK_DEAD)) - INDIRECT_CALL_1(sk->sk_data_ready, sock_def_readable, sk); + if (!sock_flag(sk, SOCK_DEAD)) { + /* Multiple threads might be blocked in recvmsg(), + * using prepare_to_wait_exclusive(). + */ + while (nb) { + INDIRECT_CALL_1(sk->sk_data_ready, + sock_def_readable, sk); + nb--; + } + } - busylock_release(busy); - return 0; + if (unlikely(to_drop)) { + for (nb = 0; to_drop != NULL; nb++) { + skb = to_drop; + to_drop = skb->next; + skb_mark_not_on_list(skb); + /* TODO: update SNMP values. */ + sk_skb_reason_drop(sk, skb, SKB_DROP_REASON_PROTO_MEM); + } + numa_drop_add(&udp_sk(sk)->drop_counters, nb); + } -uncharge_drop: - atomic_sub(skb->truesize, &sk->sk_rmem_alloc); + atomic_sub(total_size, &udp_prod_queue->rmem_alloc); + + return 0; drop: - atomic_inc(&sk->sk_drops); - busylock_release(busy); + udp_drops_inc(sk); return err; } -EXPORT_SYMBOL_GPL(__udp_enqueue_schedule_skb); +EXPORT_IPV6_MOD_GPL(__udp_enqueue_schedule_skb); void udp_destruct_common(struct sock *sk) { @@ -1566,8 +1826,9 @@ void udp_destruct_common(struct sock *sk) kfree_skb(skb); } udp_rmem_release(sk, total, 0, true); + kfree(up->udp_prod_queue); } -EXPORT_SYMBOL_GPL(udp_destruct_common); +EXPORT_IPV6_MOD_GPL(udp_destruct_common); static void udp_destruct_sock(struct sock *sk) { @@ -1577,19 +1838,21 @@ static void udp_destruct_sock(struct sock *sk) int udp_init_sock(struct sock *sk) { - udp_lib_init_sock(sk); + int res = udp_lib_init_sock(sk); + sk->sk_destruct = udp_destruct_sock; set_bit(SOCK_SUPPORT_ZC, &sk->sk_socket->flags); - return 0; + return res; } void skb_consume_udp(struct sock *sk, struct sk_buff *skb, int len) { - if (unlikely(READ_ONCE(sk->sk_peek_off) >= 0)) { - bool slow = lock_sock_fast(sk); - + if (unlikely(READ_ONCE(udp_sk(sk)->peeking_with_offset))) sk_peek_offset_bwd(sk, len); - unlock_sock_fast(sk, slow); + + if (!skb_shared(skb)) { + skb_attempt_defer_free(skb); + return; } if (!skb_unref(skb)) @@ -1602,11 +1865,11 @@ void skb_consume_udp(struct sock *sk, struct sk_buff *skb, int len) skb_release_head_state(skb); __consume_stateless_skb(skb); } -EXPORT_SYMBOL_GPL(skb_consume_udp); +EXPORT_IPV6_MOD_GPL(skb_consume_udp); static struct sk_buff *__first_packet_length(struct sock *sk, struct sk_buff_head *rcvq, - int *total) + unsigned int *total) { struct sk_buff *skb; @@ -1616,10 +1879,10 @@ static struct sk_buff *__first_packet_length(struct sock *sk, IS_UDPLITE(sk)); __UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, IS_UDPLITE(sk)); - atomic_inc(&sk->sk_drops); + udp_drops_inc(sk); __skb_unlink(skb, rcvq); *total += skb->truesize; - kfree_skb(skb); + kfree_skb_reason(skb, SKB_DROP_REASON_UDP_CSUM); } else { udp_skb_csum_unnecessary_set(skb); break; @@ -1639,8 +1902,8 @@ static int first_packet_length(struct sock *sk) { struct sk_buff_head *rcvq = &udp_sk(sk)->reader_queue; struct sk_buff_head *sk_queue = &sk->sk_receive_queue; + unsigned int total = 0; struct sk_buff *skb; - int total = 0; int res; spin_lock_bh(&rcvq->lock); @@ -1684,7 +1947,7 @@ int udp_ioctl(struct sock *sk, int cmd, int *karg) return 0; } -EXPORT_SYMBOL(udp_ioctl); +EXPORT_IPV6_MOD(udp_ioctl); struct sk_buff *__skb_recv_udp(struct sock *sk, unsigned int flags, int *off, int *err) @@ -1707,8 +1970,8 @@ struct sk_buff *__skb_recv_udp(struct sock *sk, unsigned int flags, error = -EAGAIN; do { spin_lock_bh(&queue->lock); - skb = __skb_try_recv_from_queue(sk, queue, flags, off, - err, &last); + skb = __skb_try_recv_from_queue(queue, flags, off, err, + &last); if (skb) { if (!(flags & MSG_PEEK)) udp_skb_destructor(sk, skb); @@ -1729,8 +1992,8 @@ struct sk_buff *__skb_recv_udp(struct sock *sk, unsigned int flags, spin_lock(&sk_queue->lock); skb_queue_splice_tail_init(sk_queue, queue); - skb = __skb_try_recv_from_queue(sk, queue, flags, off, - err, &last); + skb = __skb_try_recv_from_queue(queue, flags, off, err, + &last); if (skb && !(flags & MSG_PEEK)) udp_skb_dtor_locked(sk, skb); spin_unlock(&sk_queue->lock); @@ -1772,15 +2035,15 @@ try_again: __UDP_INC_STATS(net, UDP_MIB_CSUMERRORS, is_udplite); __UDP_INC_STATS(net, UDP_MIB_INERRORS, is_udplite); - atomic_inc(&sk->sk_drops); - kfree_skb(skb); + udp_drops_inc(sk); + kfree_skb_reason(skb, SKB_DROP_REASON_UDP_CSUM); goto try_again; } WARN_ON_ONCE(!skb_set_owner_sk_safe(skb, sk)); return recv_actor(sk, skb); } -EXPORT_SYMBOL(udp_read_skb); +EXPORT_IPV6_MOD(udp_read_skb); /* * This should be easy, if there is something there we @@ -1842,7 +2105,7 @@ try_again: if (unlikely(err)) { if (!peeking) { - atomic_inc(&sk->sk_drops); + udp_drops_inc(sk); UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite); } @@ -1865,10 +2128,11 @@ try_again: *addr_len = sizeof(*sin); BPF_CGROUP_RUN_PROG_UDP4_RECVMSG_LOCK(sk, - (struct sockaddr *)sin); + (struct sockaddr *)sin, + addr_len); } - if (udp_sk(sk)->gro_enabled) + if (udp_test_bit(GRO_ENABLED, sk)) udp_cmsg_recv(msg, sk, skb); if (inet_cmsg_flags(inet)) @@ -1887,7 +2151,7 @@ csum_copy_err: UDP_INC_STATS(sock_net(sk), UDP_MIB_CSUMERRORS, is_udplite); UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite); } - kfree_skb(skb); + kfree_skb_reason(skb, SKB_DROP_REASON_UDP_CSUM); /* starting over for a new packet, but check if we need to yield */ cond_resched(); @@ -1895,7 +2159,8 @@ csum_copy_err: goto try_again; } -int udp_pre_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int udp_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { /* This check is replicated from __ip4_datagram_connect() and * intended to prevent BPF program called below from accessing bytes @@ -1904,9 +2169,22 @@ int udp_pre_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) if (addr_len < sizeof(struct sockaddr_in)) return -EINVAL; - return BPF_CGROUP_RUN_PROG_INET4_CONNECT_LOCK(sk, uaddr); + return BPF_CGROUP_RUN_PROG_INET4_CONNECT_LOCK(sk, uaddr, &addr_len); +} +EXPORT_IPV6_MOD(udp_pre_connect); + +static int udp_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) +{ + int res; + + lock_sock(sk); + res = __ip4_datagram_connect(sk, uaddr, addr_len); + if (!res) + udp4_hash4(sk); + release_sock(sk); + return res; } -EXPORT_SYMBOL(udp_pre_connect); int __udp_disconnect(struct sock *sk, int flags) { @@ -1943,7 +2221,7 @@ int udp_disconnect(struct sock *sk, int flags) release_sock(sk); return 0; } -EXPORT_SYMBOL(udp_disconnect); +EXPORT_IPV6_MOD(udp_disconnect); void udp_lib_unhash(struct sock *sk) { @@ -1951,6 +2229,7 @@ void udp_lib_unhash(struct sock *sk) struct udp_table *udptable = udp_get_table_prot(sk); struct udp_hslot *hslot, *hslot2; + sock_rps_delete_flow(sk); hslot = udp_hashslot(udptable, sock_net(sk), udp_sk(sk)->udp_port_hash); hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); @@ -1967,29 +2246,31 @@ void udp_lib_unhash(struct sock *sk) hlist_del_init_rcu(&udp_sk(sk)->udp_portaddr_node); hslot2->count--; spin_unlock(&hslot2->lock); + + udp_unhash4(udptable, sk); } spin_unlock_bh(&hslot->lock); } } -EXPORT_SYMBOL(udp_lib_unhash); +EXPORT_IPV6_MOD(udp_lib_unhash); /* * inet_rcv_saddr was changed, we must rehash secondary hash */ -void udp_lib_rehash(struct sock *sk, u16 newhash) +void udp_lib_rehash(struct sock *sk, u16 newhash, u16 newhash4) { if (sk_hashed(sk)) { struct udp_table *udptable = udp_get_table_prot(sk); struct udp_hslot *hslot, *hslot2, *nhslot2; + hslot = udp_hashslot(udptable, sock_net(sk), + udp_sk(sk)->udp_port_hash); hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); nhslot2 = udp_hashslot2(udptable, newhash); udp_sk(sk)->udp_portaddr_hash = newhash; if (hslot2 != nhslot2 || rcu_access_pointer(sk->sk_reuseport_cb)) { - hslot = udp_hashslot(udptable, sock_net(sk), - udp_sk(sk)->udp_port_hash); /* we must lock primary chain too */ spin_lock_bh(&hslot->lock); if (rcu_access_pointer(sk->sk_reuseport_cb)) @@ -2010,16 +2291,43 @@ void udp_lib_rehash(struct sock *sk, u16 newhash) spin_unlock_bh(&hslot->lock); } + + /* Now process hash4 if necessary: + * (1) update hslot4; + * (2) update hslot2->hash4_cnt. + * Note that hslot2/hslot4 should be checked separately, as + * either of them may change with the other unchanged. + */ + if (udp_hashed4(sk)) { + spin_lock_bh(&hslot->lock); + + udp_rehash4(udptable, sk, newhash4); + if (hslot2 != nhslot2) { + spin_lock(&hslot2->lock); + udp_hash4_dec(hslot2); + spin_unlock(&hslot2->lock); + + spin_lock(&nhslot2->lock); + udp_hash4_inc(nhslot2); + spin_unlock(&nhslot2->lock); + } + + spin_unlock_bh(&hslot->lock); + } } } -EXPORT_SYMBOL(udp_lib_rehash); +EXPORT_IPV6_MOD(udp_lib_rehash); void udp_v4_rehash(struct sock *sk) { u16 new_hash = ipv4_portaddr_hash(sock_net(sk), inet_sk(sk)->inet_rcv_saddr, inet_sk(sk)->inet_num); - udp_lib_rehash(sk, new_hash); + u16 new_hash4 = udp_ehashfn(sock_net(sk), + sk->sk_rcv_saddr, sk->sk_num, + sk->sk_daddr, sk->sk_dport); + + udp_lib_rehash(sk, new_hash, new_hash4); } static int __udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb) @@ -2050,8 +2358,8 @@ static int __udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb) drop_reason = SKB_DROP_REASON_PROTO_MEM; } UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite); - kfree_skb_reason(skb, drop_reason); - trace_udp_fail_queue_rcv_skb(rc, sk); + trace_udp_fail_queue_rcv_skb(rc, sk, skb); + sk_skb_reason_drop(sk, skb, drop_reason); return -1; } @@ -2068,7 +2376,7 @@ static int __udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb) */ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) { - int drop_reason = SKB_DROP_REASON_NOT_SPECIFIED; + enum skb_drop_reason drop_reason = SKB_DROP_REASON_NOT_SPECIFIED; struct udp_sock *up = udp_sk(sk); int is_udplite = IS_UDPLITE(sk); @@ -2081,7 +2389,8 @@ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) } nf_reset_ct(skb); - if (static_branch_unlikely(&udp_encap_needed_key) && up->encap_type) { + if (static_branch_unlikely(&udp_encap_needed_key) && + READ_ONCE(up->encap_type)) { int (*encap_rcv)(struct sock *sk, struct sk_buff *skb); /* @@ -2119,7 +2428,8 @@ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) /* * UDP-Lite specific tests, ignored on UDP sockets */ - if ((up->pcflag & UDPLITE_RECV_CC) && UDP_SKB_CB(skb)->partial_cov) { + if (udp_test_bit(UDPLITE_RECV_CC, sk) && UDP_SKB_CB(skb)->partial_cov) { + u16 pcrlen = READ_ONCE(up->pcrlen); /* * MIB statistics other than incrementing the error count are @@ -2132,7 +2442,7 @@ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) * delivery of packets with coverage values less than a value * provided by the application." */ - if (up->pcrlen == 0) { /* full coverage was set */ + if (pcrlen == 0) { /* full coverage was set */ net_dbg_ratelimited("UDPLite: partial coverage %d while full coverage %d requested\n", UDP_SKB_CB(skb)->cscov, skb->len); goto drop; @@ -2143,9 +2453,9 @@ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) * that it wants x while sender emits packets of smaller size y. * Therefore the above ...()->partial_cov statement is essential. */ - if (UDP_SKB_CB(skb)->cscov < up->pcrlen) { + if (UDP_SKB_CB(skb)->cscov < pcrlen) { net_dbg_ratelimited("UDPLite: coverage %d too small, need min %d\n", - UDP_SKB_CB(skb)->cscov, up->pcrlen); + UDP_SKB_CB(skb)->cscov, pcrlen); goto drop; } } @@ -2155,14 +2465,12 @@ static int udp_queue_rcv_one_skb(struct sock *sk, struct sk_buff *skb) udp_lib_checksum_complete(skb)) goto csum_error; - if (sk_filter_trim_cap(sk, skb, sizeof(struct udphdr))) { - drop_reason = SKB_DROP_REASON_SOCKET_FILTER; + if (sk_filter_trim_cap(sk, skb, sizeof(struct udphdr), &drop_reason)) goto drop; - } udp_csum_pull_header(skb); - ipv4_pktinfo_prepare(sk, skb); + ipv4_pktinfo_prepare(sk, skb, true); return __udp_queue_rcv_skb(sk, skb); csum_error: @@ -2170,8 +2478,8 @@ csum_error: __UDP_INC_STATS(sock_net(sk), UDP_MIB_CSUMERRORS, is_udplite); drop: __UDP_INC_STATS(sock_net(sk), UDP_MIB_INERRORS, is_udplite); - atomic_inc(&sk->sk_drops); - kfree_skb_reason(skb, drop_reason); + udp_drops_inc(sk); + sk_skb_reason_drop(sk, skb, drop_reason); return -1; } @@ -2205,13 +2513,13 @@ bool udp_sk_rx_dst_set(struct sock *sk, struct dst_entry *dst) struct dst_entry *old; if (dst_hold_safe(dst)) { - old = xchg((__force struct dst_entry **)&sk->sk_rx_dst, dst); + old = unrcu_pointer(xchg(&sk->sk_rx_dst, RCU_INITIALIZER(dst))); dst_release(old); return old != dst; } return false; } -EXPORT_SYMBOL(udp_sk_rx_dst_set); +EXPORT_IPV6_MOD(udp_sk_rx_dst_set); /* * Multicasts and broadcasts go to each listener. @@ -2239,7 +2547,7 @@ static int __udp4_lib_mcast_deliver(struct net *net, struct sk_buff *skb, udptable->mask; hash2 = ipv4_portaddr_hash(net, daddr, hnum) & udptable->mask; start_lookup: - hslot = &udptable->hash2[hash2]; + hslot = &udptable->hash2[hash2].hslot; offset = offsetof(typeof(*sk), __sk_common.skc_portaddr_node); } @@ -2255,7 +2563,7 @@ start_lookup: nskb = skb_clone(skb, GFP_ATOMIC); if (unlikely(!nskb)) { - atomic_inc(&sk->sk_drops); + udp_drops_inc(sk); __UDP_INC_STATS(net, UDP_MIB_RCVBUFERRORS, IS_UDPLITE(sk)); __UDP_INC_STATS(net, UDP_MIB_INERRORS, @@ -2330,7 +2638,7 @@ static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh, return 0; } -/* wrapper for udp_queue_rcv_skb tacking care of csum conversion and +/* wrapper for udp_queue_rcv_skb taking care of csum conversion and * return code conversion for ip layer consumption */ static int udp_unicast_rcv_skb(struct sock *sk, struct sk_buff *skb, @@ -2358,7 +2666,7 @@ static int udp_unicast_rcv_skb(struct sock *sk, struct sk_buff *skb, int __udp4_lib_rcv(struct sk_buff *skb, struct udp_table *udptable, int proto) { - struct sock *sk; + struct sock *sk = NULL; struct udphdr *uh; unsigned short ulen; struct rtable *rt = skb_rtable(skb); @@ -2435,7 +2743,7 @@ no_sk: * Hmm. We got an UDP packet to a port to which we * don't wanna listen. Ignore it. */ - kfree_skb_reason(skb, drop_reason); + sk_skb_reason_drop(sk, skb, drop_reason); return 0; short_packet: @@ -2460,7 +2768,7 @@ csum_error: __UDP_INC_STATS(net, UDP_MIB_CSUMERRORS, proto == IPPROTO_UDPLITE); drop: __UDP_INC_STATS(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE); - kfree_skb_reason(skb, drop_reason); + sk_skb_reason_drop(sk, skb, drop_reason); return 0; } @@ -2510,14 +2818,13 @@ static struct sock *__udp4_lib_demux_lookup(struct net *net, struct udp_table *udptable = net->ipv4.udp_table; INET_ADDR_COOKIE(acookie, rmt_addr, loc_addr); unsigned short hnum = ntohs(loc_port); - unsigned int hash2, slot2; struct udp_hslot *hslot2; + unsigned int hash2; __portpair ports; struct sock *sk; hash2 = ipv4_portaddr_hash(net, loc_addr, hnum); - slot2 = hash2 & udptable->mask; - hslot2 = &udptable->hash2[slot2]; + hslot2 = udp_hashslot2(udptable, hash2); ports = INET_COMBINED_PORTS(rmt_port, hnum); udp_portaddr_for_each_entry_rcu(sk, &hslot2->head) { @@ -2529,7 +2836,7 @@ static struct sock *__udp4_lib_demux_lookup(struct net *net, return NULL; } -int udp_v4_early_demux(struct sk_buff *skb) +enum skb_drop_reason udp_v4_early_demux(struct sk_buff *skb) { struct net *net = dev_net(skb->dev); struct in_device *in_dev = NULL; @@ -2543,7 +2850,7 @@ int udp_v4_early_demux(struct sk_buff *skb) /* validate the packet */ if (!pskb_may_pull(skb, skb_transport_offset(skb) + sizeof(struct udphdr))) - return 0; + return SKB_NOT_DROPPED_YET; iph = ip_hdr(skb); uh = udp_hdr(skb); @@ -2552,12 +2859,12 @@ int udp_v4_early_demux(struct sk_buff *skb) in_dev = __in_dev_get_rcu(skb->dev); if (!in_dev) - return 0; + return SKB_NOT_DROPPED_YET; ours = ip_check_mc_rcu(in_dev, iph->daddr, iph->saddr, iph->protocol); if (!ours) - return 0; + return SKB_NOT_DROPPED_YET; sk = __udp4_lib_mcast_demux_lookup(net, uh->dest, iph->daddr, uh->source, iph->saddr, @@ -2567,11 +2874,12 @@ int udp_v4_early_demux(struct sk_buff *skb) uh->source, iph->saddr, dif, sdif); } - if (!sk || !refcount_inc_not_zero(&sk->sk_refcnt)) - return 0; + if (!sk) + return SKB_NOT_DROPPED_YET; skb->sk = sk; - skb->destructor = sock_efree; + DEBUG_NET_WARN_ON_ONCE(sk_is_refcounted(sk)); + skb->destructor = sock_pfree; dst = rcu_dereference(sk->sk_rx_dst); if (dst) @@ -2591,10 +2899,10 @@ int udp_v4_early_demux(struct sk_buff *skb) if (!inet_sk(sk)->inet_daddr && in_dev) return ip_mc_validate_source(skb, iph->daddr, iph->saddr, - iph->tos & IPTOS_RT_MASK, + ip4h_dscp(iph), skb->dev, in_dev, &itag); } - return 0; + return SKB_NOT_DROPPED_YET; } int udp_rcv(struct sk_buff *skb) @@ -2618,11 +2926,44 @@ void udp_destroy_sock(struct sock *sk) if (encap_destroy) encap_destroy(sk); } - if (up->encap_enabled) + if (udp_test_bit(ENCAP_ENABLED, sk)) { static_branch_dec(&udp_encap_needed_key); + udp_tunnel_cleanup_gro(sk); + } } } +typedef struct sk_buff *(*udp_gro_receive_t)(struct sock *sk, + struct list_head *head, + struct sk_buff *skb); + +static void set_xfrm_gro_udp_encap_rcv(__u16 encap_type, unsigned short family, + struct sock *sk) +{ +#ifdef CONFIG_XFRM + udp_gro_receive_t new_gro_receive; + + if (udp_test_bit(GRO_ENABLED, sk) && encap_type == UDP_ENCAP_ESPINUDP) { + if (IS_ENABLED(CONFIG_IPV6) && family == AF_INET6) + new_gro_receive = ipv6_stub->xfrm6_gro_udp_encap_rcv; + else + new_gro_receive = xfrm4_gro_udp_encap_rcv; + + if (udp_sk(sk)->gro_receive != new_gro_receive) { + /* + * With IPV6_ADDRFORM the gro callback could change + * after being set, unregister the old one, if valid. + */ + if (udp_sk(sk)->gro_receive) + udp_tunnel_update_gro_rcv(sk, false); + + WRITE_ONCE(udp_sk(sk)->gro_receive, new_gro_receive); + udp_tunnel_update_gro_rcv(sk, true); + } + } +#endif +} + /* * Socket option code for UDP */ @@ -2658,9 +2999,9 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, switch (optname) { case UDP_CORK: if (val != 0) { - WRITE_ONCE(up->corkflag, 1); + udp_set_bit(CORK, sk); } else { - WRITE_ONCE(up->corkflag, 0); + udp_clear_bit(CORK, sk); lock_sock(sk); push_pending_frames(sk); release_sock(sk); @@ -2668,37 +3009,39 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, break; case UDP_ENCAP: + sockopt_lock_sock(sk); switch (val) { case 0: #ifdef CONFIG_XFRM case UDP_ENCAP_ESPINUDP: - case UDP_ENCAP_ESPINUDP_NON_IKE: + set_xfrm_gro_udp_encap_rcv(val, sk->sk_family, sk); #if IS_ENABLED(CONFIG_IPV6) if (sk->sk_family == AF_INET6) - up->encap_rcv = ipv6_stub->xfrm6_udp_encap_rcv; + WRITE_ONCE(up->encap_rcv, + ipv6_stub->xfrm6_udp_encap_rcv); else #endif - up->encap_rcv = xfrm4_udp_encap_rcv; + WRITE_ONCE(up->encap_rcv, + xfrm4_udp_encap_rcv); #endif fallthrough; case UDP_ENCAP_L2TPINUDP: - up->encap_type = val; - lock_sock(sk); - udp_tunnel_encap_enable(sk->sk_socket); - release_sock(sk); + WRITE_ONCE(up->encap_type, val); + udp_tunnel_encap_enable(sk); break; default: err = -ENOPROTOOPT; break; } + sockopt_release_sock(sk); break; case UDP_NO_CHECK6_TX: - up->no_check6_tx = valbool; + udp_set_no_check6_tx(sk, valbool); break; case UDP_NO_CHECK6_RX: - up->no_check6_rx = valbool; + udp_set_no_check6_rx(sk, valbool); break; case UDP_SEGMENT: @@ -2708,14 +3051,14 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, break; case UDP_GRO: - lock_sock(sk); - + sockopt_lock_sock(sk); /* when enabling GRO, accept the related GSO packet type */ if (valbool) - udp_tunnel_encap_enable(sk->sk_socket); - up->gro_enabled = valbool; - up->accept_udp_l4 = valbool; - release_sock(sk); + udp_tunnel_encap_enable(sk); + udp_assign_bit(GRO_ENABLED, sk, valbool); + udp_assign_bit(ACCEPT_L4, sk, valbool); + set_xfrm_gro_udp_encap_rcv(up->encap_type, sk->sk_family, sk); + sockopt_release_sock(sk); break; /* @@ -2730,8 +3073,8 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, val = 8; else if (val > USHRT_MAX) val = USHRT_MAX; - up->pcslen = val; - up->pcflag |= UDPLITE_SEND_CC; + WRITE_ONCE(up->pcslen, val); + udp_set_bit(UDPLITE_SEND_CC, sk); break; /* The receiver specifies a minimum checksum coverage value. To make @@ -2744,8 +3087,8 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, val = 8; else if (val > USHRT_MAX) val = USHRT_MAX; - up->pcrlen = val; - up->pcflag |= UDPLITE_RECV_CC; + WRITE_ONCE(up->pcrlen, val); + udp_set_bit(UDPLITE_RECV_CC, sk); break; default: @@ -2755,7 +3098,7 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname, return err; } -EXPORT_SYMBOL(udp_lib_setsockopt); +EXPORT_IPV6_MOD(udp_lib_setsockopt); int udp_setsockopt(struct sock *sk, int level, int optname, sockptr_t optval, unsigned int optlen) @@ -2776,26 +3119,26 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname, if (get_user(len, optlen)) return -EFAULT; - len = min_t(unsigned int, len, sizeof(int)); - if (len < 0) return -EINVAL; + len = min_t(unsigned int, len, sizeof(int)); + switch (optname) { case UDP_CORK: - val = READ_ONCE(up->corkflag); + val = udp_test_bit(CORK, sk); break; case UDP_ENCAP: - val = up->encap_type; + val = READ_ONCE(up->encap_type); break; case UDP_NO_CHECK6_TX: - val = up->no_check6_tx; + val = udp_get_no_check6_tx(sk); break; case UDP_NO_CHECK6_RX: - val = up->no_check6_rx; + val = udp_get_no_check6_rx(sk); break; case UDP_SEGMENT: @@ -2803,17 +3146,17 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname, break; case UDP_GRO: - val = up->gro_enabled; + val = udp_test_bit(GRO_ENABLED, sk); break; /* The following two cannot be changed on UDP sockets, the return is * always 0 (which corresponds to the full checksum coverage of UDP). */ case UDPLITE_SEND_CSCOV: - val = up->pcslen; + val = READ_ONCE(up->pcslen); break; case UDPLITE_RECV_CSCOV: - val = up->pcrlen; + val = READ_ONCE(up->pcrlen); break; default: @@ -2826,7 +3169,7 @@ int udp_lib_getsockopt(struct sock *sk, int level, int optname, return -EFAULT; return 0; } -EXPORT_SYMBOL(udp_lib_getsockopt); +EXPORT_IPV6_MOD(udp_lib_getsockopt); int udp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen) @@ -2868,7 +3211,7 @@ __poll_t udp_poll(struct file *file, struct socket *sock, poll_table *wait) return mask; } -EXPORT_SYMBOL(udp_poll); +EXPORT_IPV6_MOD(udp_poll); int udp_abort(struct sock *sk, int err) { @@ -2891,14 +3234,14 @@ out: return 0; } -EXPORT_SYMBOL_GPL(udp_abort); +EXPORT_IPV6_MOD_GPL(udp_abort); struct proto udp_prot = { .name = "UDP", .owner = THIS_MODULE, .close = udp_lib_close, .pre_connect = udp_pre_connect, - .connect = ip4_datagram_connect, + .connect = udp_connect, .disconnect = udp_disconnect, .ioctl = udp_ioctl, .init = udp_init_sock, @@ -2917,7 +3260,7 @@ struct proto udp_prot = { #ifdef CONFIG_BPF_SYSCALL .psock_update_sk_prot = udp_bpf_update_proto, #endif - .memory_allocated = &udp_memory_allocated, + .memory_allocated = &net_aligned_data.udp_memory_allocated, .per_cpu_fw_alloc = &udp_memory_per_cpu_fw_alloc, .sysctl_mem = sysctl_udp_mem, @@ -3025,7 +3368,7 @@ void *udp_seq_start(struct seq_file *seq, loff_t *pos) return *pos ? udp_get_idx(seq, *pos-1) : SEQ_START_TOKEN; } -EXPORT_SYMBOL(udp_seq_start); +EXPORT_IPV6_MOD(udp_seq_start); void *udp_seq_next(struct seq_file *seq, void *v, loff_t *pos) { @@ -3039,7 +3382,7 @@ void *udp_seq_next(struct seq_file *seq, void *v, loff_t *pos) ++*pos; return sk; } -EXPORT_SYMBOL(udp_seq_next); +EXPORT_IPV6_MOD(udp_seq_next); void udp_seq_stop(struct seq_file *seq, void *v) { @@ -3051,7 +3394,7 @@ void udp_seq_stop(struct seq_file *seq, void *v) if (state->bucket <= udptable->mask) spin_unlock_bh(&udptable->hash[state->bucket].lock); } -EXPORT_SYMBOL(udp_seq_stop); +EXPORT_IPV6_MOD(udp_seq_stop); /* ------------------------------------------------------------------------ */ static void udp4_format_sock(struct sock *sp, struct seq_file *f, @@ -3069,10 +3412,10 @@ static void udp4_format_sock(struct sock *sp, struct seq_file *f, sk_wmem_alloc_get(sp), udp_rqueue_get(sp), 0, 0L, 0, - from_kuid_munged(seq_user_ns(f), sock_i_uid(sp)), + from_kuid_munged(seq_user_ns(f), sk_uid(sp)), 0, sock_i_ino(sp), refcount_read(&sp->sk_refcnt), sp, - atomic_read(&sp->sk_drops)); + sk_drops_read(sp)); } int udp4_seq_show(struct seq_file *seq, void *v) @@ -3099,33 +3442,56 @@ struct bpf_iter__udp { int bucket __aligned(8); }; +union bpf_udp_iter_batch_item { + struct sock *sk; + __u64 cookie; +}; + struct bpf_udp_iter_state { struct udp_iter_state state; unsigned int cur_sk; unsigned int end_sk; unsigned int max_sk; - int offset; - struct sock **batch; - bool st_bucket_done; + union bpf_udp_iter_batch_item *batch; }; static int bpf_iter_udp_realloc_batch(struct bpf_udp_iter_state *iter, - unsigned int new_batch_sz); + unsigned int new_batch_sz, gfp_t flags); +static struct sock *bpf_iter_udp_resume(struct sock *first_sk, + union bpf_udp_iter_batch_item *cookies, + int n_cookies) +{ + struct sock *sk = NULL; + int i; + + for (i = 0; i < n_cookies; i++) { + sk = first_sk; + udp_portaddr_for_each_entry_from(sk) + if (cookies[i].cookie == atomic64_read(&sk->sk_cookie)) + goto done; + } +done: + return sk; +} + static struct sock *bpf_iter_udp_batch(struct seq_file *seq) { struct bpf_udp_iter_state *iter = seq->private; struct udp_iter_state *state = &iter->state; + unsigned int find_cookie, end_cookie; struct net *net = seq_file_net(seq); struct udp_table *udptable; unsigned int batch_sks = 0; - bool resized = false; + int resume_bucket; + int resizes = 0; struct sock *sk; + int err = 0; + + resume_bucket = state->bucket; /* The current batch is done, so advance the bucket. */ - if (iter->st_bucket_done) { + if (iter->cur_sk == iter->end_sk) state->bucket++; - iter->offset = 0; - } udptable = udp_get_table_seq(seq, net); @@ -3137,66 +3503,89 @@ again: * before releasing the bucket lock. This allows BPF programs that are * called in seq_show to acquire the bucket lock if needed. */ + find_cookie = iter->cur_sk; + end_cookie = iter->end_sk; iter->cur_sk = 0; iter->end_sk = 0; - iter->st_bucket_done = false; batch_sks = 0; for (; state->bucket <= udptable->mask; state->bucket++) { - struct udp_hslot *hslot2 = &udptable->hash2[state->bucket]; + struct udp_hslot *hslot2 = &udptable->hash2[state->bucket].hslot; - if (hlist_empty(&hslot2->head)) { - iter->offset = 0; - continue; - } + if (hlist_empty(&hslot2->head)) + goto next_bucket; spin_lock_bh(&hslot2->lock); - udp_portaddr_for_each_entry(sk, &hslot2->head) { + sk = hlist_entry_safe(hslot2->head.first, struct sock, + __sk_common.skc_portaddr_node); + /* Resume from the first (in iteration order) unseen socket from + * the last batch that still exists in resume_bucket. Most of + * the time this will just be where the last iteration left off + * in resume_bucket unless that socket disappeared between + * reads. + */ + if (state->bucket == resume_bucket) + sk = bpf_iter_udp_resume(sk, &iter->batch[find_cookie], + end_cookie - find_cookie); +fill_batch: + udp_portaddr_for_each_entry_from(sk) { if (seq_sk_match(seq, sk)) { - /* Resume from the last iterated socket at the - * offset in the bucket before iterator was stopped. - */ - if (iter->offset) { - --iter->offset; - continue; - } if (iter->end_sk < iter->max_sk) { sock_hold(sk); - iter->batch[iter->end_sk++] = sk; + iter->batch[iter->end_sk++].sk = sk; } batch_sks++; } } + + /* Allocate a larger batch and try again. */ + if (unlikely(resizes <= 1 && iter->end_sk && + iter->end_sk != batch_sks)) { + resizes++; + + /* First, try with GFP_USER to maximize the chances of + * grabbing more memory. + */ + if (resizes == 1) { + spin_unlock_bh(&hslot2->lock); + err = bpf_iter_udp_realloc_batch(iter, + batch_sks * 3 / 2, + GFP_USER); + if (err) + return ERR_PTR(err); + /* Start over. */ + goto again; + } + + /* Next, hold onto the lock, so the bucket doesn't + * change while we get the rest of the sockets. + */ + err = bpf_iter_udp_realloc_batch(iter, batch_sks, + GFP_NOWAIT); + if (err) { + spin_unlock_bh(&hslot2->lock); + return ERR_PTR(err); + } + + /* Pick up where we left off. */ + sk = iter->batch[iter->end_sk - 1].sk; + sk = hlist_entry_safe(sk->__sk_common.skc_portaddr_node.next, + struct sock, + __sk_common.skc_portaddr_node); + batch_sks = iter->end_sk; + goto fill_batch; + } + spin_unlock_bh(&hslot2->lock); if (iter->end_sk) break; - - /* Reset the current bucket's offset before moving to the next bucket. */ - iter->offset = 0; +next_bucket: + resizes = 0; } - /* All done: no batch made. */ - if (!iter->end_sk) - return NULL; - - if (iter->end_sk == batch_sks) { - /* Batching is done for the current bucket; return the first - * socket to be iterated from the batch. - */ - iter->st_bucket_done = true; - goto done; - } - if (!resized && !bpf_iter_udp_realloc_batch(iter, batch_sks * 3 / 2)) { - resized = true; - /* After allocating a larger batch, retry one more time to grab - * the whole bucket. - */ - state->bucket--; - goto again; - } -done: - return iter->batch[0]; + WARN_ON_ONCE(iter->end_sk != batch_sks); + return iter->end_sk ? iter->batch[0].sk : NULL; } static void *bpf_iter_udp_seq_next(struct seq_file *seq, void *v, loff_t *pos) @@ -3207,16 +3596,14 @@ static void *bpf_iter_udp_seq_next(struct seq_file *seq, void *v, loff_t *pos) /* Whenever seq_next() is called, the iter->cur_sk is * done with seq_show(), so unref the iter->cur_sk. */ - if (iter->cur_sk < iter->end_sk) { - sock_put(iter->batch[iter->cur_sk++]); - ++iter->offset; - } + if (iter->cur_sk < iter->end_sk) + sock_put(iter->batch[iter->cur_sk++].sk); /* After updating iter->cur_sk, check if there are more sockets * available in the current bucket batch. */ if (iter->cur_sk < iter->end_sk) - sk = iter->batch[iter->cur_sk]; + sk = iter->batch[iter->cur_sk].sk; else /* Prepare a new batch. */ sk = bpf_iter_udp_batch(seq); @@ -3268,7 +3655,7 @@ static int bpf_iter_udp_seq_show(struct seq_file *seq, void *v) goto unlock; } - uid = from_kuid_munged(seq_user_ns(seq), sock_i_uid(sk)); + uid = from_kuid_munged(seq_user_ns(seq), sk_uid(sk)); meta.seq = seq; prog = bpf_iter_get_info(&meta, false); ret = udp_prog_seq_show(prog, &meta, v, uid, state->bucket); @@ -3280,8 +3667,19 @@ unlock: static void bpf_iter_udp_put_batch(struct bpf_udp_iter_state *iter) { - while (iter->cur_sk < iter->end_sk) - sock_put(iter->batch[iter->cur_sk++]); + union bpf_udp_iter_batch_item *item; + unsigned int cur_sk = iter->cur_sk; + __u64 cookie; + + /* Remember the cookies of the sockets we haven't seen yet, so we can + * pick up where we left off next time around. + */ + while (cur_sk < iter->end_sk) { + item = &iter->batch[cur_sk++]; + cookie = sock_gen_cookie(item->sk); + sock_put(item->sk); + item->cookie = cookie; + } } static void bpf_iter_udp_seq_stop(struct seq_file *seq, void *v) @@ -3297,10 +3695,8 @@ static void bpf_iter_udp_seq_stop(struct seq_file *seq, void *v) (void)udp_prog_seq_show(prog, &meta, v, 0, 0); } - if (iter->cur_sk < iter->end_sk) { + if (iter->cur_sk < iter->end_sk) bpf_iter_udp_put_batch(iter); - iter->st_bucket_done = false; - } } static const struct seq_operations bpf_iter_udp_seq_ops = { @@ -3332,7 +3728,7 @@ const struct seq_operations udp_seq_ops = { .stop = udp_seq_stop, .show = udp4_seq_show, }; -EXPORT_SYMBOL(udp_seq_ops); +EXPORT_IPV6_MOD(udp_seq_ops); static struct udp_seq_afinfo udp4_seq_afinfo = { .family = AF_INET, @@ -3388,10 +3784,12 @@ __setup("uhash_entries=", set_uhash_entries); void __init udp_table_init(struct udp_table *table, const char *name) { - unsigned int i; + unsigned int i, slot_size; + slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main) + + udp_hash4_slot_size(); table->hash = alloc_large_system_hash(name, - 2 * sizeof(struct udp_hslot), + slot_size, uhash_entries, 21, /* one slot per 2 MB */ 0, @@ -3400,17 +3798,18 @@ void __init udp_table_init(struct udp_table *table, const char *name) UDP_HTABLE_SIZE_MIN, UDP_HTABLE_SIZE_MAX); - table->hash2 = table->hash + (table->mask + 1); + table->hash2 = (void *)(table->hash + (table->mask + 1)); for (i = 0; i <= table->mask; i++) { INIT_HLIST_HEAD(&table->hash[i].head); table->hash[i].count = 0; spin_lock_init(&table->hash[i].lock); } for (i = 0; i <= table->mask; i++) { - INIT_HLIST_HEAD(&table->hash2[i].head); - table->hash2[i].count = 0; - spin_lock_init(&table->hash2[i].lock); + INIT_HLIST_HEAD(&table->hash2[i].hslot.head); + table->hash2[i].hslot.count = 0; + spin_lock_init(&table->hash2[i].hslot.lock); } + udp_table_hash4_init(table); } u32 udp_flow_hashrnd(void) @@ -3436,18 +3835,21 @@ static void __net_init udp_sysctl_init(struct net *net) static struct udp_table __net_init *udp_pernet_table_alloc(unsigned int hash_entries) { struct udp_table *udptable; + unsigned int slot_size; int i; udptable = kmalloc(sizeof(*udptable), GFP_KERNEL); if (!udptable) goto out; - udptable->hash = vmalloc_huge(hash_entries * 2 * sizeof(struct udp_hslot), + slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main) + + udp_hash4_slot_size(); + udptable->hash = vmalloc_huge(hash_entries * slot_size, GFP_KERNEL_ACCOUNT); if (!udptable->hash) goto free_table; - udptable->hash2 = udptable->hash + hash_entries; + udptable->hash2 = (void *)(udptable->hash + hash_entries); udptable->mask = hash_entries - 1; udptable->log = ilog2(hash_entries); @@ -3456,10 +3858,11 @@ static struct udp_table __net_init *udp_pernet_table_alloc(unsigned int hash_ent udptable->hash[i].count = 0; spin_lock_init(&udptable->hash[i].lock); - INIT_HLIST_HEAD(&udptable->hash2[i].head); - udptable->hash2[i].count = 0; - spin_lock_init(&udptable->hash2[i].lock); + INIT_HLIST_HEAD(&udptable->hash2[i].hslot.head); + udptable->hash2[i].hslot.count = 0; + spin_lock_init(&udptable->hash2[i].hslot.lock); } + udp_table_hash4_init(udptable); return udptable; @@ -3514,6 +3917,15 @@ fallback: static int __net_init udp_pernet_init(struct net *net) { +#if IS_ENABLED(CONFIG_NET_UDP_TUNNEL) + int i; + + /* No tunnel is configured */ + for (i = 0; i < ARRAY_SIZE(net->ipv4.udp_tunnel_gro); ++i) { + INIT_HLIST_HEAD(&net->ipv4.udp_tunnel_gro[i].list); + RCU_INIT_POINTER(net->ipv4.udp_tunnel_gro[i].sk, NULL); + } +#endif udp_sysctl_init(net); udp_set_table(net); @@ -3535,16 +3947,19 @@ DEFINE_BPF_ITER_FUNC(udp, struct bpf_iter_meta *meta, struct udp_sock *udp_sk, uid_t uid, int bucket) static int bpf_iter_udp_realloc_batch(struct bpf_udp_iter_state *iter, - unsigned int new_batch_sz) + unsigned int new_batch_sz, gfp_t flags) { - struct sock **new_batch; + union bpf_udp_iter_batch_item *new_batch; new_batch = kvmalloc_array(new_batch_sz, sizeof(*new_batch), - GFP_USER | __GFP_NOWARN); + flags | __GFP_NOWARN); if (!new_batch) return -ENOMEM; - bpf_iter_udp_put_batch(iter); + if (flags != GFP_NOWAIT) + bpf_iter_udp_put_batch(iter); + + memcpy(new_batch, iter->batch, sizeof(*iter->batch) * iter->end_sk); kvfree(iter->batch); iter->batch = new_batch; iter->max_sk = new_batch_sz; @@ -3563,10 +3978,12 @@ static int bpf_iter_init_udp(void *priv_data, struct bpf_iter_aux_info *aux) if (ret) return ret; - ret = bpf_iter_udp_realloc_batch(iter, INIT_BATCH_SZ); + ret = bpf_iter_udp_realloc_batch(iter, INIT_BATCH_SZ, GFP_USER); if (ret) bpf_iter_fini_seq_net(priv_data); + iter->state.bucket = -1; + return ret; } @@ -3606,7 +4023,6 @@ static void __init bpf_iter_register(void) void __init udp_init(void) { unsigned long limit; - unsigned int i; udp_table_init(&udp_table, "UDP"); limit = nr_free_buffer_pages() / 8; @@ -3615,15 +4031,6 @@ void __init udp_init(void) sysctl_udp_mem[1] = limit; sysctl_udp_mem[2] = sysctl_udp_mem[0] * 2; - /* 16 spinlocks per cpu */ - udp_busylocks_log = ilog2(nr_cpu_ids) + 4; - udp_busylocks = kmalloc(sizeof(spinlock_t) << udp_busylocks_log, - GFP_KERNEL); - if (!udp_busylocks) - panic("UDP: failed to alloc udp_busylocks\n"); - for (i = 0; i < (1U << udp_busylocks_log); i++) - spin_lock_init(udp_busylocks + i); - if (register_pernet_subsys(&udp_sysctl_ops)) panic("UDP: failed to init sysctl parameters.\n"); |