From 130ffbc2638ddc290fcbabe1b9ce6a5d333a6a97 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Wed, 12 Jun 2013 17:54:51 +0200 Subject: [PATCH 01/12] netfilter: check return code from nla_parse_tested These are the only calls under net/ that do not check nla_parse_nested() for its error code, but simply continue execution. If parsing of netlink attributes fails, we should return with an error instead of continuing. In nearly all of these calls we have a policy attached, that is being type verified during nla_parse_nested(), which we would miss checking for otherwise. Signed-off-by: Daniel Borkmann Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_netlink.c | 30 +++++++++++++++++++++------- net/netfilter/nfnetlink_cthelper.c | 16 +++++++++++---- net/netfilter/nfnetlink_cttimeout.c | 6 ++++-- 3 files changed, 39 insertions(+), 13 deletions(-) diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 6d0f8a17c5b7..f83a52298efe 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -828,7 +828,9 @@ ctnetlink_parse_tuple_ip(struct nlattr *attr, struct nf_conntrack_tuple *tuple) struct nf_conntrack_l3proto *l3proto; int ret = 0; - nla_parse_nested(tb, CTA_IP_MAX, attr, NULL); + ret = nla_parse_nested(tb, CTA_IP_MAX, attr, NULL); + if (ret < 0) + return ret; rcu_read_lock(); l3proto = __nf_ct_l3proto_find(tuple->src.l3num); @@ -895,7 +897,9 @@ ctnetlink_parse_tuple(const struct nlattr * const cda[], memset(tuple, 0, sizeof(*tuple)); - nla_parse_nested(tb, CTA_TUPLE_MAX, cda[type], tuple_nla_policy); + err = nla_parse_nested(tb, CTA_TUPLE_MAX, cda[type], tuple_nla_policy); + if (err < 0) + return err; if (!tb[CTA_TUPLE_IP]) return -EINVAL; @@ -946,9 +950,12 @@ static inline int ctnetlink_parse_help(const struct nlattr *attr, char **helper_name, struct nlattr **helpinfo) { + int err; struct nlattr *tb[CTA_HELP_MAX+1]; - nla_parse_nested(tb, CTA_HELP_MAX, attr, help_nla_policy); + err = nla_parse_nested(tb, CTA_HELP_MAX, attr, help_nla_policy); + if (err < 0) + return err; if (!tb[CTA_HELP_NAME]) return -EINVAL; @@ -1431,7 +1438,9 @@ ctnetlink_change_protoinfo(struct nf_conn *ct, const struct nlattr * const cda[] struct nf_conntrack_l4proto *l4proto; int err = 0; - nla_parse_nested(tb, CTA_PROTOINFO_MAX, attr, protoinfo_policy); + err = nla_parse_nested(tb, CTA_PROTOINFO_MAX, attr, protoinfo_policy); + if (err < 0) + return err; rcu_read_lock(); l4proto = __nf_ct_l4proto_find(nf_ct_l3num(ct), nf_ct_protonum(ct)); @@ -1452,9 +1461,12 @@ static const struct nla_policy nat_seq_policy[CTA_NAT_SEQ_MAX+1] = { static inline int change_nat_seq_adj(struct nf_nat_seq *natseq, const struct nlattr * const attr) { + int err; struct nlattr *cda[CTA_NAT_SEQ_MAX+1]; - nla_parse_nested(cda, CTA_NAT_SEQ_MAX, attr, nat_seq_policy); + err = nla_parse_nested(cda, CTA_NAT_SEQ_MAX, attr, nat_seq_policy); + if (err < 0) + return err; if (!cda[CTA_NAT_SEQ_CORRECTION_POS]) return -EINVAL; @@ -2115,7 +2127,9 @@ ctnetlink_nfqueue_parse(const struct nlattr *attr, struct nf_conn *ct) struct nlattr *cda[CTA_MAX+1]; int ret; - nla_parse_nested(cda, CTA_MAX, attr, ct_nla_policy); + ret = nla_parse_nested(cda, CTA_MAX, attr, ct_nla_policy); + if (ret < 0) + return ret; spin_lock_bh(&nf_conntrack_lock); ret = ctnetlink_nfqueue_parse_ct((const struct nlattr **)cda, ct); @@ -2710,7 +2724,9 @@ ctnetlink_parse_expect_nat(const struct nlattr *attr, struct nf_conntrack_tuple nat_tuple = {}; int err; - nla_parse_nested(tb, CTA_EXPECT_NAT_MAX, attr, exp_nat_nla_policy); + err = nla_parse_nested(tb, CTA_EXPECT_NAT_MAX, attr, exp_nat_nla_policy); + if (err < 0) + return err; if (!tb[CTA_EXPECT_NAT_DIR] || !tb[CTA_EXPECT_NAT_TUPLE]) return -EINVAL; diff --git a/net/netfilter/nfnetlink_cthelper.c b/net/netfilter/nfnetlink_cthelper.c index a191b6db657e..9e287cb56a04 100644 --- a/net/netfilter/nfnetlink_cthelper.c +++ b/net/netfilter/nfnetlink_cthelper.c @@ -67,9 +67,12 @@ static int nfnl_cthelper_parse_tuple(struct nf_conntrack_tuple *tuple, const struct nlattr *attr) { + int err; struct nlattr *tb[NFCTH_TUPLE_MAX+1]; - nla_parse_nested(tb, NFCTH_TUPLE_MAX, attr, nfnl_cthelper_tuple_pol); + err = nla_parse_nested(tb, NFCTH_TUPLE_MAX, attr, nfnl_cthelper_tuple_pol); + if (err < 0) + return err; if (!tb[NFCTH_TUPLE_L3PROTONUM] || !tb[NFCTH_TUPLE_L4PROTONUM]) return -EINVAL; @@ -121,9 +124,12 @@ static int nfnl_cthelper_expect_policy(struct nf_conntrack_expect_policy *expect_policy, const struct nlattr *attr) { + int err; struct nlattr *tb[NFCTH_POLICY_MAX+1]; - nla_parse_nested(tb, NFCTH_POLICY_MAX, attr, nfnl_cthelper_expect_pol); + err = nla_parse_nested(tb, NFCTH_POLICY_MAX, attr, nfnl_cthelper_expect_pol); + if (err < 0) + return err; if (!tb[NFCTH_POLICY_NAME] || !tb[NFCTH_POLICY_EXPECT_MAX] || @@ -153,8 +159,10 @@ nfnl_cthelper_parse_expect_policy(struct nf_conntrack_helper *helper, struct nf_conntrack_expect_policy *expect_policy; struct nlattr *tb[NFCTH_POLICY_SET_MAX+1]; - nla_parse_nested(tb, NFCTH_POLICY_SET_MAX, attr, - nfnl_cthelper_expect_policy_set); + ret = nla_parse_nested(tb, NFCTH_POLICY_SET_MAX, attr, + nfnl_cthelper_expect_policy_set); + if (ret < 0) + return ret; if (!tb[NFCTH_POLICY_SET_NUM]) return -EINVAL; diff --git a/net/netfilter/nfnetlink_cttimeout.c b/net/netfilter/nfnetlink_cttimeout.c index 65074dfb9383..50580494148d 100644 --- a/net/netfilter/nfnetlink_cttimeout.c +++ b/net/netfilter/nfnetlink_cttimeout.c @@ -59,8 +59,10 @@ ctnl_timeout_parse_policy(struct ctnl_timeout *timeout, if (likely(l4proto->ctnl_timeout.nlattr_to_obj)) { struct nlattr *tb[l4proto->ctnl_timeout.nlattr_max+1]; - nla_parse_nested(tb, l4proto->ctnl_timeout.nlattr_max, - attr, l4proto->ctnl_timeout.nla_policy); + ret = nla_parse_nested(tb, l4proto->ctnl_timeout.nlattr_max, + attr, l4proto->ctnl_timeout.nla_policy); + if (ret < 0) + return ret; ret = l4proto->ctnl_timeout.nlattr_to_obj(tb, net, &timeout->data); From 6547a221871f139cc56328a38105d47c14874cbe Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 13 Jun 2013 17:31:28 +0200 Subject: [PATCH 02/12] netfilter: nf_conntrack: avoid large timeout for mid-stream pickup When loose tracking is enabled (default), non-syn packets cause creation of new conntracks in established state with default timeout for established state (5 days). This causes the table to fill up with UNREPLIED when the 'new ack' packet happened to be the last-ack of a previous, already timed-out connection. Consider: A 192.168.x.52792 > 10.184.y.80: F, 426:426(0) ack 9237 win 255 B 10.184.y.80 > 192.168.x.52792: ., ack 427 win 123 <61 second pause> C 10.184.y.80 > 192.168.x.52792: F, 9237:9237(0) ack 427 win 123 D 192.168.x.52792 > 10.184.y.80: ., ack 9238 win 255 B moves conntrack to CLOSE_WAIT and will kill it after 60 second timeout, C is ignored (FIN set), but last packet (D) causes new ct with 5-days timeout. Use UNACK timeout (5 minutes) instead to get rid of these entries sooner when in ESTABLISHED state without having seen traffic in both directions. Signed-off-by: Florian Westphal Acked-by: Jozsef Kadlecsik Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_proto_tcp.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c index 4d4d8f1d01fc..7dcc376eea5f 100644 --- a/net/netfilter/nf_conntrack_proto_tcp.c +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -1043,6 +1043,12 @@ static int tcp_packet(struct nf_conn *ct, nf_ct_kill_acct(ct, ctinfo, skb); return NF_ACCEPT; } + /* ESTABLISHED without SEEN_REPLY, i.e. mid-connection + * pickup with loose=1. Avoid large ESTABLISHED timeout. + */ + if (new_state == TCP_CONNTRACK_ESTABLISHED && + timeout > timeouts[TCP_CONNTRACK_UNACK]) + timeout = timeouts[TCP_CONNTRACK_UNACK]; } else if (!test_bit(IPS_ASSURED_BIT, &ct->status) && (old_state == TCP_CONNTRACK_SYN_RECV || old_state == TCP_CONNTRACK_ESTABLISHED) From 681f130f39e10087475383e6771b9366e26bab0c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 20 Jun 2013 05:52:22 -0700 Subject: [PATCH 03/12] netfilter: xt_socket: add XT_SOCKET_NOWILDCARD flag xt_socket module can be a nice replacement to conntrack module in some cases (SYN filtering for example) But it lacks the ability to match the 3rd packet of TCP handshake (ACK coming from the client). Add a XT_SOCKET_NOWILDCARD flag to disable the wildcard mechanism. The wildcard is the legacy socket match behavior, that ignores LISTEN sockets bound to INADDR_ANY (or ipv6 equivalent) iptables -I INPUT -p tcp --syn -j SYN_CHAIN iptables -I INPUT -m socket --nowildcard -j ACCEPT Signed-off-by: Eric Dumazet Cc: Patrick McHardy Cc: Jesper Dangaard Brouer Signed-off-by: Pablo Neira Ayuso --- include/uapi/linux/netfilter/xt_socket.h | 7 +++ net/netfilter/xt_socket.c | 70 +++++++++++++++++++++--- 2 files changed, 69 insertions(+), 8 deletions(-) diff --git a/include/uapi/linux/netfilter/xt_socket.h b/include/uapi/linux/netfilter/xt_socket.h index 26d7217bd4f1..6315e2ac3474 100644 --- a/include/uapi/linux/netfilter/xt_socket.h +++ b/include/uapi/linux/netfilter/xt_socket.h @@ -5,10 +5,17 @@ enum { XT_SOCKET_TRANSPARENT = 1 << 0, + XT_SOCKET_NOWILDCARD = 1 << 1, }; struct xt_socket_mtinfo1 { __u8 flags; }; +#define XT_SOCKET_FLAGS_V1 XT_SOCKET_TRANSPARENT + +struct xt_socket_mtinfo2 { + __u8 flags; +}; +#define XT_SOCKET_FLAGS_V2 (XT_SOCKET_TRANSPARENT | XT_SOCKET_NOWILDCARD) #endif /* _XT_SOCKET_H */ diff --git a/net/netfilter/xt_socket.c b/net/netfilter/xt_socket.c index 02704245710e..f8b71911037a 100644 --- a/net/netfilter/xt_socket.c +++ b/net/netfilter/xt_socket.c @@ -163,8 +163,11 @@ socket_match(const struct sk_buff *skb, struct xt_action_param *par, bool wildcard; bool transparent = true; - /* Ignore sockets listening on INADDR_ANY */ - wildcard = (sk->sk_state != TCP_TIME_WAIT && + /* Ignore sockets listening on INADDR_ANY, + * unless XT_SOCKET_NOWILDCARD is set + */ + wildcard = (!(info->flags & XT_SOCKET_NOWILDCARD) && + sk->sk_state != TCP_TIME_WAIT && inet_sk(sk)->inet_rcv_saddr == 0); /* Ignore non-transparent sockets, @@ -197,7 +200,7 @@ socket_mt4_v0(const struct sk_buff *skb, struct xt_action_param *par) } static bool -socket_mt4_v1(const struct sk_buff *skb, struct xt_action_param *par) +socket_mt4_v1_v2(const struct sk_buff *skb, struct xt_action_param *par) { return socket_match(skb, par, par->matchinfo); } @@ -259,7 +262,7 @@ extract_icmp6_fields(const struct sk_buff *skb, } static bool -socket_mt6_v1(const struct sk_buff *skb, struct xt_action_param *par) +socket_mt6_v1_v2(const struct sk_buff *skb, struct xt_action_param *par) { struct ipv6hdr *iph = ipv6_hdr(skb); struct udphdr _hdr, *hp = NULL; @@ -302,8 +305,11 @@ socket_mt6_v1(const struct sk_buff *skb, struct xt_action_param *par) bool wildcard; bool transparent = true; - /* Ignore sockets listening on INADDR_ANY */ - wildcard = (sk->sk_state != TCP_TIME_WAIT && + /* Ignore sockets listening on INADDR_ANY + * unless XT_SOCKET_NOWILDCARD is set + */ + wildcard = (!(info->flags & XT_SOCKET_NOWILDCARD) && + sk->sk_state != TCP_TIME_WAIT && ipv6_addr_any(&inet6_sk(sk)->rcv_saddr)); /* Ignore non-transparent sockets, @@ -331,6 +337,28 @@ socket_mt6_v1(const struct sk_buff *skb, struct xt_action_param *par) } #endif +static int socket_mt_v1_check(const struct xt_mtchk_param *par) +{ + const struct xt_socket_mtinfo1 *info = (struct xt_socket_mtinfo1 *) par->matchinfo; + + if (info->flags & ~XT_SOCKET_FLAGS_V1) { + pr_info("unknown flags 0x%x\n", info->flags & ~XT_SOCKET_FLAGS_V1); + return -EINVAL; + } + return 0; +} + +static int socket_mt_v2_check(const struct xt_mtchk_param *par) +{ + const struct xt_socket_mtinfo2 *info = (struct xt_socket_mtinfo2 *) par->matchinfo; + + if (info->flags & ~XT_SOCKET_FLAGS_V2) { + pr_info("unknown flags 0x%x\n", info->flags & ~XT_SOCKET_FLAGS_V2); + return -EINVAL; + } + return 0; +} + static struct xt_match socket_mt_reg[] __read_mostly = { { .name = "socket", @@ -345,7 +373,8 @@ static struct xt_match socket_mt_reg[] __read_mostly = { .name = "socket", .revision = 1, .family = NFPROTO_IPV4, - .match = socket_mt4_v1, + .match = socket_mt4_v1_v2, + .checkentry = socket_mt_v1_check, .matchsize = sizeof(struct xt_socket_mtinfo1), .hooks = (1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_IN), @@ -356,7 +385,32 @@ static struct xt_match socket_mt_reg[] __read_mostly = { .name = "socket", .revision = 1, .family = NFPROTO_IPV6, - .match = socket_mt6_v1, + .match = socket_mt6_v1_v2, + .checkentry = socket_mt_v1_check, + .matchsize = sizeof(struct xt_socket_mtinfo1), + .hooks = (1 << NF_INET_PRE_ROUTING) | + (1 << NF_INET_LOCAL_IN), + .me = THIS_MODULE, + }, +#endif + { + .name = "socket", + .revision = 2, + .family = NFPROTO_IPV4, + .match = socket_mt4_v1_v2, + .checkentry = socket_mt_v2_check, + .matchsize = sizeof(struct xt_socket_mtinfo1), + .hooks = (1 << NF_INET_PRE_ROUTING) | + (1 << NF_INET_LOCAL_IN), + .me = THIS_MODULE, + }, +#ifdef XT_SOCKET_HAVE_IPV6 + { + .name = "socket", + .revision = 2, + .family = NFPROTO_IPV6, + .match = socket_mt6_v1_v2, + .checkentry = socket_mt_v2_check, .matchsize = sizeof(struct xt_socket_mtinfo1), .hooks = (1 << NF_INET_PRE_ROUTING) | (1 << NF_INET_LOCAL_IN), From bba54de5bdd107d3841b560f1a9cb0ed06e79533 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Sun, 16 Jun 2013 09:09:36 +0300 Subject: [PATCH 04/12] ipvs: provide iph to schedulers Before now the schedulers needed access only to IP addresses and it was easy to get them from skb by using ip_vs_fill_iph_addr_only. New changes for the SH scheduler will need the protocol and ports which is difficult to get from skb for the IPv6 case. As we have all the data in the iph structure, to avoid the same slow lookups provide the iph to schedulers. Signed-off-by: Julian Anastasov Acked-by: Hans Schillstrom Signed-off-by: Simon Horman --- include/net/ip_vs.h | 28 ++-------------------------- net/netfilter/ipvs/ip_vs_core.c | 4 ++-- net/netfilter/ipvs/ip_vs_dh.c | 10 ++++------ net/netfilter/ipvs/ip_vs_lblc.c | 12 +++++------- net/netfilter/ipvs/ip_vs_lblcr.c | 12 +++++------- net/netfilter/ipvs/ip_vs_lc.c | 3 ++- net/netfilter/ipvs/ip_vs_nq.c | 3 ++- net/netfilter/ipvs/ip_vs_rr.c | 3 ++- net/netfilter/ipvs/ip_vs_sed.c | 3 ++- net/netfilter/ipvs/ip_vs_sh.c | 10 ++++------ net/netfilter/ipvs/ip_vs_wlc.c | 3 ++- net/netfilter/ipvs/ip_vs_wrr.c | 3 ++- 12 files changed, 34 insertions(+), 60 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 4405886980c7..f5faf859876e 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -197,31 +197,6 @@ ip_vs_fill_iph_skb(int af, const struct sk_buff *skb, struct ip_vs_iphdr *iphdr) } } -/* This function is a faster version of ip_vs_fill_iph_skb(). - * Where we only populate {s,d}addr (and avoid calling ipv6_find_hdr()). - * This is used by the some of the ip_vs_*_schedule() functions. - * (Mostly done to avoid ABI breakage of external schedulers) - */ -static inline void -ip_vs_fill_iph_addr_only(int af, const struct sk_buff *skb, - struct ip_vs_iphdr *iphdr) -{ -#ifdef CONFIG_IP_VS_IPV6 - if (af == AF_INET6) { - const struct ipv6hdr *iph = - (struct ipv6hdr *)skb_network_header(skb); - iphdr->saddr.in6 = iph->saddr; - iphdr->daddr.in6 = iph->daddr; - } else -#endif - { - const struct iphdr *iph = - (struct iphdr *)skb_network_header(skb); - iphdr->saddr.ip = iph->saddr; - iphdr->daddr.ip = iph->daddr; - } -} - static inline void ip_vs_addr_copy(int af, union nf_inet_addr *dst, const union nf_inet_addr *src) { @@ -814,7 +789,8 @@ struct ip_vs_scheduler { /* selecting a server from the given service */ struct ip_vs_dest* (*schedule)(struct ip_vs_service *svc, - const struct sk_buff *skb); + const struct sk_buff *skb, + struct ip_vs_iphdr *iph); }; /* The persistence engine object */ diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 05565d2b3a61..e9b0330f220d 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -305,7 +305,7 @@ ip_vs_sched_persist(struct ip_vs_service *svc, * return *ignored=0 i.e. ICMP and NF_DROP */ sched = rcu_dereference(svc->scheduler); - dest = sched->schedule(svc, skb); + dest = sched->schedule(svc, skb, iph); if (!dest) { IP_VS_DBG(1, "p-schedule: no dest found.\n"); kfree(param.pe_data); @@ -452,7 +452,7 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb, } sched = rcu_dereference(svc->scheduler); - dest = sched->schedule(svc, skb); + dest = sched->schedule(svc, skb, iph); if (dest == NULL) { IP_VS_DBG(1, "Schedule: no dest found.\n"); return NULL; diff --git a/net/netfilter/ipvs/ip_vs_dh.c b/net/netfilter/ipvs/ip_vs_dh.c index ccab120df45e..c3b84546ea9e 100644 --- a/net/netfilter/ipvs/ip_vs_dh.c +++ b/net/netfilter/ipvs/ip_vs_dh.c @@ -214,18 +214,16 @@ static inline int is_overloaded(struct ip_vs_dest *dest) * Destination hashing scheduling */ static struct ip_vs_dest * -ip_vs_dh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_dh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest; struct ip_vs_dh_state *s; - struct ip_vs_iphdr iph; - - ip_vs_fill_iph_addr_only(svc->af, skb, &iph); IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); s = (struct ip_vs_dh_state *) svc->sched_data; - dest = ip_vs_dh_get(svc->af, s, &iph.daddr); + dest = ip_vs_dh_get(svc->af, s, &iph->daddr); if (!dest || !(dest->flags & IP_VS_DEST_F_AVAILABLE) || atomic_read(&dest->weight) <= 0 @@ -235,7 +233,7 @@ ip_vs_dh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) } IP_VS_DBG_BUF(6, "DH: destination IP address %s --> server %s:%d\n", - IP_VS_DBG_ADDR(svc->af, &iph.daddr), + IP_VS_DBG_ADDR(svc->af, &iph->daddr), IP_VS_DBG_ADDR(svc->af, &dest->addr), ntohs(dest->port)); diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c index 44595b8ae37f..1383b0eadc0e 100644 --- a/net/netfilter/ipvs/ip_vs_lblc.c +++ b/net/netfilter/ipvs/ip_vs_lblc.c @@ -487,19 +487,17 @@ is_overloaded(struct ip_vs_dest *dest, struct ip_vs_service *svc) * Locality-Based (weighted) Least-Connection scheduling */ static struct ip_vs_dest * -ip_vs_lblc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_lblc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_lblc_table *tbl = svc->sched_data; - struct ip_vs_iphdr iph; struct ip_vs_dest *dest = NULL; struct ip_vs_lblc_entry *en; - ip_vs_fill_iph_addr_only(svc->af, skb, &iph); - IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* First look in our cache */ - en = ip_vs_lblc_get(svc->af, tbl, &iph.daddr); + en = ip_vs_lblc_get(svc->af, tbl, &iph->daddr); if (en) { /* We only hold a read lock, but this is atomic */ en->lastuse = jiffies; @@ -529,12 +527,12 @@ ip_vs_lblc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) /* If we fail to create a cache entry, we'll just use the valid dest */ spin_lock_bh(&svc->sched_lock); if (!tbl->dead) - ip_vs_lblc_new(tbl, &iph.daddr, dest); + ip_vs_lblc_new(tbl, &iph->daddr, dest); spin_unlock_bh(&svc->sched_lock); out: IP_VS_DBG_BUF(6, "LBLC: destination IP address %s --> server %s:%d\n", - IP_VS_DBG_ADDR(svc->af, &iph.daddr), + IP_VS_DBG_ADDR(svc->af, &iph->daddr), IP_VS_DBG_ADDR(svc->af, &dest->addr), ntohs(dest->port)); return dest; diff --git a/net/netfilter/ipvs/ip_vs_lblcr.c b/net/netfilter/ipvs/ip_vs_lblcr.c index 876937db0bf4..3cd85b2fc67c 100644 --- a/net/netfilter/ipvs/ip_vs_lblcr.c +++ b/net/netfilter/ipvs/ip_vs_lblcr.c @@ -655,19 +655,17 @@ is_overloaded(struct ip_vs_dest *dest, struct ip_vs_service *svc) * Locality-Based (weighted) Least-Connection scheduling */ static struct ip_vs_dest * -ip_vs_lblcr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_lblcr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_lblcr_table *tbl = svc->sched_data; - struct ip_vs_iphdr iph; struct ip_vs_dest *dest; struct ip_vs_lblcr_entry *en; - ip_vs_fill_iph_addr_only(svc->af, skb, &iph); - IP_VS_DBG(6, "%s(): Scheduling...\n", __func__); /* First look in our cache */ - en = ip_vs_lblcr_get(svc->af, tbl, &iph.daddr); + en = ip_vs_lblcr_get(svc->af, tbl, &iph->daddr); if (en) { en->lastuse = jiffies; @@ -718,12 +716,12 @@ ip_vs_lblcr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) /* If we fail to create a cache entry, we'll just use the valid dest */ spin_lock_bh(&svc->sched_lock); if (!tbl->dead) - ip_vs_lblcr_new(tbl, &iph.daddr, dest); + ip_vs_lblcr_new(tbl, &iph->daddr, dest); spin_unlock_bh(&svc->sched_lock); out: IP_VS_DBG_BUF(6, "LBLCR: destination IP address %s --> server %s:%d\n", - IP_VS_DBG_ADDR(svc->af, &iph.daddr), + IP_VS_DBG_ADDR(svc->af, &iph->daddr), IP_VS_DBG_ADDR(svc->af, &dest->addr), ntohs(dest->port)); return dest; diff --git a/net/netfilter/ipvs/ip_vs_lc.c b/net/netfilter/ipvs/ip_vs_lc.c index 5128e338a749..2bdcb1cf2127 100644 --- a/net/netfilter/ipvs/ip_vs_lc.c +++ b/net/netfilter/ipvs/ip_vs_lc.c @@ -26,7 +26,8 @@ * Least Connection scheduling */ static struct ip_vs_dest * -ip_vs_lc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_lc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest, *least = NULL; unsigned int loh = 0, doh; diff --git a/net/netfilter/ipvs/ip_vs_nq.c b/net/netfilter/ipvs/ip_vs_nq.c index 646cfd4baa73..d8d9860934fe 100644 --- a/net/netfilter/ipvs/ip_vs_nq.c +++ b/net/netfilter/ipvs/ip_vs_nq.c @@ -55,7 +55,8 @@ ip_vs_nq_dest_overhead(struct ip_vs_dest *dest) * Weighted Least Connection scheduling */ static struct ip_vs_dest * -ip_vs_nq_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_nq_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest, *least = NULL; unsigned int loh = 0, doh; diff --git a/net/netfilter/ipvs/ip_vs_rr.c b/net/netfilter/ipvs/ip_vs_rr.c index c35986c793d9..176b87c35e34 100644 --- a/net/netfilter/ipvs/ip_vs_rr.c +++ b/net/netfilter/ipvs/ip_vs_rr.c @@ -55,7 +55,8 @@ static int ip_vs_rr_del_dest(struct ip_vs_service *svc, struct ip_vs_dest *dest) * Round-Robin Scheduling */ static struct ip_vs_dest * -ip_vs_rr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_rr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct list_head *p; struct ip_vs_dest *dest, *last; diff --git a/net/netfilter/ipvs/ip_vs_sed.c b/net/netfilter/ipvs/ip_vs_sed.c index f3205925359a..a5284cc3d882 100644 --- a/net/netfilter/ipvs/ip_vs_sed.c +++ b/net/netfilter/ipvs/ip_vs_sed.c @@ -59,7 +59,8 @@ ip_vs_sed_dest_overhead(struct ip_vs_dest *dest) * Weighted Least Connection scheduling */ static struct ip_vs_dest * -ip_vs_sed_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_sed_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest, *least; unsigned int loh, doh; diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c index a65edfe4b16c..e0d5d1653566 100644 --- a/net/netfilter/ipvs/ip_vs_sh.c +++ b/net/netfilter/ipvs/ip_vs_sh.c @@ -227,18 +227,16 @@ static inline int is_overloaded(struct ip_vs_dest *dest) * Source Hashing scheduling */ static struct ip_vs_dest * -ip_vs_sh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_sh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest; struct ip_vs_sh_state *s; - struct ip_vs_iphdr iph; - - ip_vs_fill_iph_addr_only(svc->af, skb, &iph); IP_VS_DBG(6, "ip_vs_sh_schedule(): Scheduling...\n"); s = (struct ip_vs_sh_state *) svc->sched_data; - dest = ip_vs_sh_get(svc->af, s, &iph.saddr); + dest = ip_vs_sh_get(svc->af, s, &iph->saddr); if (!dest || !(dest->flags & IP_VS_DEST_F_AVAILABLE) || atomic_read(&dest->weight) <= 0 @@ -248,7 +246,7 @@ ip_vs_sh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) } IP_VS_DBG_BUF(6, "SH: source IP address %s --> server %s:%d\n", - IP_VS_DBG_ADDR(svc->af, &iph.saddr), + IP_VS_DBG_ADDR(svc->af, &iph->saddr), IP_VS_DBG_ADDR(svc->af, &dest->addr), ntohs(dest->port)); diff --git a/net/netfilter/ipvs/ip_vs_wlc.c b/net/netfilter/ipvs/ip_vs_wlc.c index c60a81c4ce9a..6dc1fa128840 100644 --- a/net/netfilter/ipvs/ip_vs_wlc.c +++ b/net/netfilter/ipvs/ip_vs_wlc.c @@ -31,7 +31,8 @@ * Weighted Least Connection scheduling */ static struct ip_vs_dest * -ip_vs_wlc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_wlc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest, *least; unsigned int loh, doh; diff --git a/net/netfilter/ipvs/ip_vs_wrr.c b/net/netfilter/ipvs/ip_vs_wrr.c index 0e68555bceb9..0546cd572d6b 100644 --- a/net/netfilter/ipvs/ip_vs_wrr.c +++ b/net/netfilter/ipvs/ip_vs_wrr.c @@ -162,7 +162,8 @@ static int ip_vs_wrr_dest_changed(struct ip_vs_service *svc, * Weighted Round-Robin Scheduling */ static struct ip_vs_dest * -ip_vs_wrr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb) +ip_vs_wrr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, + struct ip_vs_iphdr *iph) { struct ip_vs_dest *dest, *last, *stop = NULL; struct ip_vs_wrr_mark *mark = svc->sched_data; From c6c96c188336b2b95d5f14facd101f1e4165a9d3 Mon Sep 17 00:00:00 2001 From: Alexander Frolkin Date: Thu, 13 Jun 2013 08:56:15 +0100 Subject: [PATCH 05/12] ipvs: sloppy TCP and SCTP This adds support for sloppy TCP and SCTP modes to IPVS. When enabled (sysctls net.ipv4.vs.sloppy_tcp and net.ipv4.vs.sloppy_sctp), allows IPVS to create connection state on any packet, not just a TCP SYN (or SCTP INIT). This allows connections to fail over from one IPVS director to another mid-flight. Signed-off-by: Alexander Frolkin Signed-off-by: Julian Anastasov Signed-off-by: Simon Horman --- include/net/ip_vs.h | 24 ++++++++++++++++++++++++ net/netfilter/ipvs/ip_vs_ctl.c | 14 ++++++++++++++ net/netfilter/ipvs/ip_vs_proto_sctp.c | 18 ++++++++++-------- net/netfilter/ipvs/ip_vs_proto_tcp.c | 14 ++++++++------ 4 files changed, 56 insertions(+), 14 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index f5faf859876e..95860dfdfbe3 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -978,6 +978,8 @@ struct netns_ipvs { int sysctl_sync_sock_size; int sysctl_cache_bypass; int sysctl_expire_nodest_conn; + int sysctl_sloppy_tcp; + int sysctl_sloppy_sctp; int sysctl_expire_quiescent_template; int sysctl_sync_threshold[2]; unsigned int sysctl_sync_refresh_period; @@ -1020,6 +1022,8 @@ struct netns_ipvs { #define DEFAULT_SYNC_THRESHOLD 3 #define DEFAULT_SYNC_PERIOD 50 #define DEFAULT_SYNC_VER 1 +#define DEFAULT_SLOPPY_TCP 0 +#define DEFAULT_SLOPPY_SCTP 0 #define DEFAULT_SYNC_REFRESH_PERIOD (0U * HZ) #define DEFAULT_SYNC_RETRIES 0 #define IPVS_SYNC_WAKEUP_RATE 8 @@ -1056,6 +1060,16 @@ static inline int sysctl_sync_ver(struct netns_ipvs *ipvs) return ipvs->sysctl_sync_ver; } +static inline int sysctl_sloppy_tcp(struct netns_ipvs *ipvs) +{ + return ipvs->sysctl_sloppy_tcp; +} + +static inline int sysctl_sloppy_sctp(struct netns_ipvs *ipvs) +{ + return ipvs->sysctl_sloppy_sctp; +} + static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) { return ACCESS_ONCE(ipvs->sysctl_sync_ports); @@ -1109,6 +1123,16 @@ static inline int sysctl_sync_ver(struct netns_ipvs *ipvs) return DEFAULT_SYNC_VER; } +static inline int sysctl_sloppy_tcp(struct netns_ipvs *ipvs) +{ + return DEFAULT_SLOPPY_TCP; +} + +static inline int sysctl_sloppy_sctp(struct netns_ipvs *ipvs) +{ + return DEFAULT_SLOPPY_SCTP; +} + static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) { return 1; diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 47e510819f54..da035fc01eb2 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -1738,6 +1738,18 @@ static struct ctl_table vs_vars[] = { .mode = 0644, .proc_handler = proc_dointvec, }, + { + .procname = "sloppy_tcp", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "sloppy_sctp", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, { .procname = "expire_quiescent_template", .maxlen = sizeof(int), @@ -3723,6 +3735,8 @@ static int __net_init ip_vs_control_net_init_sysctl(struct net *net) tbl[idx++].data = &ipvs->sysctl_sync_sock_size; tbl[idx++].data = &ipvs->sysctl_cache_bypass; tbl[idx++].data = &ipvs->sysctl_expire_nodest_conn; + tbl[idx++].data = &ipvs->sysctl_sloppy_tcp; + tbl[idx++].data = &ipvs->sysctl_sloppy_sctp; tbl[idx++].data = &ipvs->sysctl_expire_quiescent_template; ipvs->sysctl_sync_threshold[0] = DEFAULT_SYNC_THRESHOLD; ipvs->sysctl_sync_threshold[1] = DEFAULT_SYNC_PERIOD; diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c index 86464881cd20..df29d6417043 100644 --- a/net/netfilter/ipvs/ip_vs_proto_sctp.c +++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c @@ -15,6 +15,7 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, { struct net *net; struct ip_vs_service *svc; + struct netns_ipvs *ipvs; sctp_chunkhdr_t _schunkh, *sch; sctp_sctphdr_t *sh, _sctph; @@ -27,13 +28,14 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, if (sch == NULL) return 0; net = skb_net(skb); + ipvs = net_ipvs(net); rcu_read_lock(); - if ((sch->type == SCTP_CID_INIT) && + if ((sch->type == SCTP_CID_INIT || sysctl_sloppy_sctp(ipvs)) && (svc = ip_vs_service_find(net, af, skb->mark, iph->protocol, &iph->daddr, sh->dest))) { int ignored; - if (ip_vs_todrop(net_ipvs(net))) { + if (ip_vs_todrop(ipvs)) { /* * It seems that we are very loaded. * We have to drop this packet :( @@ -232,21 +234,21 @@ static struct ipvs_sctp_nextstate * STATE : IP_VS_SCTP_S_NONE */ /*next state *//*event */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, + {{IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_DATA_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, + {IP_VS_SCTP_S_INIT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, + {IP_VS_SCTP_S_ECHO_CLI /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, + {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, + {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, + {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ }, diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index 50a15944c6c1..e3a697234a98 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -39,6 +39,7 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, struct net *net; struct ip_vs_service *svc; struct tcphdr _tcph, *th; + struct netns_ipvs *ipvs; th = skb_header_pointer(skb, iph->len, sizeof(_tcph), &_tcph); if (th == NULL) { @@ -46,14 +47,15 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, return 0; } net = skb_net(skb); + ipvs = net_ipvs(net); /* No !th->ack check to allow scheduling on SYN+ACK for Active FTP */ rcu_read_lock(); - if (th->syn && + if ((th->syn || sysctl_sloppy_tcp(ipvs)) && !th->rst && (svc = ip_vs_service_find(net, af, skb->mark, iph->protocol, &iph->daddr, th->dest))) { int ignored; - if (ip_vs_todrop(net_ipvs(net))) { + if (ip_vs_todrop(ipvs)) { /* * It seems that we are very loaded. * We have to drop this packet :( @@ -401,7 +403,7 @@ static struct tcp_states_t tcp_states [] = { /* sNO, sES, sSS, sSR, sFW, sTW, sCL, sCW, sLA, sLI, sSA */ /*syn*/ {{sSR, sES, sES, sSR, sSR, sSR, sSR, sSR, sSR, sSR, sSR }}, /*fin*/ {{sCL, sCW, sSS, sTW, sTW, sTW, sCL, sCW, sLA, sLI, sTW }}, -/*ack*/ {{sCL, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, +/*ack*/ {{sES, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, /*rst*/ {{sCL, sCL, sCL, sSR, sCL, sCL, sCL, sCL, sLA, sLI, sSR }}, /* OUTPUT */ @@ -415,7 +417,7 @@ static struct tcp_states_t tcp_states [] = { /* sNO, sES, sSS, sSR, sFW, sTW, sCL, sCW, sLA, sLI, sSA */ /*syn*/ {{sSR, sES, sES, sSR, sSR, sSR, sSR, sSR, sSR, sSR, sSR }}, /*fin*/ {{sCL, sFW, sSS, sTW, sFW, sTW, sCL, sCW, sLA, sLI, sTW }}, -/*ack*/ {{sCL, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, +/*ack*/ {{sES, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, /*rst*/ {{sCL, sCL, sCL, sSR, sCL, sCL, sCL, sCL, sLA, sLI, sCL }}, }; @@ -424,7 +426,7 @@ static struct tcp_states_t tcp_states_dos [] = { /* sNO, sES, sSS, sSR, sFW, sTW, sCL, sCW, sLA, sLI, sSA */ /*syn*/ {{sSR, sES, sES, sSR, sSR, sSR, sSR, sSR, sSR, sSR, sSA }}, /*fin*/ {{sCL, sCW, sSS, sTW, sTW, sTW, sCL, sCW, sLA, sLI, sSA }}, -/*ack*/ {{sCL, sES, sSS, sSR, sFW, sTW, sCL, sCW, sCL, sLI, sSA }}, +/*ack*/ {{sES, sES, sSS, sSR, sFW, sTW, sCL, sCW, sCL, sLI, sSA }}, /*rst*/ {{sCL, sCL, sCL, sSR, sCL, sCL, sCL, sCL, sLA, sLI, sCL }}, /* OUTPUT */ @@ -438,7 +440,7 @@ static struct tcp_states_t tcp_states_dos [] = { /* sNO, sES, sSS, sSR, sFW, sTW, sCL, sCW, sLA, sLI, sSA */ /*syn*/ {{sSA, sES, sES, sSR, sSA, sSA, sSA, sSA, sSA, sSA, sSA }}, /*fin*/ {{sCL, sFW, sSS, sTW, sFW, sTW, sCL, sCW, sLA, sLI, sTW }}, -/*ack*/ {{sCL, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, +/*ack*/ {{sES, sES, sSS, sES, sFW, sTW, sCL, sCW, sCL, sLI, sES }}, /*rst*/ {{sCL, sCL, sCL, sSR, sCL, sCL, sCL, sCL, sLA, sLI, sCL }}, }; From 61e7c420b4b2a797ac209106ba743ab6ebe984d8 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Tue, 18 Jun 2013 10:08:07 +0300 Subject: [PATCH 06/12] ipvs: replace the SCTP state machine Convert the SCTP state table, so that it is more readable. Change the states to be according to the diagram in RFC 2960 and add more states suitable for middle box. Still, such change in states adds incompatibility if systems in sync setup include this change and others do not include it. With this change we also have proper transitions in INPUT-ONLY mode (DR/TUN) where we see packets only from client. Now we should not switch to 10-second CLOSED state at a time when we should stay in ESTABLISHED state. The short names for states are because we have 16-char space in ipvsadm and 11-char limit for the connection list format. It is a sequence of the TCP implementation where the longest state name is ESTABLISHED. Signed-off-by: Julian Anastasov Signed-off-by: Simon Horman --- include/net/ip_vs.h | 21 +- net/netfilter/ipvs/ip_vs_proto_sctp.c | 858 +++++--------------------- net/netfilter/ipvs/ip_vs_sync.c | 7 +- 3 files changed, 170 insertions(+), 716 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 95860dfdfbe3..e667df171003 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -380,17 +380,18 @@ enum { */ enum ip_vs_sctp_states { IP_VS_SCTP_S_NONE, - IP_VS_SCTP_S_INIT_CLI, - IP_VS_SCTP_S_INIT_SER, - IP_VS_SCTP_S_INIT_ACK_CLI, - IP_VS_SCTP_S_INIT_ACK_SER, - IP_VS_SCTP_S_ECHO_CLI, - IP_VS_SCTP_S_ECHO_SER, + IP_VS_SCTP_S_INIT1, + IP_VS_SCTP_S_INIT, + IP_VS_SCTP_S_COOKIE_SENT, + IP_VS_SCTP_S_COOKIE_REPLIED, + IP_VS_SCTP_S_COOKIE_WAIT, + IP_VS_SCTP_S_COOKIE, + IP_VS_SCTP_S_COOKIE_ECHOED, IP_VS_SCTP_S_ESTABLISHED, - IP_VS_SCTP_S_SHUT_CLI, - IP_VS_SCTP_S_SHUT_SER, - IP_VS_SCTP_S_SHUT_ACK_CLI, - IP_VS_SCTP_S_SHUT_ACK_SER, + IP_VS_SCTP_S_SHUTDOWN_SENT, + IP_VS_SCTP_S_SHUTDOWN_RECEIVED, + IP_VS_SCTP_S_SHUTDOWN_ACK_SENT, + IP_VS_SCTP_S_REJECTED, IP_VS_SCTP_S_CLOSED, IP_VS_SCTP_S_LAST }; diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c index df29d6417043..3c0da8728036 100644 --- a/net/netfilter/ipvs/ip_vs_proto_sctp.c +++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c @@ -185,710 +185,159 @@ sctp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp) return 1; } -struct ipvs_sctp_nextstate { - int next_state; -}; enum ipvs_sctp_event_t { - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_DATA_SER, - IP_VS_SCTP_EVE_INIT_CLI, - IP_VS_SCTP_EVE_INIT_SER, - IP_VS_SCTP_EVE_INIT_ACK_CLI, - IP_VS_SCTP_EVE_INIT_ACK_SER, - IP_VS_SCTP_EVE_COOKIE_ECHO_CLI, - IP_VS_SCTP_EVE_COOKIE_ECHO_SER, - IP_VS_SCTP_EVE_COOKIE_ACK_CLI, - IP_VS_SCTP_EVE_COOKIE_ACK_SER, - IP_VS_SCTP_EVE_ABORT_CLI, - IP_VS_SCTP_EVE__ABORT_SER, - IP_VS_SCTP_EVE_SHUT_CLI, - IP_VS_SCTP_EVE_SHUT_SER, - IP_VS_SCTP_EVE_SHUT_ACK_CLI, - IP_VS_SCTP_EVE_SHUT_ACK_SER, - IP_VS_SCTP_EVE_SHUT_COM_CLI, - IP_VS_SCTP_EVE_SHUT_COM_SER, - IP_VS_SCTP_EVE_LAST + IP_VS_SCTP_DATA = 0, /* DATA, SACK, HEARTBEATs */ + IP_VS_SCTP_INIT, + IP_VS_SCTP_INIT_ACK, + IP_VS_SCTP_COOKIE_ECHO, + IP_VS_SCTP_COOKIE_ACK, + IP_VS_SCTP_SHUTDOWN, + IP_VS_SCTP_SHUTDOWN_ACK, + IP_VS_SCTP_SHUTDOWN_COMPLETE, + IP_VS_SCTP_ERROR, + IP_VS_SCTP_ABORT, + IP_VS_SCTP_EVENT_LAST }; -static enum ipvs_sctp_event_t sctp_events[256] = { - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_INIT_CLI, - IP_VS_SCTP_EVE_INIT_ACK_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_ABORT_CLI, - IP_VS_SCTP_EVE_SHUT_CLI, - IP_VS_SCTP_EVE_SHUT_ACK_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_COOKIE_ECHO_CLI, - IP_VS_SCTP_EVE_COOKIE_ACK_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_DATA_CLI, - IP_VS_SCTP_EVE_SHUT_COM_CLI, +/* RFC 2960, 3.2 Chunk Field Descriptions */ +static __u8 sctp_events[] = { + [SCTP_CID_DATA] = IP_VS_SCTP_DATA, + [SCTP_CID_INIT] = IP_VS_SCTP_INIT, + [SCTP_CID_INIT_ACK] = IP_VS_SCTP_INIT_ACK, + [SCTP_CID_SACK] = IP_VS_SCTP_DATA, + [SCTP_CID_HEARTBEAT] = IP_VS_SCTP_DATA, + [SCTP_CID_HEARTBEAT_ACK] = IP_VS_SCTP_DATA, + [SCTP_CID_ABORT] = IP_VS_SCTP_ABORT, + [SCTP_CID_SHUTDOWN] = IP_VS_SCTP_SHUTDOWN, + [SCTP_CID_SHUTDOWN_ACK] = IP_VS_SCTP_SHUTDOWN_ACK, + [SCTP_CID_ERROR] = IP_VS_SCTP_ERROR, + [SCTP_CID_COOKIE_ECHO] = IP_VS_SCTP_COOKIE_ECHO, + [SCTP_CID_COOKIE_ACK] = IP_VS_SCTP_COOKIE_ACK, + [SCTP_CID_ECN_ECNE] = IP_VS_SCTP_DATA, + [SCTP_CID_ECN_CWR] = IP_VS_SCTP_DATA, + [SCTP_CID_SHUTDOWN_COMPLETE] = IP_VS_SCTP_SHUTDOWN_COMPLETE, }; -static struct ipvs_sctp_nextstate - sctp_states_table[IP_VS_SCTP_S_LAST][IP_VS_SCTP_EVE_LAST] = { - /* - * STATE : IP_VS_SCTP_S_NONE - */ - /*next state *//*event */ - {{IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - {IP_VS_SCTP_S_INIT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - {IP_VS_SCTP_S_ECHO_CLI /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ }, - }, - /* - * STATE : IP_VS_SCTP_S_INIT_CLI - * Cient sent INIT and is waiting for reply from server(In ECHO_WAIT) - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_INIT_ACK_SER /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ECHO_CLI */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_ECHO_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_INIT_SER - * Server sent INIT and waiting for INIT ACK from the client - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - {IP_VS_SCTP_S_INIT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_INIT_ACK_CLI - * Client sent INIT ACK and waiting for ECHO from the server - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK has been resent by the client, let us stay is in - * the same state - */ - {IP_VS_SCTP_S_INIT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - /* - * INIT_ACK sent by the server, close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * ECHO by client, it should not happen, close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - /* - * ECHO by server, this is what we are expecting, move to ECHO_SER - */ - {IP_VS_SCTP_S_ECHO_SER /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, it should not happen, close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - /* - * Unexpected COOKIE ACK from server, staty in the same state - */ - {IP_VS_SCTP_S_INIT_ACK_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_INIT_ACK_SER - * Server sent INIT ACK and waiting for ECHO from the client - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * Unexpected INIT_ACK by the client, let us close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - /* - * INIT_ACK resent by the server, let us move to same state - */ - {IP_VS_SCTP_S_INIT_ACK_SER /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client send the ECHO, this is what we are expecting, - * move to ECHO_CLI - */ - {IP_VS_SCTP_S_ECHO_CLI /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - /* - * ECHO received from the server, Not sure what to do, - * let us close it - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, let us stay in the same state - */ - {IP_VS_SCTP_S_INIT_ACK_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - /* - * COOKIE ACK from server, hmm... this should not happen, lets close - * the connection. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_ECHO_CLI - * Cient sent ECHO and waiting COOKEI ACK from the Server - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK has been by the client, let us close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_ECHO_CLI /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client resent the ECHO, let us stay in the same state - */ - {IP_VS_SCTP_S_ECHO_CLI /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - /* - * ECHO received from the server, Not sure what to do, - * let us close it - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, this shoud not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - /* - * COOKIE ACK from server, this is what we are awaiting,lets move to - * ESTABLISHED. - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_ECHO_SER - * Server sent ECHO and waiting COOKEI ACK from the client - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_ECHO_SER /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - /* - * INIT_ACK has been by the server, let us close the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent the ECHO, not sure what to do, let's close the - * connection. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - /* - * ECHO resent by the server, stay in the same state - */ - {IP_VS_SCTP_S_ECHO_SER /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, this is what we are expecting, let's move - * to ESTABLISHED. - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - /* - * COOKIE ACK from server, this should not happen, lets close the - * connection. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_ESTABLISHED - * Association established - */ - {{IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent ECHO, Spec(sec 5.2.4) says it may be handled by the - * peer and peer shall move to the ESTABISHED. if it doesn't handle - * it will send ERROR chunk. So, stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, not sure what to do stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - /* - * SHUTDOWN from the client, move to SHUDDOWN_CLI - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - /* - * SHUTDOWN from the server, move to SHUTDOWN_SER - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_SHUT_SER */ }, - /* - * client sent SHUDTDOWN_ACK, this should not happen, let's close - * the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_SHUT_CLI - * SHUTDOWN sent from the client, waitinf for SHUT ACK from the server - */ - /* - * We received the data chuck, keep the state unchanged. I assume - * that still data chuncks can be received by both the peers in - * SHUDOWN state - */ - - {{IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent ECHO, Spec(sec 5.2.4) says it may be handled by the - * peer and peer shall move to the ESTABISHED. if it doesn't handle - * it will send ERROR chunk. So, stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, not sure what to do stay in the same state - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - /* - * SHUTDOWN resent from the client, move to SHUDDOWN_CLI - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - /* - * SHUTDOWN from the server, move to SHUTDOWN_SER - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_SHUT_SER */ }, - /* - * client sent SHUDTDOWN_ACK, this should not happen, let's close - * the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - /* - * Server sent SHUTDOWN ACK, this is what we are expecting, let's move - * to SHUDOWN_ACK_SER - */ - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - /* - * SHUTDOWN COM from client, this should not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_SHUT_SER - * SHUTDOWN sent from the server, waitinf for SHUTDOWN ACK from client - */ - /* - * We received the data chuck, keep the state unchanged. I assume - * that still data chuncks can be received by both the peers in - * SHUDOWN state - */ - - {{IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent ECHO, Spec(sec 5.2.4) says it may be handled by the - * peer and peer shall move to the ESTABISHED. if it doesn't handle - * it will send ERROR chunk. So, stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, not sure what to do stay in the same state - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - /* - * SHUTDOWN resent from the client, move to SHUDDOWN_CLI - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - /* - * SHUTDOWN resent from the server, move to SHUTDOWN_SER - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_SHUT_SER */ }, - /* - * client sent SHUDTDOWN_ACK, this is what we are expecting, let's - * move to SHUT_ACK_CLI - */ - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - /* - * Server sent SHUTDOWN ACK, this should not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - /* - * SHUTDOWN COM from client, this should not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - - /* - * State : IP_VS_SCTP_S_SHUT_ACK_CLI - * SHUTDOWN ACK from the client, awaiting for SHUTDOWN COM from server - */ - /* - * We received the data chuck, keep the state unchanged. I assume - * that still data chuncks can be received by both the peers in - * SHUDOWN state - */ - - {{IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent ECHO, Spec(sec 5.2.4) says it may be handled by the - * peer and peer shall move to the ESTABISHED. if it doesn't handle - * it will send ERROR chunk. So, stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, not sure what to do stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - /* - * SHUTDOWN sent from the client, move to SHUDDOWN_CLI - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - /* - * SHUTDOWN sent from the server, move to SHUTDOWN_SER - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_SHUT_SER */ }, - /* - * client resent SHUDTDOWN_ACK, let's stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_CLI /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - /* - * Server sent SHUTDOWN ACK, this should not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - /* - * SHUTDOWN COM from client, this should not happen, let's close the - * connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - /* - * SHUTDOWN COMPLETE from server this is what we are expecting. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - - /* - * State : IP_VS_SCTP_S_SHUT_ACK_SER - * SHUTDOWN ACK from the server, awaiting for SHUTDOWN COM from client - */ - /* - * We received the data chuck, keep the state unchanged. I assume - * that still data chuncks can be received by both the peers in - * SHUDOWN state - */ - - {{IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_DATA_SER */ }, - /* - * We have got an INIT from client. From the spec.“Upon receipt of - * an INIT in the COOKIE-WAIT state, an endpoint MUST respond with - * an INIT ACK using the same parameters it sent in its original - * INIT chunk (including its Initiate Tag, unchanged”). - */ - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - /* - * INIT_ACK sent by the server, Unexpected INIT ACK, spec says, - * “If an INIT ACK is received by an endpoint in any state other - * than the COOKIE-WAIT state, the endpoint should discard the - * INIT ACK chunk”. Stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - /* - * Client sent ECHO, Spec(sec 5.2.4) says it may be handled by the - * peer and peer shall move to the ESTABISHED. if it doesn't handle - * it will send ERROR chunk. So, stay in the same state - */ - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_ESTABLISHED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - /* - * COOKIE ACK from client, not sure what to do stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - /* - * SHUTDOWN sent from the client, move to SHUDDOWN_CLI - */ - {IP_VS_SCTP_S_SHUT_CLI /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - /* - * SHUTDOWN sent from the server, move to SHUTDOWN_SER - */ - {IP_VS_SCTP_S_SHUT_SER /* IP_VS_SCTP_EVE_SHUT_SER */ }, - /* - * client sent SHUDTDOWN_ACK, this should not happen let's close - * the connection. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - /* - * Server resent SHUTDOWN ACK, stay in the same state - */ - {IP_VS_SCTP_S_SHUT_ACK_SER /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - /* - * SHUTDOWN COM from client, this what we are expecting, let's close - * the connection - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - /* - * SHUTDOWN COMPLETE from server this should not happen. - */ - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - }, - /* - * State : IP_VS_SCTP_S_CLOSED - */ - {{IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_DATA_SER */ }, - {IP_VS_SCTP_S_INIT_CLI /* IP_VS_SCTP_EVE_INIT_CLI */ }, - {IP_VS_SCTP_S_INIT_SER /* IP_VS_SCTP_EVE_INIT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_INIT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ECHO_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_COOKIE_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_ABORT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_ACK_SER */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_CLI */ }, - {IP_VS_SCTP_S_CLOSED /* IP_VS_SCTP_EVE_SHUT_COM_SER */ } - } -}; - -/* - * Timeout table[state] +/* SCTP States: + * See RFC 2960, 4. SCTP Association State Diagram + * + * New states (not in diagram): + * - INIT1 state: use shorter timeout for dropped INIT packets + * - REJECTED state: use shorter timeout if INIT is rejected with ABORT + * - INIT, COOKIE_SENT, COOKIE_REPLIED, COOKIE states: for better debugging + * + * The states are as seen in real server. In the diagram, INIT1, INIT, + * COOKIE_SENT and COOKIE_REPLIED processing happens in CLOSED state. + * + * States as per packets from client (C) and server (S): + * + * Setup of client connection: + * IP_VS_SCTP_S_INIT1: First C:INIT sent, wait for S:INIT-ACK + * IP_VS_SCTP_S_INIT: Next C:INIT sent, wait for S:INIT-ACK + * IP_VS_SCTP_S_COOKIE_SENT: S:INIT-ACK sent, wait for C:COOKIE-ECHO + * IP_VS_SCTP_S_COOKIE_REPLIED: C:COOKIE-ECHO sent, wait for S:COOKIE-ACK + * + * Setup of server connection: + * IP_VS_SCTP_S_COOKIE_WAIT: S:INIT sent, wait for C:INIT-ACK + * IP_VS_SCTP_S_COOKIE: C:INIT-ACK sent, wait for S:COOKIE-ECHO + * IP_VS_SCTP_S_COOKIE_ECHOED: S:COOKIE-ECHO sent, wait for C:COOKIE-ACK */ + +#define sNO IP_VS_SCTP_S_NONE +#define sI1 IP_VS_SCTP_S_INIT1 +#define sIN IP_VS_SCTP_S_INIT +#define sCS IP_VS_SCTP_S_COOKIE_SENT +#define sCR IP_VS_SCTP_S_COOKIE_REPLIED +#define sCW IP_VS_SCTP_S_COOKIE_WAIT +#define sCO IP_VS_SCTP_S_COOKIE +#define sCE IP_VS_SCTP_S_COOKIE_ECHOED +#define sES IP_VS_SCTP_S_ESTABLISHED +#define sSS IP_VS_SCTP_S_SHUTDOWN_SENT +#define sSR IP_VS_SCTP_S_SHUTDOWN_RECEIVED +#define sSA IP_VS_SCTP_S_SHUTDOWN_ACK_SENT +#define sRJ IP_VS_SCTP_S_REJECTED +#define sCL IP_VS_SCTP_S_CLOSED + +static const __u8 sctp_states + [IP_VS_DIR_LAST][IP_VS_SCTP_EVENT_LAST][IP_VS_SCTP_S_LAST] = { + { /* INPUT */ +/* sNO, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL*/ +/* d */{sES, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* i */{sI1, sIN, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sIN, sIN}, +/* i_a */{sCW, sCW, sCW, sCS, sCR, sCO, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_e */{sCR, sIN, sIN, sCR, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_a */{sES, sI1, sIN, sCS, sCR, sCW, sCO, sES, sES, sSS, sSR, sSA, sRJ, sCL}, +/* s */{sSR, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sSR, sSS, sSR, sSA, sRJ, sCL}, +/* s_a */{sCL, sIN, sIN, sCS, sCR, sCW, sCO, sCE, sES, sCL, sSR, sCL, sRJ, sCL}, +/* s_c */{sCL, sCL, sCL, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sCL, sRJ, sCL}, +/* err */{sCL, sI1, sIN, sCS, sCR, sCW, sCO, sCL, sES, sSS, sSR, sSA, sRJ, sCL}, +/* ab */{sCL, sCL, sCL, sCL, sCL, sRJ, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL}, + }, + { /* OUTPUT */ +/* sNO, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL*/ +/* d */{sES, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* i */{sCW, sCW, sCW, sCW, sCW, sCW, sCW, sCW, sES, sCW, sCW, sCW, sCW, sCW}, +/* i_a */{sCS, sCS, sCS, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_e */{sCE, sCE, sCE, sCE, sCE, sCE, sCE, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_a */{sES, sES, sES, sES, sES, sES, sES, sES, sES, sSS, sSR, sSA, sRJ, sCL}, +/* s */{sSS, sSS, sSS, sSS, sSS, sSS, sSS, sSS, sSS, sSS, sSR, sSA, sRJ, sCL}, +/* s_a */{sSA, sSA, sSA, sSA, sSA, sCW, sCO, sCE, sES, sSA, sSA, sSA, sRJ, sCL}, +/* s_c */{sCL, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* err */{sCL, sCL, sCL, sCL, sCL, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* ab */{sCL, sRJ, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL}, + }, + { /* INPUT-ONLY */ +/* sNO, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL*/ +/* d */{sES, sI1, sIN, sCS, sCR, sES, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* i */{sI1, sIN, sIN, sIN, sIN, sIN, sCO, sCE, sES, sSS, sSR, sSA, sIN, sIN}, +/* i_a */{sCE, sCE, sCE, sCE, sCE, sCE, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_e */{sES, sES, sES, sES, sES, sES, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* c_a */{sES, sI1, sIN, sES, sES, sCW, sES, sES, sES, sSS, sSR, sSA, sRJ, sCL}, +/* s */{sSR, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sSR, sSS, sSR, sSA, sRJ, sCL}, +/* s_a */{sCL, sIN, sIN, sCS, sCR, sCW, sCO, sCE, sCL, sCL, sSR, sCL, sRJ, sCL}, +/* s_c */{sCL, sCL, sCL, sCL, sCL, sCW, sCO, sCE, sES, sSS, sCL, sCL, sRJ, sCL}, +/* err */{sCL, sI1, sIN, sCS, sCR, sCW, sCO, sCE, sES, sSS, sSR, sSA, sRJ, sCL}, +/* ab */{sCL, sCL, sCL, sCL, sCL, sRJ, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL}, + }, +}; + +#define IP_VS_SCTP_MAX_RTO ((60 + 1) * HZ) + +/* Timeout table[state] */ static const int sctp_timeouts[IP_VS_SCTP_S_LAST + 1] = { - [IP_VS_SCTP_S_NONE] = 2 * HZ, - [IP_VS_SCTP_S_INIT_CLI] = 1 * 60 * HZ, - [IP_VS_SCTP_S_INIT_SER] = 1 * 60 * HZ, - [IP_VS_SCTP_S_INIT_ACK_CLI] = 1 * 60 * HZ, - [IP_VS_SCTP_S_INIT_ACK_SER] = 1 * 60 * HZ, - [IP_VS_SCTP_S_ECHO_CLI] = 1 * 60 * HZ, - [IP_VS_SCTP_S_ECHO_SER] = 1 * 60 * HZ, - [IP_VS_SCTP_S_ESTABLISHED] = 15 * 60 * HZ, - [IP_VS_SCTP_S_SHUT_CLI] = 1 * 60 * HZ, - [IP_VS_SCTP_S_SHUT_SER] = 1 * 60 * HZ, - [IP_VS_SCTP_S_SHUT_ACK_CLI] = 1 * 60 * HZ, - [IP_VS_SCTP_S_SHUT_ACK_SER] = 1 * 60 * HZ, - [IP_VS_SCTP_S_CLOSED] = 10 * HZ, - [IP_VS_SCTP_S_LAST] = 2 * HZ, + [IP_VS_SCTP_S_NONE] = 2 * HZ, + [IP_VS_SCTP_S_INIT1] = (0 + 3 + 1) * HZ, + [IP_VS_SCTP_S_INIT] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_COOKIE_SENT] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_COOKIE_REPLIED] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_COOKIE_WAIT] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_COOKIE] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_COOKIE_ECHOED] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_ESTABLISHED] = 15 * 60 * HZ, + [IP_VS_SCTP_S_SHUTDOWN_SENT] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_SHUTDOWN_RECEIVED] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_SHUTDOWN_ACK_SENT] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_REJECTED] = (0 + 3 + 1) * HZ, + [IP_VS_SCTP_S_CLOSED] = IP_VS_SCTP_MAX_RTO, + [IP_VS_SCTP_S_LAST] = 2 * HZ, }; static const char *sctp_state_name_table[IP_VS_SCTP_S_LAST + 1] = { - [IP_VS_SCTP_S_NONE] = "NONE", - [IP_VS_SCTP_S_INIT_CLI] = "INIT_CLI", - [IP_VS_SCTP_S_INIT_SER] = "INIT_SER", - [IP_VS_SCTP_S_INIT_ACK_CLI] = "INIT_ACK_CLI", - [IP_VS_SCTP_S_INIT_ACK_SER] = "INIT_ACK_SER", - [IP_VS_SCTP_S_ECHO_CLI] = "COOKIE_ECHO_CLI", - [IP_VS_SCTP_S_ECHO_SER] = "COOKIE_ECHO_SER", - [IP_VS_SCTP_S_ESTABLISHED] = "ESTABISHED", - [IP_VS_SCTP_S_SHUT_CLI] = "SHUTDOWN_CLI", - [IP_VS_SCTP_S_SHUT_SER] = "SHUTDOWN_SER", - [IP_VS_SCTP_S_SHUT_ACK_CLI] = "SHUTDOWN_ACK_CLI", - [IP_VS_SCTP_S_SHUT_ACK_SER] = "SHUTDOWN_ACK_SER", - [IP_VS_SCTP_S_CLOSED] = "CLOSED", - [IP_VS_SCTP_S_LAST] = "BUG!" + [IP_VS_SCTP_S_NONE] = "NONE", + [IP_VS_SCTP_S_INIT1] = "INIT1", + [IP_VS_SCTP_S_INIT] = "INIT", + [IP_VS_SCTP_S_COOKIE_SENT] = "C-SENT", + [IP_VS_SCTP_S_COOKIE_REPLIED] = "C-REPLIED", + [IP_VS_SCTP_S_COOKIE_WAIT] = "C-WAIT", + [IP_VS_SCTP_S_COOKIE] = "COOKIE", + [IP_VS_SCTP_S_COOKIE_ECHOED] = "C-ECHOED", + [IP_VS_SCTP_S_ESTABLISHED] = "ESTABLISHED", + [IP_VS_SCTP_S_SHUTDOWN_SENT] = "S-SENT", + [IP_VS_SCTP_S_SHUTDOWN_RECEIVED] = "S-RECEIVED", + [IP_VS_SCTP_S_SHUTDOWN_ACK_SENT] = "S-ACK-SENT", + [IP_VS_SCTP_S_REJECTED] = "REJECTED", + [IP_VS_SCTP_S_CLOSED] = "CLOSED", + [IP_VS_SCTP_S_LAST] = "BUG!", }; @@ -945,17 +394,20 @@ set_sctp_state(struct ip_vs_proto_data *pd, struct ip_vs_conn *cp, } } - event = sctp_events[chunk_type]; + event = (chunk_type < sizeof(sctp_events)) ? + sctp_events[chunk_type] : IP_VS_SCTP_DATA; - /* - * If the direction is IP_VS_DIR_OUTPUT, this event is from server + /* Update direction to INPUT_ONLY if necessary + * or delete NO_OUTPUT flag if output packet detected */ - if (direction == IP_VS_DIR_OUTPUT) - event++; - /* - * get next state - */ - next_state = sctp_states_table[cp->state][event].next_state; + if (cp->flags & IP_VS_CONN_F_NOOUTPUT) { + if (direction == IP_VS_DIR_OUTPUT) + cp->flags &= ~IP_VS_CONN_F_NOOUTPUT; + else + direction = IP_VS_DIR_INPUT_ONLY; + } + + next_state = sctp_states[direction][event][cp->state]; if (next_state != cp->state) { struct ip_vs_dest *dest = cp->dest; diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index f6046d9af8d3..2fc66394d86d 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -461,9 +461,10 @@ static int ip_vs_sync_conn_needed(struct netns_ipvs *ipvs, } else if (unlikely(cp->protocol == IPPROTO_SCTP)) { if (!((1 << cp->state) & ((1 << IP_VS_SCTP_S_ESTABLISHED) | - (1 << IP_VS_SCTP_S_CLOSED) | - (1 << IP_VS_SCTP_S_SHUT_ACK_CLI) | - (1 << IP_VS_SCTP_S_SHUT_ACK_SER)))) + (1 << IP_VS_SCTP_S_SHUTDOWN_SENT) | + (1 << IP_VS_SCTP_S_SHUTDOWN_RECEIVED) | + (1 << IP_VS_SCTP_S_SHUTDOWN_ACK_SENT) | + (1 << IP_VS_SCTP_S_CLOSED)))) return 0; force = cp->state != cp->old_state; if (force && cp->state != IP_VS_SCTP_S_ESTABLISHED) From acaac5d8bbedf6bd96f53960780942e1ad90d70e Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Tue, 18 Jun 2013 10:08:08 +0300 Subject: [PATCH 07/12] ipvs: drop SCTP connections depending on state Drop SCTP connections under load (dropentry context) depending on the protocol state, just like for TCP: INIT conns are dropped immediately, established are dropped randomly while connections in progress or shutdown are skipped. Signed-off-by: Julian Anastasov Signed-off-by: Simon Horman --- net/netfilter/ipvs/ip_vs_conn.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index c8c52a98590b..4c8e5c0aa1ab 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -1231,6 +1231,18 @@ void ip_vs_random_dropentry(struct net *net) default: continue; } + } else if (cp->protocol == IPPROTO_SCTP) { + switch (cp->state) { + case IP_VS_SCTP_S_INIT1: + case IP_VS_SCTP_S_INIT: + break; + case IP_VS_SCTP_S_ESTABLISHED: + if (todrop_entry(cp)) + break; + continue; + default: + continue; + } } else { if (!todrop_entry(cp)) continue; From eba3b5a78799d21dea05118b294524958f0ab592 Mon Sep 17 00:00:00 2001 From: Alexander Frolkin Date: Wed, 19 Jun 2013 10:54:25 +0100 Subject: [PATCH 08/12] ipvs: SH fallback and L4 hashing By default the SH scheduler rejects connections that are hashed onto a realserver of weight 0. This patch adds a flag to make SH choose a different realserver in this case, instead of rejecting the connection. The patch also adds a flag to make SH include the source port (TCP, UDP, SCTP) in the hash as well as the source address. This basically allows for deterministic round-robin load balancing (i.e., where any director in a cluster of directors with identical config will send the same packet the same way). The flags are service flags (IP_VS_SVC_F_SCHED*) so that these options can be set per service. They are set using a new option to ipvsadm. Signed-off-by: Alexander Frolkin Acked-by: Julian Anastasov Signed-off-by: Simon Horman --- include/uapi/linux/ip_vs.h | 6 ++ net/netfilter/ipvs/ip_vs_sh.c | 100 +++++++++++++++++++++++++++++----- 2 files changed, 91 insertions(+), 15 deletions(-) diff --git a/include/uapi/linux/ip_vs.h b/include/uapi/linux/ip_vs.h index a24537725e80..29458223d044 100644 --- a/include/uapi/linux/ip_vs.h +++ b/include/uapi/linux/ip_vs.h @@ -20,6 +20,12 @@ #define IP_VS_SVC_F_PERSISTENT 0x0001 /* persistent port */ #define IP_VS_SVC_F_HASHED 0x0002 /* hashed entry */ #define IP_VS_SVC_F_ONEPACKET 0x0004 /* one-packet scheduling */ +#define IP_VS_SVC_F_SCHED1 0x0008 /* scheduler flag 1 */ +#define IP_VS_SVC_F_SCHED2 0x0010 /* scheduler flag 2 */ +#define IP_VS_SVC_F_SCHED3 0x0020 /* scheduler flag 3 */ + +#define IP_VS_SVC_F_SCHED_SH_FALLBACK IP_VS_SVC_F_SCHED1 /* SH fallback */ +#define IP_VS_SVC_F_SCHED_SH_PORT IP_VS_SVC_F_SCHED2 /* SH use port */ /* * Destination Server Flags diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c index e0d5d1653566..f16c027df15b 100644 --- a/net/netfilter/ipvs/ip_vs_sh.c +++ b/net/netfilter/ipvs/ip_vs_sh.c @@ -48,6 +48,10 @@ #include +#include +#include +#include + /* * IPVS SH bucket @@ -71,10 +75,19 @@ struct ip_vs_sh_state { struct ip_vs_sh_bucket buckets[IP_VS_SH_TAB_SIZE]; }; +/* Helper function to determine if server is unavailable */ +static inline bool is_unavailable(struct ip_vs_dest *dest) +{ + return atomic_read(&dest->weight) <= 0 || + dest->flags & IP_VS_DEST_F_OVERLOAD; +} + /* * Returns hash value for IPVS SH entry */ -static inline unsigned int ip_vs_sh_hashkey(int af, const union nf_inet_addr *addr) +static inline unsigned int +ip_vs_sh_hashkey(int af, const union nf_inet_addr *addr, + __be16 port, unsigned int offset) { __be32 addr_fold = addr->ip; @@ -83,7 +96,8 @@ static inline unsigned int ip_vs_sh_hashkey(int af, const union nf_inet_addr *ad addr_fold = addr->ip6[0]^addr->ip6[1]^ addr->ip6[2]^addr->ip6[3]; #endif - return (ntohl(addr_fold)*2654435761UL) & IP_VS_SH_TAB_MASK; + return (offset + (ntohs(port) + ntohl(addr_fold))*2654435761UL) & + IP_VS_SH_TAB_MASK; } @@ -91,12 +105,42 @@ static inline unsigned int ip_vs_sh_hashkey(int af, const union nf_inet_addr *ad * Get ip_vs_dest associated with supplied parameters. */ static inline struct ip_vs_dest * -ip_vs_sh_get(int af, struct ip_vs_sh_state *s, const union nf_inet_addr *addr) +ip_vs_sh_get(struct ip_vs_service *svc, struct ip_vs_sh_state *s, + const union nf_inet_addr *addr, __be16 port) { - return rcu_dereference(s->buckets[ip_vs_sh_hashkey(af, addr)].dest); + unsigned int hash = ip_vs_sh_hashkey(svc->af, addr, port, 0); + struct ip_vs_dest *dest = rcu_dereference(s->buckets[hash].dest); + + return (!dest || is_unavailable(dest)) ? NULL : dest; } +/* As ip_vs_sh_get, but with fallback if selected server is unavailable */ +static inline struct ip_vs_dest * +ip_vs_sh_get_fallback(struct ip_vs_service *svc, struct ip_vs_sh_state *s, + const union nf_inet_addr *addr, __be16 port) +{ + unsigned int offset; + unsigned int hash; + struct ip_vs_dest *dest; + + for (offset = 0; offset < IP_VS_SH_TAB_SIZE; offset++) { + hash = ip_vs_sh_hashkey(svc->af, addr, port, offset); + dest = rcu_dereference(s->buckets[hash].dest); + if (!dest) + break; + if (is_unavailable(dest)) + IP_VS_DBG_BUF(6, "SH: selected unavailable server " + "%s:%d (offset %d)", + IP_VS_DBG_ADDR(svc->af, &dest->addr), + ntohs(dest->port), offset); + else + return dest; + } + + return NULL; +} + /* * Assign all the hash buckets of the specified table with the service. */ @@ -213,13 +257,33 @@ static int ip_vs_sh_dest_changed(struct ip_vs_service *svc, } -/* - * If the dest flags is set with IP_VS_DEST_F_OVERLOAD, - * consider that the server is overloaded here. - */ -static inline int is_overloaded(struct ip_vs_dest *dest) +/* Helper function to get port number */ +static inline __be16 +ip_vs_sh_get_port(const struct sk_buff *skb, struct ip_vs_iphdr *iph) { - return dest->flags & IP_VS_DEST_F_OVERLOAD; + __be16 port; + struct tcphdr _tcph, *th; + struct udphdr _udph, *uh; + sctp_sctphdr_t _sctph, *sh; + + switch (iph->protocol) { + case IPPROTO_TCP: + th = skb_header_pointer(skb, iph->len, sizeof(_tcph), &_tcph); + port = th->source; + break; + case IPPROTO_UDP: + uh = skb_header_pointer(skb, iph->len, sizeof(_udph), &_udph); + port = uh->source; + break; + case IPPROTO_SCTP: + sh = skb_header_pointer(skb, iph->len, sizeof(_sctph), &_sctph); + port = sh->source; + break; + default: + port = 0; + } + + return port; } @@ -232,15 +296,21 @@ ip_vs_sh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb, { struct ip_vs_dest *dest; struct ip_vs_sh_state *s; + __be16 port = 0; IP_VS_DBG(6, "ip_vs_sh_schedule(): Scheduling...\n"); + if (svc->flags & IP_VS_SVC_F_SCHED_SH_PORT) + port = ip_vs_sh_get_port(skb, iph); + s = (struct ip_vs_sh_state *) svc->sched_data; - dest = ip_vs_sh_get(svc->af, s, &iph->saddr); - if (!dest - || !(dest->flags & IP_VS_DEST_F_AVAILABLE) - || atomic_read(&dest->weight) <= 0 - || is_overloaded(dest)) { + + if (svc->flags & IP_VS_SVC_F_SCHED_SH_FALLBACK) + dest = ip_vs_sh_get_fallback(svc, s, &iph->saddr, port); + else + dest = ip_vs_sh_get(svc, s, &iph->saddr, port); + + if (!dest) { ip_vs_scheduler_err(svc, "no destination available"); return NULL; } From 4d0c875dcc4923476f364e83912d134da2df224c Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Mon, 24 Jun 2013 22:44:41 +0300 Subject: [PATCH 09/12] ipvs: add sync_persist_mode flag Add sync_persist_mode flag to reduce sync traffic by syncing only persistent templates. Signed-off-by: Julian Anastasov Tested-by: Aleksey Chudov Signed-off-by: Simon Horman --- Documentation/networking/ipvs-sysctl.txt | 13 +++++++++++++ include/net/ip_vs.h | 11 +++++++++++ net/netfilter/ipvs/ip_vs_ctl.c | 7 +++++++ net/netfilter/ipvs/ip_vs_sync.c | 12 ++++++++++++ 4 files changed, 43 insertions(+) diff --git a/Documentation/networking/ipvs-sysctl.txt b/Documentation/networking/ipvs-sysctl.txt index 9573d0c48c6e..7a3c04729591 100644 --- a/Documentation/networking/ipvs-sysctl.txt +++ b/Documentation/networking/ipvs-sysctl.txt @@ -181,6 +181,19 @@ snat_reroute - BOOLEAN always be the same as the original route so it is an optimisation to disable snat_reroute and avoid the recalculation. +sync_persist_mode - INTEGER + default 0 + + Controls the synchronisation of connections when using persistence + + 0: All types of connections are synchronised + 1: Attempt to reduce the synchronisation traffic depending on + the connection type. For persistent services avoid synchronisation + for normal connections, do it only for persistence templates. + In such case, for TCP and SCTP it may need enabling sloppy_tcp and + sloppy_sctp flags on backup servers. For non-persistent services + such optimization is not applied, mode 0 is assumed. + sync_version - INTEGER default 1 diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index e667df171003..f0d70f066f3d 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -975,6 +975,7 @@ struct netns_ipvs { int sysctl_snat_reroute; int sysctl_sync_ver; int sysctl_sync_ports; + int sysctl_sync_persist_mode; unsigned long sysctl_sync_qlen_max; int sysctl_sync_sock_size; int sysctl_cache_bypass; @@ -1076,6 +1077,11 @@ static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) return ACCESS_ONCE(ipvs->sysctl_sync_ports); } +static inline int sysctl_sync_persist_mode(struct netns_ipvs *ipvs) +{ + return ipvs->sysctl_sync_persist_mode; +} + static inline unsigned long sysctl_sync_qlen_max(struct netns_ipvs *ipvs) { return ipvs->sysctl_sync_qlen_max; @@ -1139,6 +1145,11 @@ static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) return 1; } +static inline int sysctl_sync_persist_mode(struct netns_ipvs *ipvs) +{ + return 0; +} + static inline unsigned long sysctl_sync_qlen_max(struct netns_ipvs *ipvs) { return IPVS_SYNC_QLEN_MAX; diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index da035fc01eb2..c8148e487386 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -1714,6 +1714,12 @@ static struct ctl_table vs_vars[] = { .mode = 0644, .proc_handler = &proc_do_sync_ports, }, + { + .procname = "sync_persist_mode", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, { .procname = "sync_qlen_max", .maxlen = sizeof(unsigned long), @@ -3729,6 +3735,7 @@ static int __net_init ip_vs_control_net_init_sysctl(struct net *net) tbl[idx++].data = &ipvs->sysctl_sync_ver; ipvs->sysctl_sync_ports = 1; tbl[idx++].data = &ipvs->sysctl_sync_ports; + tbl[idx++].data = &ipvs->sysctl_sync_persist_mode; ipvs->sysctl_sync_qlen_max = nr_free_buffer_pages() / 32; tbl[idx++].data = &ipvs->sysctl_sync_qlen_max; ipvs->sysctl_sync_sock_size = 0; diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 2fc66394d86d..f4484719f3e6 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -425,6 +425,16 @@ ip_vs_sync_buff_create_v0(struct netns_ipvs *ipvs) return sb; } +/* Check if connection is controlled by persistence */ +static inline bool in_persistence(struct ip_vs_conn *cp) +{ + for (cp = cp->control; cp; cp = cp->control) { + if (cp->flags & IP_VS_CONN_F_TEMPLATE) + return true; + } + return false; +} + /* Check if conn should be synced. * pkts: conn packets, use sysctl_sync_threshold to avoid packet check * - (1) sync_refresh_period: reduce sync rate. Additionally, retry @@ -447,6 +457,8 @@ static int ip_vs_sync_conn_needed(struct netns_ipvs *ipvs, /* Check if we sync in current state */ if (unlikely(cp->flags & IP_VS_CONN_F_TEMPLATE)) force = 0; + else if (unlikely(sysctl_sync_persist_mode(ipvs) && in_persistence(cp))) + return 0; else if (likely(cp->protocol == IPPROTO_TCP)) { if (!((1 << cp->state) & ((1 << IP_VS_TCP_S_ESTABLISHED) | From d0667186eb0eab78dcca9f75af6ed03873ca8d9f Mon Sep 17 00:00:00 2001 From: JunweiZhang Date: Wed, 26 Jun 2013 16:40:05 +0800 Subject: [PATCH 10/12] kernel: remove unnecessary head file ip_vs.h is not necessary for sysctl_binary.c. prepare for the next patch to avoid compile issue. Signed-off-by: JunweiZhang Signed-off-by: Nicolas Dichtel Reviewed-by: Julian Anastasov Signed-off-by: Simon Horman --- kernel/sysctl_binary.c | 1 - 1 file changed, 1 deletion(-) diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c index aea4a9ea6fc8..b609213ca9a2 100644 --- a/kernel/sysctl_binary.c +++ b/kernel/sysctl_binary.c @@ -3,7 +3,6 @@ #include "../fs/xfs/xfs_sysctl.h" #include #include -#include #include #include #include From 8b4d14d8eb36874daf159d33dcccd4746a6f3189 Mon Sep 17 00:00:00 2001 From: JunweiZhang Date: Wed, 26 Jun 2013 16:40:06 +0800 Subject: [PATCH 11/12] netns: exclude ipvs from struct net when IPVS disabled no real problem is fixed, just save a few bytes in net_namespace structure. Signed-off-by: JunweiZhang Signed-off-by: Nicolas Dichtel Reviewed-by: Julian Anastasov Signed-off-by: Simon Horman --- include/net/net_namespace.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 495bc57f292c..84e37b1ca9e1 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -115,7 +115,9 @@ struct net { #ifdef CONFIG_XFRM struct netns_xfrm xfrm; #endif +#if IS_ENABLED(CONFIG_IP_VS) struct netns_ipvs *ipvs; +#endif struct sock *diag_nlsk; atomic_t rt_genid; atomic_t fnhe_genid; From 496e4ae7dc944faa1721bfda7e9d834d5611a874 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Sat, 29 Jun 2013 14:15:47 +0200 Subject: [PATCH 12/12] netfilter: nf_queue: add NFQA_SKB_CSUM_NOTVERIFIED info flag The common case is that TCP/IP checksums have already been verified, e.g. by hardware (rx checksum offload), or conntrack. Userspace can use this flag to determine when the checksum has not been validated yet. If the flag is set, this doesn't necessarily mean that the packet has an invalid checksum, e.g. if NIC doesn't support rx checksum. Userspace that sucessfully enabled NFQA_CFG_F_GSO queue feature flag can infer that IP/TCP checksum has already been validated if either the SKB_INFO attribute is not present or the NFQA_SKB_CSUM_NOTVERIFIED flag is unset. Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- include/uapi/linux/netfilter/nfnetlink_queue.h | 2 ++ net/netfilter/nfnetlink_queue_core.c | 16 ++++++++++++++-- 2 files changed, 16 insertions(+), 2 deletions(-) diff --git a/include/uapi/linux/netfilter/nfnetlink_queue.h b/include/uapi/linux/netfilter/nfnetlink_queue.h index a2308ae5a73d..3a9b92147339 100644 --- a/include/uapi/linux/netfilter/nfnetlink_queue.h +++ b/include/uapi/linux/netfilter/nfnetlink_queue.h @@ -105,5 +105,7 @@ enum nfqnl_attr_config { #define NFQA_SKB_CSUMNOTREADY (1 << 0) /* packet is GSO (i.e., exceeds device mtu) */ #define NFQA_SKB_GSO (1 << 1) +/* csum not validated (incoming device doesn't support hw checksum, etc.) */ +#define NFQA_SKB_CSUM_NOTVERIFIED (1 << 2) #endif /* _NFNETLINK_QUEUE_H */ diff --git a/net/netfilter/nfnetlink_queue_core.c b/net/netfilter/nfnetlink_queue_core.c index 299a48ae5dc9..971ea145ab3e 100644 --- a/net/netfilter/nfnetlink_queue_core.c +++ b/net/netfilter/nfnetlink_queue_core.c @@ -280,12 +280,17 @@ nfqnl_zcopy(struct sk_buff *to, const struct sk_buff *from, int len, int hlen) skb_shinfo(to)->nr_frags = j; } -static int nfqnl_put_packet_info(struct sk_buff *nlskb, struct sk_buff *packet) +static int +nfqnl_put_packet_info(struct sk_buff *nlskb, struct sk_buff *packet, + bool csum_verify) { __u32 flags = 0; if (packet->ip_summed == CHECKSUM_PARTIAL) flags = NFQA_SKB_CSUMNOTREADY; + else if (csum_verify) + flags = NFQA_SKB_CSUM_NOTVERIFIED; + if (skb_is_gso(packet)) flags |= NFQA_SKB_GSO; @@ -310,6 +315,7 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue, struct net_device *outdev; struct nf_conn *ct = NULL; enum ip_conntrack_info uninitialized_var(ctinfo); + bool csum_verify; size = nlmsg_total_size(sizeof(struct nfgenmsg)) + nla_total_size(sizeof(struct nfqnl_msg_packet_hdr)) @@ -327,6 +333,12 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue, if (entskb->tstamp.tv64) size += nla_total_size(sizeof(struct nfqnl_msg_packet_timestamp)); + if (entry->hook <= NF_INET_FORWARD || + (entry->hook == NF_INET_POST_ROUTING && entskb->sk == NULL)) + csum_verify = !skb_csum_unnecessary(entskb); + else + csum_verify = false; + outdev = entry->outdev; switch ((enum nfqnl_config_mode)ACCESS_ONCE(queue->copy_mode)) { @@ -476,7 +488,7 @@ nfqnl_build_packet_message(struct nfqnl_instance *queue, nla_put_be32(skb, NFQA_CAP_LEN, htonl(cap_len))) goto nla_put_failure; - if (nfqnl_put_packet_info(skb, entskb)) + if (nfqnl_put_packet_info(skb, entskb, csum_verify)) goto nla_put_failure; if (data_len) {