From patchwork Thu Jan 13 01:53:03 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Simon Horman X-Patchwork-Id: 78634 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 322EFB7043 for ; Thu, 13 Jan 2011 12:53:38 +1100 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932597Ab1AMBxa (ORCPT ); Wed, 12 Jan 2011 20:53:30 -0500 Received: from kirsty.vergenet.net ([202.4.237.240]:39623 "EHLO kirsty.vergenet.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932516Ab1AMBxU (ORCPT ); Wed, 12 Jan 2011 20:53:20 -0500 Received: from ayumi.akashicho.tokyo.vergenet.net (219-109-213-121.bitcat.net [219.109.213.121]) by kirsty.vergenet.net (Postfix) with ESMTP id 183CA245BD; Thu, 13 Jan 2011 12:53:18 +1100 (EST) Received: by ayumi.akashicho.tokyo.vergenet.net (Postfix, from userid 7100) id 16118EDE191; Thu, 13 Jan 2011 10:53:10 +0900 (JST) From: Simon Horman To: netfilter-devel@vger.kernel.org, lvs-devel@vger.kernel.org, netdev@vger.kernel.org Cc: Patrick McHardy , Pablo Neira Ayuso , Julian Anastasov , Hans Schillstrom , Simon Horman Subject: [PATCH 17/22] IPVS: netns, ip_vs_ctl local vars moved to ipvs struct. Date: Thu, 13 Jan 2011 10:53:03 +0900 Message-Id: <1294883588-5683-18-git-send-email-horms@verge.net.au> X-Mailer: git-send-email 1.7.2.3 In-Reply-To: <1294883588-5683-1-git-send-email-horms@verge.net.au> References: <1294883588-5683-1-git-send-email-horms@verge.net.au> Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org From: Hans Schillstrom Moving global vars to ipvs struct, except for svc table lock. Next patch for ctl will be drop-rate handling. *v3 __ip_vs_mutex remains global ip_vs_conntrack_enabled(struct netns_ipvs *ipvs) Signed-off-by: Hans Schillstrom Acked-by: Julian Anastasov Signed-off-by: Simon Horman --- include/net/ip_vs.h | 27 ++-- include/net/netns/ip_vs.h | 37 ++++- net/netfilter/ipvs/ip_vs_conn.c | 7 +- net/netfilter/ipvs/ip_vs_core.c | 34 +++-- net/netfilter/ipvs/ip_vs_ctl.c | 291 +++++++++++++++++---------------- net/netfilter/ipvs/ip_vs_proto_sctp.c | 2 +- net/netfilter/ipvs/ip_vs_proto_tcp.c | 2 +- net/netfilter/ipvs/ip_vs_proto_udp.c | 2 +- net/netfilter/ipvs/ip_vs_sync.c | 9 +- 9 files changed, 230 insertions(+), 181 deletions(-) diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index f82c0ff..af9acf4 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -41,7 +41,7 @@ static inline struct netns_ipvs *net_ipvs(struct net* net) * Get net ptr from skb in traffic cases * use skb_sknet when call is from userland (ioctl or netlink) */ -static inline struct net *skb_net(struct sk_buff *skb) +static inline struct net *skb_net(const struct sk_buff *skb) { #ifdef CONFIG_NET_NS #ifdef CONFIG_IP_VS_DEBUG @@ -69,7 +69,7 @@ static inline struct net *skb_net(struct sk_buff *skb) #endif } -static inline struct net *skb_sknet(struct sk_buff *skb) +static inline struct net *skb_sknet(const struct sk_buff *skb) { #ifdef CONFIG_NET_NS #ifdef CONFIG_IP_VS_DEBUG @@ -1023,13 +1023,6 @@ extern int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb, /* * IPVS control data and functions (from ip_vs_ctl.c) */ -extern int sysctl_ip_vs_cache_bypass; -extern int sysctl_ip_vs_expire_nodest_conn; -extern int sysctl_ip_vs_expire_quiescent_template; -extern int sysctl_ip_vs_sync_threshold[2]; -extern int sysctl_ip_vs_nat_icmp_send; -extern int sysctl_ip_vs_conntrack; -extern int sysctl_ip_vs_snat_reroute; extern struct ip_vs_stats ip_vs_stats; extern const struct ctl_path net_vs_ctl_path[]; extern int sysctl_ip_vs_sync_ver; @@ -1119,11 +1112,13 @@ extern int ip_vs_icmp_xmit_v6 extern int ip_vs_drop_rate; extern int ip_vs_drop_counter; -static __inline__ int ip_vs_todrop(void) +static inline int ip_vs_todrop(struct netns_ipvs *ipvs) { - if (!ip_vs_drop_rate) return 0; - if (--ip_vs_drop_counter > 0) return 0; - ip_vs_drop_counter = ip_vs_drop_rate; + if (!ipvs->drop_rate) + return 0; + if (--ipvs->drop_counter > 0) + return 0; + ipvs->drop_counter = ipvs->drop_rate; return 1; } @@ -1211,9 +1206,9 @@ static inline void ip_vs_notrack(struct sk_buff *skb) * Netfilter connection tracking * (from ip_vs_nfct.c) */ -static inline int ip_vs_conntrack_enabled(void) +static inline int ip_vs_conntrack_enabled(struct netns_ipvs *ipvs) { - return sysctl_ip_vs_conntrack; + return ipvs->sysctl_conntrack; } extern void ip_vs_update_conntrack(struct sk_buff *skb, struct ip_vs_conn *cp, @@ -1226,7 +1221,7 @@ extern void ip_vs_conn_drop_conntrack(struct ip_vs_conn *cp); #else -static inline int ip_vs_conntrack_enabled(void) +static inline int ip_vs_conntrack_enabled(struct netns_ipvs *ipvs) { return 0; } diff --git a/include/net/netns/ip_vs.h b/include/net/netns/ip_vs.h index 1acfb33..c4b1abf 100644 --- a/include/net/netns/ip_vs.h +++ b/include/net/netns/ip_vs.h @@ -61,13 +61,46 @@ struct netns_ipvs { struct list_head sctp_apps[SCTP_APP_TAB_SIZE]; spinlock_t sctp_app_lock; #endif + /* ip_vs_conn */ + atomic_t conn_count; /* connection counter */ + /* ip_vs_ctl */ struct ip_vs_stats *tot_stats; /* Statistics & est. */ struct ip_vs_cpu_stats __percpu *cpustats; /* Stats per cpu */ seqcount_t *ustats_seq; /* u64 read retry */ - /* ip_vs_conn */ - atomic_t conn_count; /* connection counter */ + int num_services; /* no of virtual services */ + /* 1/rate drop and drop-entry variables */ + int drop_rate; + int drop_counter; + atomic_t dropentry; + /* locks in ctl.c */ + spinlock_t dropentry_lock; /* drop entry handling */ + spinlock_t droppacket_lock; /* drop packet handling */ + spinlock_t securetcp_lock; /* state and timeout tables */ + rwlock_t rs_lock; /* real services table */ + /* semaphore for IPVS sockopts. And, [gs]etsockopt may sleep. */ + struct lock_class_key ctl_key; /* ctl_mutex debuging */ + /* sys-ctl struct */ + struct ctl_table_header *sysctl_hdr; + struct ctl_table *sysctl_tbl; + /* sysctl variables */ + int sysctl_amemthresh; + int sysctl_am_droprate; + int sysctl_drop_entry; + int sysctl_drop_packet; + int sysctl_secure_tcp; +#ifdef CONFIG_IP_VS_NFCT + int sysctl_conntrack; +#endif + int sysctl_snat_reroute; + int sysctl_sync_ver; + int sysctl_cache_bypass; + int sysctl_expire_nodest_conn; + int sysctl_expire_quiescent_template; + int sysctl_sync_threshold[2]; + int sysctl_nat_icmp_send; + /* ip_vs_lblc */ int sysctl_lblc_expiration; struct ctl_table_header *lblc_ctl_header; diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 0d5e4fe..5ba205a 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -686,13 +686,14 @@ static inline void ip_vs_unbind_dest(struct ip_vs_conn *cp) int ip_vs_check_template(struct ip_vs_conn *ct) { struct ip_vs_dest *dest = ct->dest; + struct netns_ipvs *ipvs = net_ipvs(ip_vs_conn_net(ct)); /* * Checking the dest server status. */ if ((dest == NULL) || !(dest->flags & IP_VS_DEST_F_AVAILABLE) || - (sysctl_ip_vs_expire_quiescent_template && + (ipvs->sysctl_expire_quiescent_template && (atomic_read(&dest->weight) == 0))) { IP_VS_DBG_BUF(9, "check_template: dest not available for " "protocol %s s:%s:%d v:%s:%d " @@ -879,7 +880,7 @@ ip_vs_conn_new(const struct ip_vs_conn_param *p, * IP_VS_CONN_F_ONE_PACKET too. */ - if (ip_vs_conntrack_enabled()) + if (ip_vs_conntrack_enabled(ipvs)) cp->flags |= IP_VS_CONN_F_NFCT; /* Hash it in the ip_vs_conn_tab finally */ @@ -1198,7 +1199,7 @@ static void ip_vs_conn_flush(struct net *net) struct ip_vs_conn *cp; struct netns_ipvs *ipvs = net_ipvs(net); - flush_again: +flush_again: for (idx = 0; idx < ip_vs_conn_tab_size; idx++) { /* * Lock is actually needed in this loop. diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index 7205b49..a7c59a7 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -499,6 +499,7 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb, int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb, struct ip_vs_proto_data *pd) { + struct netns_ipvs *ipvs; __be16 _ports[2], *pptr; struct ip_vs_iphdr iph; int unicast; @@ -521,7 +522,8 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb, /* if it is fwmark-based service, the cache_bypass sysctl is up and the destination is a non-local unicast, then create a cache_bypass connection entry */ - if (sysctl_ip_vs_cache_bypass && svc->fwmark && unicast) { + ipvs = net_ipvs(skb_net(skb)); + if (ipvs->sysctl_cache_bypass && svc->fwmark && unicast) { int ret, cs; struct ip_vs_conn *cp; unsigned int flags = (svc->flags & IP_VS_SVC_F_ONEPACKET && @@ -733,6 +735,7 @@ static int handle_response_icmp(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, unsigned int offset, unsigned int ihl) { + struct netns_ipvs *ipvs; unsigned int verdict = NF_DROP; if (IP_VS_FWD_METHOD(cp) != 0) { @@ -754,6 +757,8 @@ static int handle_response_icmp(int af, struct sk_buff *skb, if (!skb_make_writable(skb, offset)) goto out; + ipvs = net_ipvs(skb_net(skb)); + #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) ip_vs_nat_icmp_v6(skb, pp, cp, 1); @@ -763,11 +768,11 @@ static int handle_response_icmp(int af, struct sk_buff *skb, #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) { - if (sysctl_ip_vs_snat_reroute && ip6_route_me_harder(skb) != 0) + if (ipvs->sysctl_snat_reroute && ip6_route_me_harder(skb) != 0) goto out; } else #endif - if ((sysctl_ip_vs_snat_reroute || + if ((ipvs->sysctl_snat_reroute || skb_rtable(skb)->rt_flags & RTCF_LOCAL) && ip_route_me_harder(skb, RTN_LOCAL) != 0) goto out; @@ -979,6 +984,7 @@ handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, struct ip_vs_conn *cp, int ihl) { struct ip_vs_protocol *pp = pd->pp; + struct netns_ipvs *ipvs; IP_VS_DBG_PKT(11, af, pp, skb, 0, "Outgoing packet"); @@ -1014,13 +1020,15 @@ handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, * if it came from this machine itself. So re-compute * the routing information. */ + ipvs = net_ipvs(skb_net(skb)); + #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) { - if (sysctl_ip_vs_snat_reroute && ip6_route_me_harder(skb) != 0) + if (ipvs->sysctl_snat_reroute && ip6_route_me_harder(skb) != 0) goto drop; } else #endif - if ((sysctl_ip_vs_snat_reroute || + if ((ipvs->sysctl_snat_reroute || skb_rtable(skb)->rt_flags & RTCF_LOCAL) && ip_route_me_harder(skb, RTN_LOCAL) != 0) goto drop; @@ -1057,6 +1065,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af) struct ip_vs_protocol *pp; struct ip_vs_proto_data *pd; struct ip_vs_conn *cp; + struct netns_ipvs *ipvs; EnterFunction(11); @@ -1131,10 +1140,11 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af) * Check if the packet belongs to an existing entry */ cp = pp->conn_out_get(af, skb, &iph, iph.len, 0); + ipvs = net_ipvs(net); if (likely(cp)) return handle_response(af, skb, pd, cp, iph.len); - if (sysctl_ip_vs_nat_icmp_send && + if (ipvs->sysctl_nat_icmp_send && (pp->protocol == IPPROTO_TCP || pp->protocol == IPPROTO_UDP || pp->protocol == IPPROTO_SCTP)) { @@ -1580,7 +1590,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) if (cp->dest && !(cp->dest->flags & IP_VS_DEST_F_AVAILABLE)) { /* the destination server is not available */ - if (sysctl_ip_vs_expire_nodest_conn) { + if (ipvs->sysctl_expire_nodest_conn) { /* try to expire the connection immediately */ ip_vs_conn_expire_now(cp); } @@ -1610,15 +1620,15 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) */ if (cp->flags & IP_VS_CONN_F_ONE_PACKET) - pkts = sysctl_ip_vs_sync_threshold[0]; + pkts = ipvs->sysctl_sync_threshold[0]; else pkts = atomic_add_return(1, &cp->in_pkts); if ((ipvs->sync_state & IP_VS_STATE_MASTER) && cp->protocol == IPPROTO_SCTP) { if ((cp->state == IP_VS_SCTP_S_ESTABLISHED && - (pkts % sysctl_ip_vs_sync_threshold[1] - == sysctl_ip_vs_sync_threshold[0])) || + (pkts % ipvs->sysctl_sync_threshold[1] + == ipvs->sysctl_sync_threshold[0])) || (cp->old_state != cp->state && ((cp->state == IP_VS_SCTP_S_CLOSED) || (cp->state == IP_VS_SCTP_S_SHUT_ACK_CLI) || @@ -1632,8 +1642,8 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af) else if ((ipvs->sync_state & IP_VS_STATE_MASTER) && (((cp->protocol != IPPROTO_TCP || cp->state == IP_VS_TCP_S_ESTABLISHED) && - (pkts % sysctl_ip_vs_sync_threshold[1] - == sysctl_ip_vs_sync_threshold[0])) || + (pkts % ipvs->sysctl_sync_threshold[1] + == ipvs->sysctl_sync_threshold[0])) || ((cp->protocol == IPPROTO_TCP) && (cp->old_state != cp->state) && ((cp->state == IP_VS_TCP_S_FIN_WAIT) || (cp->state == IP_VS_TCP_S_CLOSE) || diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index cbd58c6..183ac18 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -58,42 +58,7 @@ static DEFINE_MUTEX(__ip_vs_mutex); /* lock for service table */ static DEFINE_RWLOCK(__ip_vs_svc_lock); -/* lock for table with the real services */ -static DEFINE_RWLOCK(__ip_vs_rs_lock); - -/* lock for state and timeout tables */ -static DEFINE_SPINLOCK(ip_vs_securetcp_lock); - -/* lock for drop entry handling */ -static DEFINE_SPINLOCK(__ip_vs_dropentry_lock); - -/* lock for drop packet handling */ -static DEFINE_SPINLOCK(__ip_vs_droppacket_lock); - -/* 1/rate drop and drop-entry variables */ -int ip_vs_drop_rate = 0; -int ip_vs_drop_counter = 0; -static atomic_t ip_vs_dropentry = ATOMIC_INIT(0); - -/* number of virtual services */ -static int ip_vs_num_services = 0; - /* sysctl variables */ -static int sysctl_ip_vs_drop_entry = 0; -static int sysctl_ip_vs_drop_packet = 0; -static int sysctl_ip_vs_secure_tcp = 0; -static int sysctl_ip_vs_amemthresh = 1024; -static int sysctl_ip_vs_am_droprate = 10; -int sysctl_ip_vs_cache_bypass = 0; -int sysctl_ip_vs_expire_nodest_conn = 0; -int sysctl_ip_vs_expire_quiescent_template = 0; -int sysctl_ip_vs_sync_threshold[2] = { 3, 50 }; -int sysctl_ip_vs_nat_icmp_send = 0; -#ifdef CONFIG_IP_VS_NFCT -int sysctl_ip_vs_conntrack; -#endif -int sysctl_ip_vs_snat_reroute = 1; -int sysctl_ip_vs_sync_ver = 1; /* Default version of sync proto */ #ifdef CONFIG_IP_VS_DEBUG static int sysctl_ip_vs_debug_level = 0; @@ -142,73 +107,73 @@ static void update_defense_level(struct netns_ipvs *ipvs) /* si_swapinfo(&i); */ /* availmem = availmem - (i.totalswap - i.freeswap); */ - nomem = (availmem < sysctl_ip_vs_amemthresh); + nomem = (availmem < ipvs->sysctl_amemthresh); local_bh_disable(); /* drop_entry */ - spin_lock(&__ip_vs_dropentry_lock); - switch (sysctl_ip_vs_drop_entry) { + spin_lock(&ipvs->dropentry_lock); + switch (ipvs->sysctl_drop_entry) { case 0: - atomic_set(&ip_vs_dropentry, 0); + atomic_set(&ipvs->dropentry, 0); break; case 1: if (nomem) { - atomic_set(&ip_vs_dropentry, 1); - sysctl_ip_vs_drop_entry = 2; + atomic_set(&ipvs->dropentry, 1); + ipvs->sysctl_drop_entry = 2; } else { - atomic_set(&ip_vs_dropentry, 0); + atomic_set(&ipvs->dropentry, 0); } break; case 2: if (nomem) { - atomic_set(&ip_vs_dropentry, 1); + atomic_set(&ipvs->dropentry, 1); } else { - atomic_set(&ip_vs_dropentry, 0); - sysctl_ip_vs_drop_entry = 1; + atomic_set(&ipvs->dropentry, 0); + ipvs->sysctl_drop_entry = 1; }; break; case 3: - atomic_set(&ip_vs_dropentry, 1); + atomic_set(&ipvs->dropentry, 1); break; } - spin_unlock(&__ip_vs_dropentry_lock); + spin_unlock(&ipvs->dropentry_lock); /* drop_packet */ - spin_lock(&__ip_vs_droppacket_lock); - switch (sysctl_ip_vs_drop_packet) { + spin_lock(&ipvs->droppacket_lock); + switch (ipvs->sysctl_drop_packet) { case 0: - ip_vs_drop_rate = 0; + ipvs->drop_rate = 0; break; case 1: if (nomem) { - ip_vs_drop_rate = ip_vs_drop_counter - = sysctl_ip_vs_amemthresh / - (sysctl_ip_vs_amemthresh-availmem); - sysctl_ip_vs_drop_packet = 2; + ipvs->drop_rate = ipvs->drop_counter + = ipvs->sysctl_amemthresh / + (ipvs->sysctl_amemthresh-availmem); + ipvs->sysctl_drop_packet = 2; } else { - ip_vs_drop_rate = 0; + ipvs->drop_rate = 0; } break; case 2: if (nomem) { - ip_vs_drop_rate = ip_vs_drop_counter - = sysctl_ip_vs_amemthresh / - (sysctl_ip_vs_amemthresh-availmem); + ipvs->drop_rate = ipvs->drop_counter + = ipvs->sysctl_amemthresh / + (ipvs->sysctl_amemthresh-availmem); } else { - ip_vs_drop_rate = 0; - sysctl_ip_vs_drop_packet = 1; + ipvs->drop_rate = 0; + ipvs->sysctl_drop_packet = 1; } break; case 3: - ip_vs_drop_rate = sysctl_ip_vs_am_droprate; + ipvs->drop_rate = ipvs->sysctl_am_droprate; break; } - spin_unlock(&__ip_vs_droppacket_lock); + spin_unlock(&ipvs->droppacket_lock); /* secure_tcp */ - spin_lock(&ip_vs_securetcp_lock); - switch (sysctl_ip_vs_secure_tcp) { + spin_lock(&ipvs->securetcp_lock); + switch (ipvs->sysctl_secure_tcp) { case 0: if (old_secure_tcp >= 2) to_change = 0; @@ -217,7 +182,7 @@ static void update_defense_level(struct netns_ipvs *ipvs) if (nomem) { if (old_secure_tcp < 2) to_change = 1; - sysctl_ip_vs_secure_tcp = 2; + ipvs->sysctl_secure_tcp = 2; } else { if (old_secure_tcp >= 2) to_change = 0; @@ -230,7 +195,7 @@ static void update_defense_level(struct netns_ipvs *ipvs) } else { if (old_secure_tcp >= 2) to_change = 0; - sysctl_ip_vs_secure_tcp = 1; + ipvs->sysctl_secure_tcp = 1; } break; case 3: @@ -238,11 +203,11 @@ static void update_defense_level(struct netns_ipvs *ipvs) to_change = 1; break; } - old_secure_tcp = sysctl_ip_vs_secure_tcp; + old_secure_tcp = ipvs->sysctl_secure_tcp; if (to_change >= 0) ip_vs_protocol_timeout_change(ipvs, - sysctl_ip_vs_secure_tcp > 1); - spin_unlock(&ip_vs_securetcp_lock); + ipvs->sysctl_secure_tcp > 1); + spin_unlock(&ipvs->securetcp_lock); local_bh_enable(); } @@ -260,7 +225,7 @@ static void defense_work_handler(struct work_struct *work) struct netns_ipvs *ipvs = net_ipvs(&init_net); update_defense_level(ipvs); - if (atomic_read(&ip_vs_dropentry)) + if (atomic_read(&ipvs->dropentry)) ip_vs_random_dropentry(); schedule_delayed_work(&defense_work, DEFENSE_TIMER_PERIOD); @@ -602,7 +567,7 @@ ip_vs_lookup_real_service(struct net *net, int af, __u16 protocol, */ hash = ip_vs_rs_hashkey(af, daddr, dport); - read_lock(&__ip_vs_rs_lock); + read_lock(&ipvs->rs_lock); list_for_each_entry(dest, &ipvs->rs_table[hash], d_list) { if ((dest->af == af) && ip_vs_addr_equal(af, &dest->addr, daddr) @@ -610,11 +575,11 @@ ip_vs_lookup_real_service(struct net *net, int af, __u16 protocol, && ((dest->protocol == protocol) || dest->vfwmark)) { /* HIT */ - read_unlock(&__ip_vs_rs_lock); + read_unlock(&ipvs->rs_lock); return dest; } } - read_unlock(&__ip_vs_rs_lock); + read_unlock(&ipvs->rs_lock); return NULL; } @@ -788,9 +753,9 @@ __ip_vs_update_dest(struct ip_vs_service *svc, struct ip_vs_dest *dest, * Put the real service in rs_table if not present. * For now only for NAT! */ - write_lock_bh(&__ip_vs_rs_lock); + write_lock_bh(&ipvs->rs_lock); ip_vs_rs_hash(ipvs, dest); - write_unlock_bh(&__ip_vs_rs_lock); + write_unlock_bh(&ipvs->rs_lock); } atomic_set(&dest->conn_flags, conn_flags); @@ -1022,14 +987,16 @@ ip_vs_edit_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) */ static void __ip_vs_del_dest(struct net *net, struct ip_vs_dest *dest) { + struct netns_ipvs *ipvs = net_ipvs(net); + ip_vs_kill_estimator(net, &dest->stats); /* * Remove it from the d-linked list with the real services. */ - write_lock_bh(&__ip_vs_rs_lock); + write_lock_bh(&ipvs->rs_lock); ip_vs_rs_unhash(dest); - write_unlock_bh(&__ip_vs_rs_lock); + write_unlock_bh(&ipvs->rs_lock); /* * Decrease the refcnt of the dest, and free the dest @@ -1092,7 +1059,6 @@ static int ip_vs_del_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) { struct ip_vs_dest *dest; - struct net *net = svc->net; __be16 dport = udest->port; EnterFunction(2); @@ -1121,7 +1087,7 @@ ip_vs_del_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) /* * Delete the destination */ - __ip_vs_del_dest(net, dest); + __ip_vs_del_dest(svc->net, dest); LeaveFunction(2); @@ -1140,6 +1106,7 @@ ip_vs_add_service(struct net *net, struct ip_vs_service_user_kern *u, struct ip_vs_scheduler *sched = NULL; struct ip_vs_pe *pe = NULL; struct ip_vs_service *svc = NULL; + struct netns_ipvs *ipvs = net_ipvs(net); /* increase the module use count */ ip_vs_use_count_inc(); @@ -1219,7 +1186,7 @@ ip_vs_add_service(struct net *net, struct ip_vs_service_user_kern *u, /* Count only IPv4 services for old get/setsockopt interface */ if (svc->af == AF_INET) - ip_vs_num_services++; + ipvs->num_services++; /* Hash the service into the service table */ write_lock_bh(&__ip_vs_svc_lock); @@ -1359,12 +1326,13 @@ static void __ip_vs_del_service(struct ip_vs_service *svc) struct ip_vs_dest *dest, *nxt; struct ip_vs_scheduler *old_sched; struct ip_vs_pe *old_pe; + struct netns_ipvs *ipvs = net_ipvs(svc->net); pr_info("%s: enter\n", __func__); /* Count only IPv4 services for old get/setsockopt interface */ if (svc->af == AF_INET) - ip_vs_num_services--; + ipvs->num_services--; ip_vs_kill_estimator(svc->net, &svc->stats); @@ -1589,42 +1557,31 @@ proc_do_sync_mode(ctl_table *table, int write, /* * IPVS sysctl table (under the /proc/sys/net/ipv4/vs/) + * Do not change order or insert new entries without + * align with netns init in __ip_vs_control_init() */ static struct ctl_table vs_vars[] = { { .procname = "amemthresh", - .data = &sysctl_ip_vs_amemthresh, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, -#ifdef CONFIG_IP_VS_DEBUG - { - .procname = "debug_level", - .data = &sysctl_ip_vs_debug_level, .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec, }, -#endif { .procname = "am_droprate", - .data = &sysctl_ip_vs_am_droprate, .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec, }, { .procname = "drop_entry", - .data = &sysctl_ip_vs_drop_entry, .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_do_defense_mode, }, { .procname = "drop_packet", - .data = &sysctl_ip_vs_drop_packet, .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_do_defense_mode, @@ -1632,7 +1589,6 @@ static struct ctl_table vs_vars[] = { #ifdef CONFIG_IP_VS_NFCT { .procname = "conntrack", - .data = &sysctl_ip_vs_conntrack, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &proc_dointvec, @@ -1640,25 +1596,62 @@ static struct ctl_table vs_vars[] = { #endif { .procname = "secure_tcp", - .data = &sysctl_ip_vs_secure_tcp, .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_do_defense_mode, }, { .procname = "snat_reroute", - .data = &sysctl_ip_vs_snat_reroute, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &proc_dointvec, }, { .procname = "sync_version", - .data = &sysctl_ip_vs_sync_ver, .maxlen = sizeof(int), .mode = 0644, .proc_handler = &proc_do_sync_mode, }, + { + .procname = "cache_bypass", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "expire_nodest_conn", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "expire_quiescent_template", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "sync_threshold", + .maxlen = + sizeof(((struct netns_ipvs *)0)->sysctl_sync_threshold), + .mode = 0644, + .proc_handler = proc_do_sync_threshold, + }, + { + .procname = "nat_icmp_send", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, +#ifdef CONFIG_IP_VS_DEBUG + { + .procname = "debug_level", + .data = &sysctl_ip_vs_debug_level, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, +#endif #if 0 { .procname = "timeout_established", @@ -1745,41 +1738,6 @@ static struct ctl_table vs_vars[] = { .proc_handler = proc_dointvec_jiffies, }, #endif - { - .procname = "cache_bypass", - .data = &sysctl_ip_vs_cache_bypass, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, - { - .procname = "expire_nodest_conn", - .data = &sysctl_ip_vs_expire_nodest_conn, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, - { - .procname = "expire_quiescent_template", - .data = &sysctl_ip_vs_expire_quiescent_template, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, - { - .procname = "sync_threshold", - .data = &sysctl_ip_vs_sync_threshold, - .maxlen = sizeof(sysctl_ip_vs_sync_threshold), - .mode = 0644, - .proc_handler = proc_do_sync_threshold, - }, - { - .procname = "nat_icmp_send", - .data = &sysctl_ip_vs_nat_icmp_send, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec, - }, { } }; @@ -1791,8 +1749,6 @@ const struct ctl_path net_vs_ctl_path[] = { }; EXPORT_SYMBOL_GPL(net_vs_ctl_path); -static struct ctl_table_header * sysctl_header; - #ifdef CONFIG_PROC_FS struct ip_vs_iter { @@ -2543,7 +2499,7 @@ do_ip_vs_get_ctl(struct sock *sk, int cmd, void __user *user, int *len) struct ip_vs_getinfo info; info.version = IP_VS_VERSION_CODE; info.size = ip_vs_conn_tab_size; - info.num_services = ip_vs_num_services; + info.num_services = ipvs->num_services; if (copy_to_user(user, &info, sizeof(info)) != 0) ret = -EFAULT; } @@ -3014,7 +2970,7 @@ static int ip_vs_genl_dump_dests(struct sk_buff *skb, struct ip_vs_service *svc; struct ip_vs_dest *dest; struct nlattr *attrs[IPVS_CMD_ATTR_MAX + 1]; - struct net *net; + struct net *net = skb_sknet(skb); mutex_lock(&__ip_vs_mutex); @@ -3023,7 +2979,7 @@ static int ip_vs_genl_dump_dests(struct sk_buff *skb, IPVS_CMD_ATTR_MAX, ip_vs_cmd_policy)) goto out_err; - net = skb_sknet(skb); + svc = ip_vs_genl_find_service(net, attrs[IPVS_CMD_ATTR_SERVICE]); if (IS_ERR(svc) || svc == NULL) goto out_err; @@ -3215,8 +3171,10 @@ static int ip_vs_genl_set_cmd(struct sk_buff *skb, struct genl_info *info) int ret = 0, cmd; int need_full_svc = 0, need_full_dest = 0; struct net *net; + struct netns_ipvs *ipvs; net = skb_sknet(skb); + ipvs = net_ipvs(net); cmd = info->genlhdr->cmd; mutex_lock(&__ip_vs_mutex); @@ -3326,8 +3284,10 @@ static int ip_vs_genl_get_cmd(struct sk_buff *skb, struct genl_info *info) void *reply; int ret, cmd, reply_cmd; struct net *net; + struct netns_ipvs *ipvs; net = skb_sknet(skb); + ipvs = net_ipvs(net); cmd = info->genlhdr->cmd; if (cmd == IPVS_CMD_GET_SERVICE) @@ -3530,9 +3490,21 @@ int __net_init __ip_vs_control_init(struct net *net) { int idx; struct netns_ipvs *ipvs = net_ipvs(net); + struct ctl_table *tbl; if (!net_eq(net, &init_net)) /* netns not enabled yet */ return -EPERM; + + atomic_set(&ipvs->dropentry, 0); + spin_lock_init(&ipvs->dropentry_lock); + spin_lock_init(&ipvs->droppacket_lock); + spin_lock_init(&ipvs->securetcp_lock); + ipvs->rs_lock = __RW_LOCK_UNLOCKED(ipvs->rs_lock); + + /* Initialize rs_table */ + for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++) + INIT_LIST_HEAD(&ipvs->rs_table[idx]); + /* procfs stats */ ipvs->tot_stats = kzalloc(sizeof(struct ip_vs_stats), GFP_KERNEL); if (ipvs->tot_stats == NULL) { @@ -3553,14 +3525,51 @@ int __net_init __ip_vs_control_init(struct net *net) proc_net_fops_create(net, "ip_vs_stats", 0, &ip_vs_stats_fops); proc_net_fops_create(net, "ip_vs_stats_percpu", 0, &ip_vs_stats_percpu_fops); - sysctl_header = register_net_sysctl_table(net, net_vs_ctl_path, + + if (!net_eq(net, &init_net)) { + tbl = kmemdup(vs_vars, sizeof(vs_vars), GFP_KERNEL); + if (tbl == NULL) + goto err_dup; + } else + tbl = vs_vars; + /* Initialize sysctl defaults */ + idx = 0; + ipvs->sysctl_amemthresh = 1024; + tbl[idx++].data = &ipvs->sysctl_amemthresh; + ipvs->sysctl_am_droprate = 10; + tbl[idx++].data = &ipvs->sysctl_am_droprate; + tbl[idx++].data = &ipvs->sysctl_drop_entry; + tbl[idx++].data = &ipvs->sysctl_drop_packet; +#ifdef CONFIG_IP_VS_NFCT + tbl[idx++].data = &ipvs->sysctl_conntrack; +#endif + tbl[idx++].data = &ipvs->sysctl_secure_tcp; + ipvs->sysctl_snat_reroute = 1; + tbl[idx++].data = &ipvs->sysctl_snat_reroute; + ipvs->sysctl_sync_ver = 1; + tbl[idx++].data = &ipvs->sysctl_sync_ver; + tbl[idx++].data = &ipvs->sysctl_cache_bypass; + tbl[idx++].data = &ipvs->sysctl_expire_nodest_conn; + tbl[idx++].data = &ipvs->sysctl_expire_quiescent_template; + ipvs->sysctl_sync_threshold[0] = 3; + ipvs->sysctl_sync_threshold[1] = 50; + tbl[idx].data = &ipvs->sysctl_sync_threshold; + tbl[idx++].maxlen = sizeof(ipvs->sysctl_sync_threshold); + tbl[idx++].data = &ipvs->sysctl_nat_icmp_send; + + + ipvs->sysctl_hdr = register_net_sysctl_table(net, net_vs_ctl_path, vs_vars); - if (sysctl_header == NULL) + if (ipvs->sysctl_hdr == NULL) goto err_reg; ip_vs_new_estimator(net, ipvs->tot_stats); + ipvs->sysctl_tbl = tbl; return 0; err_reg: + if (!net_eq(net, &init_net)) + kfree(tbl); +err_dup: free_percpu(ipvs->cpustats); err_alloc: kfree(ipvs->tot_stats); @@ -3575,7 +3584,7 @@ static void __net_exit __ip_vs_control_cleanup(struct net *net) return; ip_vs_kill_estimator(net, ipvs->tot_stats); - unregister_net_sysctl_table(sysctl_header); + unregister_net_sysctl_table(ipvs->sysctl_hdr); proc_net_remove(net, "ip_vs_stats_percpu"); proc_net_remove(net, "ip_vs_stats"); proc_net_remove(net, "ip_vs"); diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c index 550365a..fb2d04a 100644 --- a/net/netfilter/ipvs/ip_vs_proto_sctp.c +++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c @@ -34,7 +34,7 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, &iph.daddr, sh->dest))) { int ignored; - if (ip_vs_todrop()) { + if (ip_vs_todrop(net_ipvs(net))) { /* * It seems that we are very loaded. * We have to drop this packet :( diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index d8b3f9f..c0cc341 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -54,7 +54,7 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, &iph.daddr, th->dest))) { int ignored; - if (ip_vs_todrop()) { + if (ip_vs_todrop(net_ipvs(net))) { /* * It seems that we are very loaded. * We have to drop this packet :( diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index 581157b..f1282cb 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -50,7 +50,7 @@ udp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd, if (svc) { int ignored; - if (ip_vs_todrop()) { + if (ip_vs_todrop(net_ipvs(net))) { /* * It seems that we are very loaded. * We have to drop this packet :( diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index f85e47d..b178056 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -394,7 +394,7 @@ void ip_vs_sync_switch_mode(struct net *net, int mode) if (!ipvs->sync_state & IP_VS_STATE_MASTER) return; - if (mode == sysctl_ip_vs_sync_ver || !ipvs->sync_buff) + if (mode == ipvs->sysctl_sync_ver || !ipvs->sync_buff) return; spin_lock_bh(&ipvs->sync_buff_lock); @@ -521,7 +521,7 @@ void ip_vs_sync_conn(struct net *net, struct ip_vs_conn *cp) unsigned int len, pe_name_len, pad; /* Handle old version of the protocol */ - if (sysctl_ip_vs_sync_ver == 0) { + if (ipvs->sysctl_sync_ver == 0) { ip_vs_sync_conn_v0(net, cp); return; } @@ -650,7 +650,7 @@ control: if (cp->flags & IP_VS_CONN_F_TEMPLATE) { int pkts = atomic_add_return(1, &cp->in_pkts); - if (pkts % sysctl_ip_vs_sync_threshold[1] != 1) + if (pkts % ipvs->sysctl_sync_threshold[1] != 1) return; } goto sloop; @@ -724,6 +724,7 @@ static void ip_vs_proc_conn(struct net *net, struct ip_vs_conn_param *param, { struct ip_vs_dest *dest; struct ip_vs_conn *cp; + struct netns_ipvs *ipvs = net_ipvs(net); if (!(flags & IP_VS_CONN_F_TEMPLATE)) cp = ip_vs_conn_in_get(param); @@ -794,7 +795,7 @@ static void ip_vs_proc_conn(struct net *net, struct ip_vs_conn_param *param, if (opt) memcpy(&cp->in_seq, opt, sizeof(*opt)); - atomic_set(&cp->in_pkts, sysctl_ip_vs_sync_threshold[0]); + atomic_set(&cp->in_pkts, ipvs->sysctl_sync_threshold[0]); cp->state = state; cp->old_state = cp->state; /*