diff options
Diffstat (limited to 'net/netfilter/nf_conntrack_core.c')
-rw-r--r-- | net/netfilter/nf_conntrack_core.c | 699 |
1 files changed, 121 insertions, 578 deletions
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c index 836541e509fe..93d97d9f9da8 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -47,14 +47,10 @@ #include <linux/netdevice.h> #include <linux/socket.h> -/* This rwlock protects the main hash table, protocol/helper/expected - registrations, conntrack timers*/ -#define ASSERT_READ_LOCK(x) -#define ASSERT_WRITE_LOCK(x) - #include <net/netfilter/nf_conntrack.h> #include <net/netfilter/nf_conntrack_l3proto.h> -#include <net/netfilter/nf_conntrack_protocol.h> +#include <net/netfilter/nf_conntrack_l4proto.h> +#include <net/netfilter/nf_conntrack_expect.h> #include <net/netfilter/nf_conntrack_helper.h> #include <net/netfilter/nf_conntrack_core.h> @@ -67,92 +63,32 @@ #endif DEFINE_RWLOCK(nf_conntrack_lock); +EXPORT_SYMBOL_GPL(nf_conntrack_lock); /* nf_conntrack_standalone needs this */ atomic_t nf_conntrack_count = ATOMIC_INIT(0); +EXPORT_SYMBOL_GPL(nf_conntrack_count); -void (*nf_conntrack_destroyed)(struct nf_conn *conntrack) = NULL; -LIST_HEAD(nf_conntrack_expect_list); -struct nf_conntrack_protocol **nf_ct_protos[PF_MAX] __read_mostly; -struct nf_conntrack_l3proto *nf_ct_l3protos[PF_MAX] __read_mostly; -static LIST_HEAD(helpers); -unsigned int nf_conntrack_htable_size __read_mostly = 0; -int nf_conntrack_max __read_mostly; -struct list_head *nf_conntrack_hash __read_mostly; -static kmem_cache_t *nf_conntrack_expect_cachep __read_mostly; -struct nf_conn nf_conntrack_untracked; -unsigned int nf_ct_log_invalid __read_mostly; -static LIST_HEAD(unconfirmed); -static int nf_conntrack_vmalloc __read_mostly; - -static unsigned int nf_conntrack_next_id; -static unsigned int nf_conntrack_expect_next_id; -#ifdef CONFIG_NF_CONNTRACK_EVENTS -ATOMIC_NOTIFIER_HEAD(nf_conntrack_chain); -ATOMIC_NOTIFIER_HEAD(nf_conntrack_expect_chain); +void (*nf_conntrack_destroyed)(struct nf_conn *conntrack); +EXPORT_SYMBOL_GPL(nf_conntrack_destroyed); -DEFINE_PER_CPU(struct nf_conntrack_ecache, nf_conntrack_ecache); +unsigned int nf_conntrack_htable_size __read_mostly; +EXPORT_SYMBOL_GPL(nf_conntrack_htable_size); -/* deliver cached events and clear cache entry - must be called with locally - * disabled softirqs */ -static inline void -__nf_ct_deliver_cached_events(struct nf_conntrack_ecache *ecache) -{ - DEBUGP("ecache: delivering events for %p\n", ecache->ct); - if (nf_ct_is_confirmed(ecache->ct) && !nf_ct_is_dying(ecache->ct) - && ecache->events) - atomic_notifier_call_chain(&nf_conntrack_chain, ecache->events, - ecache->ct); - - ecache->events = 0; - nf_ct_put(ecache->ct); - ecache->ct = NULL; -} +int nf_conntrack_max __read_mostly; +EXPORT_SYMBOL_GPL(nf_conntrack_max); -/* Deliver all cached events for a particular conntrack. This is called - * by code prior to async packet handling for freeing the skb */ -void nf_ct_deliver_cached_events(const struct nf_conn *ct) -{ - struct nf_conntrack_ecache *ecache; +struct list_head *nf_conntrack_hash __read_mostly; +EXPORT_SYMBOL_GPL(nf_conntrack_hash); - local_bh_disable(); - ecache = &__get_cpu_var(nf_conntrack_ecache); - if (ecache->ct == ct) - __nf_ct_deliver_cached_events(ecache); - local_bh_enable(); -} +struct nf_conn nf_conntrack_untracked __read_mostly; +EXPORT_SYMBOL_GPL(nf_conntrack_untracked); -/* Deliver cached events for old pending events, if current conntrack != old */ -void __nf_ct_event_cache_init(struct nf_conn *ct) -{ - struct nf_conntrack_ecache *ecache; - - /* take care of delivering potentially old events */ - ecache = &__get_cpu_var(nf_conntrack_ecache); - BUG_ON(ecache->ct == ct); - if (ecache->ct) - __nf_ct_deliver_cached_events(ecache); - /* initialize for this conntrack/packet */ - ecache->ct = ct; - nf_conntrack_get(&ct->ct_general); -} - -/* flush the event cache - touches other CPU's data and must not be called - * while packets are still passing through the code */ -static void nf_ct_event_cache_flush(void) -{ - struct nf_conntrack_ecache *ecache; - int cpu; +unsigned int nf_ct_log_invalid __read_mostly; +LIST_HEAD(unconfirmed); +static int nf_conntrack_vmalloc __read_mostly; - for_each_possible_cpu(cpu) { - ecache = &per_cpu(nf_conntrack_ecache, cpu); - if (ecache->ct) - nf_ct_put(ecache->ct); - } -} -#else -static inline void nf_ct_event_cache_flush(void) {} -#endif /* CONFIG_NF_CONNTRACK_EVENTS */ +static unsigned int nf_conntrack_next_id; DEFINE_PER_CPU(struct ip_conntrack_stat, nf_conntrack_stat); EXPORT_PER_CPU_SYMBOL(nf_conntrack_stat); @@ -184,85 +120,6 @@ DEFINE_RWLOCK(nf_ct_cache_lock); /* This avoids calling kmem_cache_create() with same name simultaneously */ static DEFINE_MUTEX(nf_ct_cache_mutex); -extern struct nf_conntrack_protocol nf_conntrack_generic_protocol; -struct nf_conntrack_protocol * -__nf_ct_proto_find(u_int16_t l3proto, u_int8_t protocol) -{ - if (unlikely(l3proto >= AF_MAX || nf_ct_protos[l3proto] == NULL)) - return &nf_conntrack_generic_protocol; - - return nf_ct_protos[l3proto][protocol]; -} - -/* this is guaranteed to always return a valid protocol helper, since - * it falls back to generic_protocol */ -struct nf_conntrack_protocol * -nf_ct_proto_find_get(u_int16_t l3proto, u_int8_t protocol) -{ - struct nf_conntrack_protocol *p; - - preempt_disable(); - p = __nf_ct_proto_find(l3proto, protocol); - if (!try_module_get(p->me)) - p = &nf_conntrack_generic_protocol; - preempt_enable(); - - return p; -} - -void nf_ct_proto_put(struct nf_conntrack_protocol *p) -{ - module_put(p->me); -} - -struct nf_conntrack_l3proto * -nf_ct_l3proto_find_get(u_int16_t l3proto) -{ - struct nf_conntrack_l3proto *p; - - preempt_disable(); - p = __nf_ct_l3proto_find(l3proto); - if (!try_module_get(p->me)) - p = &nf_conntrack_generic_l3proto; - preempt_enable(); - - return p; -} - -void nf_ct_l3proto_put(struct nf_conntrack_l3proto *p) -{ - module_put(p->me); -} - -int -nf_ct_l3proto_try_module_get(unsigned short l3proto) -{ - int ret; - struct nf_conntrack_l3proto *p; - -retry: p = nf_ct_l3proto_find_get(l3proto); - if (p == &nf_conntrack_generic_l3proto) { - ret = request_module("nf_conntrack-%d", l3proto); - if (!ret) - goto retry; - - return -EPROTOTYPE; - } - - return 0; -} - -void nf_ct_l3proto_module_put(unsigned short l3proto) -{ - struct nf_conntrack_l3proto *p; - - preempt_disable(); - p = __nf_ct_l3proto_find(l3proto); - preempt_enable(); - - module_put(p->me); -} - static int nf_conntrack_hash_rnd_initted; static unsigned int nf_conntrack_hash_rnd; @@ -363,6 +220,7 @@ out_up_mutex: mutex_unlock(&nf_ct_cache_mutex); return ret; } +EXPORT_SYMBOL_GPL(nf_conntrack_register_cache); /* FIXME: In the current, only nf_conntrack_cleanup() can call this function. */ void nf_conntrack_unregister_cache(u_int32_t features) @@ -397,6 +255,7 @@ void nf_conntrack_unregister_cache(u_int32_t features) mutex_unlock(&nf_ct_cache_mutex); } +EXPORT_SYMBOL_GPL(nf_conntrack_unregister_cache); int nf_ct_get_tuple(const struct sk_buff *skb, @@ -406,7 +265,7 @@ nf_ct_get_tuple(const struct sk_buff *skb, u_int8_t protonum, struct nf_conntrack_tuple *tuple, const struct nf_conntrack_l3proto *l3proto, - const struct nf_conntrack_protocol *protocol) + const struct nf_conntrack_l4proto *l4proto) { NF_CT_TUPLE_U_BLANK(tuple); @@ -417,14 +276,15 @@ nf_ct_get_tuple(const struct sk_buff *skb, tuple->dst.protonum = protonum; tuple->dst.dir = IP_CT_DIR_ORIGINAL; - return protocol->pkt_to_tuple(skb, dataoff, tuple); + return l4proto->pkt_to_tuple(skb, dataoff, tuple); } +EXPORT_SYMBOL_GPL(nf_ct_get_tuple); int nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse, const struct nf_conntrack_tuple *orig, const struct nf_conntrack_l3proto *l3proto, - const struct nf_conntrack_protocol *protocol) + const struct nf_conntrack_l4proto *l4proto) { NF_CT_TUPLE_U_BLANK(inverse); @@ -435,111 +295,14 @@ nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse, inverse->dst.dir = !orig->dst.dir; inverse->dst.protonum = orig->dst.protonum; - return protocol->invert_tuple(inverse, orig); -} - -/* nf_conntrack_expect helper functions */ -void nf_ct_unlink_expect(struct nf_conntrack_expect *exp) -{ - struct nf_conn_help *master_help = nfct_help(exp->master); - - NF_CT_ASSERT(master_help); - ASSERT_WRITE_LOCK(&nf_conntrack_lock); - NF_CT_ASSERT(!timer_pending(&exp->timeout)); - - list_del(&exp->list); - NF_CT_STAT_INC(expect_delete); - master_help->expecting--; - nf_conntrack_expect_put(exp); -} - -static void expectation_timed_out(unsigned long ul_expect) -{ - struct nf_conntrack_expect *exp = (void *)ul_expect; - - write_lock_bh(&nf_conntrack_lock); - nf_ct_unlink_expect(exp); - write_unlock_bh(&nf_conntrack_lock); - nf_conntrack_expect_put(exp); -} - -struct nf_conntrack_expect * -__nf_conntrack_expect_find(const struct nf_conntrack_tuple *tuple) -{ - struct nf_conntrack_expect *i; - - list_for_each_entry(i, &nf_conntrack_expect_list, list) { - if (nf_ct_tuple_mask_cmp(tuple, &i->tuple, &i->mask)) { - atomic_inc(&i->use); - return i; - } - } - return NULL; -} - -/* Just find a expectation corresponding to a tuple. */ -struct nf_conntrack_expect * -nf_conntrack_expect_find(const struct nf_conntrack_tuple *tuple) -{ - struct nf_conntrack_expect *i; - - read_lock_bh(&nf_conntrack_lock); - i = __nf_conntrack_expect_find(tuple); - read_unlock_bh(&nf_conntrack_lock); - - return i; -} - -/* If an expectation for this connection is found, it gets delete from - * global list then returned. */ -static struct nf_conntrack_expect * -find_expectation(const struct nf_conntrack_tuple *tuple) -{ - struct nf_conntrack_expect *i; - - list_for_each_entry(i, &nf_conntrack_expect_list, list) { - /* If master is not in hash table yet (ie. packet hasn't left - this machine yet), how can other end know about expected? - Hence these are not the droids you are looking for (if - master ct never got confirmed, we'd hold a reference to it - and weird things would happen to future packets). */ - if (nf_ct_tuple_mask_cmp(tuple, &i->tuple, &i->mask) - && nf_ct_is_confirmed(i->master)) { - if (i->flags & NF_CT_EXPECT_PERMANENT) { - atomic_inc(&i->use); - return i; - } else if (del_timer(&i->timeout)) { - nf_ct_unlink_expect(i); - return i; - } - } - } - return NULL; -} - -/* delete all expectations for this conntrack */ -void nf_ct_remove_expectations(struct nf_conn *ct) -{ - struct nf_conntrack_expect *i, *tmp; - struct nf_conn_help *help = nfct_help(ct); - - /* Optimization: most connection never expect any others. */ - if (!help || help->expecting == 0) - return; - - list_for_each_entry_safe(i, tmp, &nf_conntrack_expect_list, list) { - if (i->master == ct && del_timer(&i->timeout)) { - nf_ct_unlink_expect(i); - nf_conntrack_expect_put(i); - } - } + return l4proto->invert_tuple(inverse, orig); } +EXPORT_SYMBOL_GPL(nf_ct_invert_tuple); static void clean_from_lists(struct nf_conn *ct) { DEBUGP("clean_from_lists(%p)\n", ct); - ASSERT_WRITE_LOCK(&nf_conntrack_lock); list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list); list_del(&ct->tuplehash[IP_CT_DIR_REPLY].list); @@ -551,8 +314,9 @@ static void destroy_conntrack(struct nf_conntrack *nfct) { struct nf_conn *ct = (struct nf_conn *)nfct; + struct nf_conn_help *help = nfct_help(ct); struct nf_conntrack_l3proto *l3proto; - struct nf_conntrack_protocol *proto; + struct nf_conntrack_l4proto *l4proto; DEBUGP("destroy_conntrack(%p)\n", ct); NF_CT_ASSERT(atomic_read(&nfct->use) == 0); @@ -561,6 +325,9 @@ destroy_conntrack(struct nf_conntrack *nfct) nf_conntrack_event(IPCT_DESTROY, ct); set_bit(IPS_DYING_BIT, &ct->status); + if (help && help->helper && help->helper->destroy) + help->helper->destroy(ct); + /* To make sure we don't get any weird locking issues here: * destroy_conntrack() MUST NOT be called with a write lock * to nf_conntrack_lock!!! -HW */ @@ -568,9 +335,9 @@ destroy_conntrack(struct nf_conntrack *nfct) if (l3proto && l3proto->destroy) l3proto->destroy(ct); - proto = __nf_ct_proto_find(ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.l3num, ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum); - if (proto && proto->destroy) - proto->destroy(ct); + l4proto = __nf_ct_l4proto_find(ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.l3num, ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum); + if (l4proto && l4proto->destroy) + l4proto->destroy(ct); if (nf_conntrack_destroyed) nf_conntrack_destroyed(ct); @@ -618,7 +385,6 @@ __nf_conntrack_find(const struct nf_conntrack_tuple *tuple, struct nf_conntrack_tuple_hash *h; unsigned int hash = hash_conntrack(tuple); - ASSERT_READ_LOCK(&nf_conntrack_lock); list_for_each_entry(h, &nf_conntrack_hash[hash], list) { if (nf_ct_tuplehash_to_ctrack(h) != ignored_conntrack && nf_ct_tuple_equal(tuple, &h->tuple)) { @@ -630,6 +396,7 @@ __nf_conntrack_find(const struct nf_conntrack_tuple *tuple, return NULL; } +EXPORT_SYMBOL_GPL(__nf_conntrack_find); /* Find a connection corresponding to a tuple. */ struct nf_conntrack_tuple_hash * @@ -646,6 +413,7 @@ nf_conntrack_find_get(const struct nf_conntrack_tuple *tuple, return h; } +EXPORT_SYMBOL_GPL(nf_conntrack_find_get); static void __nf_conntrack_hash_insert(struct nf_conn *ct, unsigned int hash, @@ -669,6 +437,7 @@ void nf_conntrack_hash_insert(struct nf_conn *ct) __nf_conntrack_hash_insert(ct, hash, repl_hash); write_unlock_bh(&nf_conntrack_lock); } +EXPORT_SYMBOL_GPL(nf_conntrack_hash_insert); /* Confirm a connection given skb; places it in hash table */ int @@ -746,6 +515,7 @@ out: write_unlock_bh(&nf_conntrack_lock); return NF_DROP; } +EXPORT_SYMBOL_GPL(__nf_conntrack_confirm); /* Returns true if a connection correspondings to the tuple (required for NAT). */ @@ -761,6 +531,7 @@ nf_conntrack_tuple_taken(const struct nf_conntrack_tuple *tuple, return h != NULL; } +EXPORT_SYMBOL_GPL(nf_conntrack_tuple_taken); /* There's a small race here where we may free a just-assured connection. Too bad: we're in trouble anyway. */ @@ -794,53 +565,13 @@ static int early_drop(struct list_head *chain) return dropped; } -static struct nf_conntrack_helper * -__nf_ct_helper_find(const struct nf_conntrack_tuple *tuple) -{ - struct nf_conntrack_helper *h; - - list_for_each_entry(h, &helpers, list) { - if (nf_ct_tuple_mask_cmp(tuple, &h->tuple, &h->mask)) - return h; - } - return NULL; -} - -struct nf_conntrack_helper * -nf_ct_helper_find_get( const struct nf_conntrack_tuple *tuple) -{ - struct nf_conntrack_helper *helper; - - /* need nf_conntrack_lock to assure that helper exists until - * try_module_get() is called */ - read_lock_bh(&nf_conntrack_lock); - - helper = __nf_ct_helper_find(tuple); - if (helper) { - /* need to increase module usage count to assure helper will - * not go away while the caller is e.g. busy putting a - * conntrack in the hash that uses the helper */ - if (!try_module_get(helper->me)) - helper = NULL; - } - - read_unlock_bh(&nf_conntrack_lock); - - return helper; -} - -void nf_ct_helper_put(struct nf_conntrack_helper *helper) -{ - module_put(helper->me); -} - static struct nf_conn * __nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, const struct nf_conntrack_tuple *repl, - const struct nf_conntrack_l3proto *l3proto) + const struct nf_conntrack_l3proto *l3proto, + u_int32_t features) { struct nf_conn *conntrack = NULL; - u_int32_t features = 0; struct nf_conntrack_helper *helper; if (unlikely(!nf_conntrack_hash_rnd_initted)) { @@ -866,12 +597,13 @@ __nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, } /* find features needed by this conntrack. */ - features = l3proto->get_features(orig); + features |= l3proto->get_features(orig); /* FIXME: protect helper list per RCU */ read_lock_bh(&nf_conntrack_lock); helper = __nf_ct_helper_find(repl); - if (helper) + /* NAT might want to assign a helper later */ + if (helper || features & NF_CT_F_NAT) features |= NF_CT_F_HELP; read_unlock_bh(&nf_conntrack_lock); @@ -893,12 +625,6 @@ __nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, memset(conntrack, 0, nf_ct_cache[features].size); conntrack->features = features; - if (helper) { - struct nf_conn_help *help = nfct_help(conntrack); - NF_CT_ASSERT(help); - help->helper = helper; - } - atomic_set(&conntrack->ct_general.use, 1); conntrack->ct_general.destroy = destroy_conntrack; conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *orig; @@ -922,8 +648,9 @@ struct nf_conn *nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, struct nf_conntrack_l3proto *l3proto; l3proto = __nf_ct_l3proto_find(orig->src.l3num); - return __nf_conntrack_alloc(orig, repl, l3proto); + return __nf_conntrack_alloc(orig, repl, l3proto, 0); } +EXPORT_SYMBOL_GPL(nf_conntrack_alloc); void nf_conntrack_free(struct nf_conn *conntrack) { @@ -934,32 +661,40 @@ void nf_conntrack_free(struct nf_conn *conntrack) kmem_cache_free(nf_ct_cache[features].cachep, conntrack); atomic_dec(&nf_conntrack_count); } +EXPORT_SYMBOL_GPL(nf_conntrack_free); /* Allocate a new conntrack: we return -ENOMEM if classification failed due to stress. Otherwise it really is unclassifiable. */ static struct nf_conntrack_tuple_hash * init_conntrack(const struct nf_conntrack_tuple *tuple, struct nf_conntrack_l3proto *l3proto, - struct nf_conntrack_protocol *protocol, + struct nf_conntrack_l4proto *l4proto, struct sk_buff *skb, unsigned int dataoff) { struct nf_conn *conntrack; struct nf_conntrack_tuple repl_tuple; struct nf_conntrack_expect *exp; + u_int32_t features = 0; - if (!nf_ct_invert_tuple(&repl_tuple, tuple, l3proto, protocol)) { + if (!nf_ct_invert_tuple(&repl_tuple, tuple, l3proto, l4proto)) { DEBUGP("Can't invert tuple.\n"); return NULL; } - conntrack = __nf_conntrack_alloc(tuple, &repl_tuple, l3proto); + read_lock_bh(&nf_conntrack_lock); + exp = __nf_conntrack_expect_find(tuple); + if (exp && exp->helper) + features = NF_CT_F_HELP; + read_unlock_bh(&nf_conntrack_lock); + + conntrack = __nf_conntrack_alloc(tuple, &repl_tuple, l3proto, features); if (conntrack == NULL || IS_ERR(conntrack)) { DEBUGP("Can't allocate conntrack.\n"); return (struct nf_conntrack_tuple_hash *)conntrack; } - if (!protocol->new(conntrack, skb, dataoff)) { + if (!l4proto->new(conntrack, skb, dataoff)) { nf_conntrack_free(conntrack); DEBUGP("init conntrack: can't track with proto module\n"); return NULL; @@ -974,6 +709,8 @@ init_conntrack(const struct nf_conntrack_tuple *tuple, /* Welcome, Mr. Bond. We've been expecting you... */ __set_bit(IPS_EXPECTED_BIT, &conntrack->status); conntrack->master = exp->master; + if (exp->helper) + nfct_help(conntrack)->helper = exp->helper; #ifdef CONFIG_NF_CONNTRACK_MARK conntrack->mark = exp->master->mark; #endif @@ -982,8 +719,13 @@ init_conntrack(const struct nf_conntrack_tuple *tuple, #endif nf_conntrack_get(&conntrack->master->ct_general); NF_CT_STAT_INC(expect_new); - } else + } else { + struct nf_conn_help *help = nfct_help(conntrack); + + if (help) + help->helper = __nf_ct_helper_find(&repl_tuple); NF_CT_STAT_INC(new); + } /* Overload tuple linked list to put us in unconfirmed list. */ list_add(&conntrack->tuplehash[IP_CT_DIR_ORIGINAL].list, &unconfirmed); @@ -1006,7 +748,7 @@ resolve_normal_ct(struct sk_buff *skb, u_int16_t l3num, u_int8_t protonum, struct nf_conntrack_l3proto *l3proto, - struct nf_conntrack_protocol *proto, + struct nf_conntrack_l4proto *l4proto, int *set_reply, enum ip_conntrack_info *ctinfo) { @@ -1016,7 +758,7 @@ resolve_normal_ct(struct sk_buff *skb, if (!nf_ct_get_tuple(skb, (unsigned int)(skb->nh.raw - skb->data), dataoff, l3num, protonum, &tuple, l3proto, - proto)) { + l4proto)) { DEBUGP("resolve_normal_ct: Can't get tuple\n"); return NULL; } @@ -1024,7 +766,7 @@ resolve_normal_ct(struct sk_buff *skb, /* look for tuple match */ h = nf_conntrack_find_get(&tuple, NULL); if (!h) { - h = init_conntrack(&tuple, l3proto, proto, skb, dataoff); + h = init_conntrack(&tuple, l3proto, l4proto, skb, dataoff); if (!h) return NULL; if (IS_ERR(h)) @@ -1062,7 +804,7 @@ nf_conntrack_in(int pf, unsigned int hooknum, struct sk_buff **pskb) struct nf_conn *ct; enum ip_conntrack_info ctinfo; struct nf_conntrack_l3proto *l3proto; - struct nf_conntrack_protocol *proto; + struct nf_conntrack_l4proto *l4proto; unsigned int dataoff; u_int8_t protonum; int set_reply = 0; @@ -1080,19 +822,19 @@ nf_conntrack_in(int pf, unsigned int hooknum, struct sk_buff **pskb) return -ret; } - proto = __nf_ct_proto_find((u_int16_t)pf, protonum); + l4proto = __nf_ct_l4proto_find((u_int16_t)pf, protonum); /* It may be an special packet, error, unclean... * inverse of the return code tells to the netfilter * core what to do with the packet. */ - if (proto->error != NULL && - (ret = proto->error(*pskb, dataoff, &ctinfo, pf, hooknum)) <= 0) { + if (l4proto->error != NULL && + (ret = l4proto->error(*pskb, dataoff, &ctinfo, pf, hooknum)) <= 0) { NF_CT_STAT_INC(error); NF_CT_STAT_INC(invalid); return -ret; } - ct = resolve_normal_ct(*pskb, dataoff, pf, protonum, l3proto, proto, + ct = resolve_normal_ct(*pskb, dataoff, pf, protonum, l3proto, l4proto, &set_reply, &ctinfo); if (!ct) { /* Not valid part of a connection */ @@ -1108,7 +850,7 @@ nf_conntrack_in(int pf, unsigned int hooknum, struct sk_buff **pskb) NF_CT_ASSERT((*pskb)->nfct); - ret = proto->packet(ct, *pskb, dataoff, ctinfo, pf, hooknum); + ret = l4proto->packet(ct, *pskb, dataoff, ctinfo, pf, hooknum); if (ret < 0) { /* Invalid: inverse of the return code tells * the netfilter core what to do */ @@ -1124,255 +866,38 @@ nf_conntrack_in(int pf, unsigned int hooknum, struct sk_buff **pskb) return ret; } +EXPORT_SYMBOL_GPL(nf_conntrack_in); int nf_ct_invert_tuplepr(struct nf_conntrack_tuple *inverse, const struct nf_conntrack_tuple *orig) { return nf_ct_invert_tuple(inverse, orig, __nf_ct_l3proto_find(orig->src.l3num), - __nf_ct_proto_find(orig->src.l3num, + __nf_ct_l4proto_find(orig->src.l3num, orig->dst.protonum)); } +EXPORT_SYMBOL_GPL(nf_ct_invert_tuplepr); -/* Would two expected things clash? */ -static inline int expect_clash(const struct nf_conntrack_expect *a, - const struct nf_conntrack_expect *b) -{ - /* Part covered by intersection of masks must be unequal, - otherwise they clash */ - struct nf_conntrack_tuple intersect_mask; - int count; - - intersect_mask.src.l3num = a->mask.src.l3num & b->mask.src.l3num; - intersect_mask.src.u.all = a->mask.src.u.all & b->mask.src.u.all; - intersect_mask.dst.u.all = a->mask.dst.u.all & b->mask.dst.u.all; - intersect_mask.dst.protonum = a->mask.dst.protonum - & b->mask.dst.protonum; - - for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ - intersect_mask.src.u3.all[count] = - a->mask.src.u3.all[count] & b->mask.src.u3.all[count]; - } - - for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ - intersect_mask.dst.u3.all[count] = - a->mask.dst.u3.all[count] & b->mask.dst.u3.all[count]; - } - - return nf_ct_tuple_mask_cmp(&a->tuple, &b->tuple, &intersect_mask); -} - -static inline int expect_matches(const struct nf_conntrack_expect *a, - const struct nf_conntrack_expect *b) -{ - return a->master == b->master - && nf_ct_tuple_equal(&a->tuple, &b->tuple) - && nf_ct_tuple_equal(&a->mask, &b->mask); -} - -/* Generally a bad idea to call this: could have matched already. */ -void nf_conntrack_unexpect_related(struct nf_conntrack_expect *exp) -{ - struct nf_conntrack_expect *i; - - write_lock_bh(&nf_conntrack_lock); - /* choose the the oldest expectation to evict */ - list_for_each_entry_reverse(i, &nf_conntrack_expect_list, list) { - if (expect_matches(i, exp) && del_timer(&i->timeout)) { - nf_ct_unlink_expect(i); - write_unlock_bh(&nf_conntrack_lock); - nf_conntrack_expect_put(i); - return; - } - } - write_unlock_bh(&nf_conntrack_lock); -} - -/* We don't increase the master conntrack refcount for non-fulfilled - * conntracks. During the conntrack destruction, the expectations are - * always killed before the conntrack itself */ -struct nf_conntrack_expect *nf_conntrack_expect_alloc(struct nf_conn *me) -{ - struct nf_conntrack_expect *new; - - new = kmem_cache_alloc(nf_conntrack_expect_cachep, GFP_ATOMIC); - if (!new) { - DEBUGP("expect_related: OOM allocating expect\n"); - return NULL; - } - new->master = me; - atomic_set(&new->use, 1); - return new; -} - -void nf_conntrack_expect_put(struct nf_conntrack_expect *exp) -{ - if (atomic_dec_and_test(&exp->use)) - kmem_cache_free(nf_conntrack_expect_cachep, exp); -} - -static void nf_conntrack_expect_insert(struct nf_conntrack_expect *exp) -{ - struct nf_conn_help *master_help = nfct_help(exp->master); - - atomic_inc(&exp->use); - master_help->expecting++; - list_add(&exp->list, &nf_conntrack_expect_list); - - init_timer(&exp->timeout); - exp->timeout.data = (unsigned long)exp; - exp->timeout.function = expectation_timed_out; - exp->timeout.expires = jiffies + master_help->helper->timeout * HZ; - add_timer(&exp->timeout); - - exp->id = ++nf_conntrack_expect_next_id; - atomic_inc(&exp->use); - NF_CT_STAT_INC(expect_create); -} - -/* Race with expectations being used means we could have none to find; OK. */ -static void evict_oldest_expect(struct nf_conn *master) -{ - struct nf_conntrack_expect *i; - - list_for_each_entry_reverse(i, &nf_conntrack_expect_list, list) { - if (i->master == master) { - if (del_timer(&i->timeout)) { - nf_ct_unlink_expect(i); - nf_conntrack_expect_put(i); - } - break; - } - } -} - -static inline int refresh_timer(struct nf_conntrack_expect *i) -{ - struct nf_conn_help *master_help = nfct_help(i->master); - - if (!del_timer(&i->timeout)) - return 0; - - i->timeout.expires = jiffies + master_help->helper->timeout*HZ; - add_timer(&i->timeout); - return 1; -} - -int nf_conntrack_expect_related(struct nf_conntrack_expect *expect) +/* Alter reply tuple (maybe alter helper). This is for NAT, and is + implicitly racy: see __nf_conntrack_confirm */ +void nf_conntrack_alter_reply(struct nf_conn *ct, + const struct nf_conntrack_tuple *newreply) { - struct nf_conntrack_expect *i; - struct nf_conn *master = expect->master; - struct nf_conn_help *master_help = nfct_help(master); - int ret; - - NF_CT_ASSERT(master_help); - - DEBUGP("nf_conntrack_expect_related %p\n", related_to); - DEBUGP("tuple: "); NF_CT_DUMP_TUPLE(&expect->tuple); - DEBUGP("mask: "); NF_CT_DUMP_TUPLE(&expect->mask); - - write_lock_bh(&nf_conntrack_lock); - list_for_each_entry(i, &nf_conntrack_expect_list, list) { - if (expect_matches(i, expect)) { - /* Refresh timer: if it's dying, ignore.. */ - if (refresh_timer(i)) { - ret = 0; - goto out; - } - } else if (expect_clash(i, expect)) { - ret = -EBUSY; - goto out; - } - } - /* Will be over limit? */ - if (master_help->helper->max_expected && - master_help->expecting >= master_help->helper->max_expected) - evict_oldest_expect(master); - - nf_conntrack_expect_insert(expect); - nf_conntrack_expect_event(IPEXP_NEW, expect); - ret = 0; -out: - write_unlock_bh(&nf_conntrack_lock); - return ret; -} - -int nf_conntrack_helper_register(struct nf_conntrack_helper *me) -{ - int ret; - BUG_ON(me->timeout == 0); - - ret = nf_conntrack_register_cache(NF_CT_F_HELP, "nf_conntrack:help", - sizeof(struct nf_conn) - + sizeof(struct nf_conn_help) - + __alignof__(struct nf_conn_help)); - if (ret < 0) { - printk(KERN_ERR "nf_conntrack_helper_reigster: Unable to create slab cache for conntracks\n"); - return ret; - } - write_lock_bh(&nf_conntrack_lock); - list_add(&me->list, &helpers); - write_unlock_bh(&nf_conntrack_lock); - - return 0; -} - -struct nf_conntrack_helper * -__nf_conntrack_helper_find_byname(const char *name) -{ - struct nf_conntrack_helper *h; - - list_for_each_entry(h, &helpers, list) { - if (!strcmp(h->name, name)) - return h; - } - - return NULL; -} - -static inline void unhelp(struct nf_conntrack_tuple_hash *i, - const struct nf_conntrack_helper *me) -{ - struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(i); struct nf_conn_help *help = nfct_help(ct); - if (help && help->helper == me) { - nf_conntrack_event(IPCT_HELPER, ct); - help->helper = NULL; - } -} - -void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me) -{ - unsigned int i; - struct nf_conntrack_tuple_hash *h; - struct nf_conntrack_expect *exp, *tmp; - - /* Need write lock here, to delete helper. */ write_lock_bh(&nf_conntrack_lock); - list_del(&me->list); - - /* Get rid of expectations */ - list_for_each_entry_safe(exp, tmp, &nf_conntrack_expect_list, list) { - struct nf_conn_help *help = nfct_help(exp->master); - if (help->helper == me && del_timer(&exp->timeout)) { - nf_ct_unlink_expect(exp); - nf_conntrack_expect_put(exp); - } - } + /* Should be unconfirmed, so not in hash table yet */ + NF_CT_ASSERT(!nf_ct_is_confirmed(ct)); - /* Get rid of expecteds, set helpers to NULL. */ - list_for_each_entry(h, &unconfirmed, list) - unhelp(h, me); - for (i = 0; i < nf_conntrack_htable_size; i++) { - list_for_each_entry(h, &nf_conntrack_hash[i], list) - unhelp(h, me); - } - write_unlock_bh(&nf_conntrack_lock); + DEBUGP("Altering reply tuple of %p to ", ct); + NF_CT_DUMP_TUPLE(newreply); - /* Someone could be still looking at the helper in a bh. */ - synchronize_net(); + ct->tuplehash[IP_CT_DIR_REPLY].tuple = *newreply; + if (!ct->master && help && help->expecting == 0) + help->helper = __nf_ct_helper_find(newreply); + write_unlock_bh(&nf_conntrack_lock); } +EXPORT_SYMBOL_GPL(nf_conntrack_alter_reply); /* Refresh conntrack for this many jiffies and do accounting if do_acct is 1 */ void __nf_ct_refresh_acct(struct nf_conn *ct, @@ -1399,9 +924,14 @@ void __nf_ct_refresh_acct(struct nf_conn *ct, ct->timeout.expires = extra_jiffies; event = IPCT_REFRESH; } else { - /* Need del_timer for race avoidance (may already be dying). */ - if (del_timer(&ct->timeout)) { - ct->timeout.expires = jiffies + extra_jiffies; + unsigned long newtime = jiffies + extra_jiffies; + + /* Only update the timeout if the new timeout is at least + HZ jiffies from the old timeout. Need del_timer for race + avoidance (may already be dying). */ + if (newtime - ct->timeout.expires >= HZ + && del_timer(&ct->timeout)) { + ct->timeout.expires = newtime; add_timer(&ct->timeout); event = IPCT_REFRESH; } @@ -1412,9 +942,10 @@ void __nf_ct_refresh_acct(struct nf_conn *ct, ct->counters[CTINFO2DIR(ctinfo)].packets++; ct->counters[CTINFO2DIR(ctinfo)].bytes += skb->len - (unsigned int)(skb->nh.raw - skb->data); - if ((ct->counters[CTINFO2DIR(ctinfo)].packets & 0x80000000) - || (ct->counters[CTINFO2DIR(ctinfo)].bytes & 0x80000000)) - event |= IPCT_COUNTER_FILLING; + + if ((ct->counters[CTINFO2DIR(ctinfo)].packets & 0x80000000) + || (ct->counters[CTINFO2DIR(ctinfo)].bytes & 0x80000000)) + event |= IPCT_COUNTER_FILLING; } #endif @@ -1424,6 +955,7 @@ void __nf_ct_refresh_acct(struct nf_conn *ct, if (event) nf_conntrack_event_cache(event, skb); } +EXPORT_SYMBOL_GPL(__nf_ct_refresh_acct); #if defined(CONFIG_NF_CT_NETLINK) || \ defined(CONFIG_NF_CT_NETLINK_MODULE) @@ -1448,6 +980,7 @@ int nf_ct_port_tuple_to_nfattr(struct sk_buff *skb, nfattr_failure: return -1; } +EXPORT_SYMBOL_GPL(nf_ct_port_tuple_to_nfattr); static const size_t cta_min_proto[CTA_PROTO_MAX] = { [CTA_PROTO_SRC_PORT-1] = sizeof(u_int16_t), @@ -1463,13 +996,12 @@ int nf_ct_port_nfattr_to_tuple(struct nfattr *tb[], if (nfattr_bad_size(tb, CTA_PROTO_MAX, cta_min_proto)) return -EINVAL; - t->src.u.tcp.port = - *(u_int16_t *)NFA_DATA(tb[CTA_PROTO_SRC_PORT-1]); - t->dst.u.tcp.port = - *(u_int16_t *)NFA_DATA(tb[CTA_PROTO_DST_PORT-1]); + t->src.u.tcp.port = *(__be16 *)NFA_DATA(tb[CTA_PROTO_SRC_PORT-1]); + t->dst.u.tcp.port = *(__be16 *)NFA_DATA(tb[CTA_PROTO_DST_PORT-1]); return 0; } +EXPORT_SYMBOL_GPL(nf_ct_port_nfattr_to_tuple); #endif /* Used by ipt_REJECT and ip6t_REJECT. */ @@ -1490,6 +1022,7 @@ void __nf_conntrack_attach(struct sk_buff *nskb, struct sk_buff *skb) nskb->nfctinfo = ctinfo; nf_conntrack_get(nskb->nfct); } +EXPORT_SYMBOL_GPL(__nf_conntrack_attach); static inline int do_iter(const struct nf_conntrack_tuple_hash *i, @@ -1543,6 +1076,7 @@ nf_ct_iterate_cleanup(int (*iter)(struct nf_conn *i, void *data), void *data) nf_ct_put(ct); } } +EXPORT_SYMBOL_GPL(nf_ct_iterate_cleanup); static int kill_all(struct nf_conn *i, void *data) { @@ -1562,6 +1096,7 @@ void nf_conntrack_flush() { nf_ct_iterate_cleanup(kill_all, NULL); } +EXPORT_SYMBOL_GPL(nf_conntrack_flush); /* Mishearing the voices in his head, our hero wonders how he's supposed to kill the mall. */ @@ -1599,6 +1134,8 @@ void nf_conntrack_cleanup(void) free_conntrack_hash(nf_conntrack_hash, nf_conntrack_vmalloc, nf_conntrack_htable_size); + nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_generic); + /* free l3proto protocol tables */ for (i = 0; i < PF_MAX; i++) if (nf_ct_protos[i]) { @@ -1724,10 +1261,14 @@ int __init nf_conntrack_init(void) goto err_free_conntrack_slab; } + ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_generic); + if (ret < 0) + goto out_free_expect_slab; + /* Don't NEED lock here, but good form anyway. */ write_lock_bh(&nf_conntrack_lock); - for (i = 0; i < PF_MAX; i++) - nf_ct_l3protos[i] = &nf_conntrack_generic_l3proto; + for (i = 0; i < AF_MAX; i++) + nf_ct_l3protos[i] = &nf_conntrack_l3proto_generic; write_unlock_bh(&nf_conntrack_lock); /* For use by REJECT target */ @@ -1741,6 +1282,8 @@ int __init nf_conntrack_init(void) return ret; +out_free_expect_slab: + kmem_cache_destroy(nf_conntrack_expect_cachep); err_free_conntrack_slab: nf_conntrack_unregister_cache(NF_CT_F_BASIC); err_free_hash: |