inet_timewait_sock.c 8.84 KB
Newer Older
1 2 3 4 5 6 7 8 9 10
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		Generic TIME_WAIT sockets functions
 *
 *		From code orinally in TCP
 */

11
#include <linux/kernel.h>
12
#include <linux/slab.h>
13
#include <linux/module.h>
14 15
#include <net/inet_hashtables.h>
#include <net/inet_timewait_sock.h>
16
#include <net/ip.h>
17

18

19 20 21 22 23 24 25 26
/**
 *	inet_twsk_bind_unhash - unhash a timewait socket from bind hash
 *	@tw: timewait socket
 *	@hashinfo: hashinfo pointer
 *
 *	unhash a timewait socket from bind hash, if hashed.
 *	bind hash lock must be held by caller.
 *	Returns 1 if caller should call inet_twsk_put() after lock release.
27
 */
28
void inet_twsk_bind_unhash(struct inet_timewait_sock *tw,
29 30 31 32 33
			  struct inet_hashinfo *hashinfo)
{
	struct inet_bind_bucket *tb = tw->tw_tb;

	if (!tb)
34
		return;
35 36 37 38

	__hlist_del(&tw->tw_bind_node);
	tw->tw_tb = NULL;
	inet_bind_bucket_destroy(hashinfo->bind_bucket_cachep, tb);
39
	__sock_put((struct sock *)tw);
40 41
}

42
/* Must be called with locally disabled BHs. */
43
static void inet_twsk_kill(struct inet_timewait_sock *tw)
44
{
45
	struct inet_hashinfo *hashinfo = tw->tw_dr->hashinfo;
46
	spinlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
47
	struct inet_bind_hashbucket *bhead;
48

49
	spin_lock(lock);
50
	sk_nulls_del_node_init_rcu((struct sock *)tw);
51
	spin_unlock(lock);
52 53

	/* Disassociate with bind bucket. */
54 55
	bhead = &hashinfo->bhash[inet_bhashfn(twsk_net(tw), tw->tw_num,
			hashinfo->bhash_size)];
56

57
	spin_lock(&bhead->lock);
58
	inet_twsk_bind_unhash(tw, hashinfo);
59
	spin_unlock(&bhead->lock);
60

61 62
	atomic_dec(&tw->tw_dr->tw_count);
	inet_twsk_put(tw);
63 64
}

Eric Dumazet's avatar
Eric Dumazet committed
65
void inet_twsk_free(struct inet_timewait_sock *tw)
66
{
67 68
	struct module *owner = tw->tw_prot->owner;
	twsk_destructor((struct sock *)tw);
69
#ifdef SOCK_REFCNT_DEBUG
70
	pr_debug("%s timewait_sock %p released\n", tw->tw_prot->name, tw);
71
#endif
72 73 74 75 76 77
	kmem_cache_free(tw->tw_prot->twsk_prot->twsk_slab, tw);
	module_put(owner);
}

void inet_twsk_put(struct inet_timewait_sock *tw)
{
78
	if (refcount_dec_and_test(&tw->tw_refcnt))
79
		inet_twsk_free(tw);
80 81 82
}
EXPORT_SYMBOL_GPL(inet_twsk_put);

Eric Dumazet's avatar
Eric Dumazet committed
83 84 85 86 87 88 89 90 91 92 93 94
static void inet_twsk_add_node_rcu(struct inet_timewait_sock *tw,
				   struct hlist_nulls_head *list)
{
	hlist_nulls_add_head_rcu(&tw->tw_node, list);
}

static void inet_twsk_add_bind_node(struct inet_timewait_sock *tw,
				    struct hlist_head *list)
{
	hlist_add_head(&tw->tw_bind_node, list);
}

95
/*
96
 * Enter the time wait state. This is called with locally disabled BH.
97 98 99
 * Essentially we whip up a timewait bucket, copy the relevant info into it
 * from the SK, and mess with hash chains and list linkage.
 */
100
void inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
101 102 103
			   struct inet_hashinfo *hashinfo)
{
	const struct inet_sock *inet = inet_sk(sk);
104
	const struct inet_connection_sock *icsk = inet_csk(sk);
105
	struct inet_ehash_bucket *ehead = inet_ehash_bucket(hashinfo, sk->sk_hash);
106
	spinlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
107 108 109 110 111
	struct inet_bind_hashbucket *bhead;
	/* Step 1: Put TW into bind hash. Original socket stays there too.
	   Note, that any socket with inet->num != 0 MUST be bound in
	   binding cache, even if it is closed.
	 */
112
	bhead = &hashinfo->bhash[inet_bhashfn(twsk_net(tw), inet->inet_num,
113
			hashinfo->bhash_size)];
114
	spin_lock(&bhead->lock);
115
	tw->tw_tb = icsk->icsk_bind_hash;
116
	WARN_ON(!icsk->icsk_bind_hash);
117 118 119
	inet_twsk_add_bind_node(tw, &tw->tw_tb->owners);
	spin_unlock(&bhead->lock);

120
	spin_lock(lock);
121

Eric Dumazet's avatar
Eric Dumazet committed
122
	inet_twsk_add_node_rcu(tw, &ehead->chain);
123

Eric Dumazet's avatar
Eric Dumazet committed
124
	/* Step 3: Remove SK from hash chain */
125 126
	if (__sk_nulls_del_node_init_rcu(sk))
		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
127

128
	spin_unlock(lock);
129 130 131 132 133 134 135 136 137 138 139

	/* tw_refcnt is set to 3 because we have :
	 * - one reference for bhash chain.
	 * - one reference for ehash chain.
	 * - one reference for timer.
	 * We can use atomic_set() because prior spin_lock()/spin_unlock()
	 * committed into memory all tw fields.
	 * Also note that after this point, we lost our implicit reference
	 * so we are not allowed to use tw anymore.
	 */
	refcount_set(&tw->tw_refcnt, 3);
140
}
141
EXPORT_SYMBOL_GPL(inet_twsk_hashdance);
142

143
static void tw_timer_handler(struct timer_list *t)
144
{
145
	struct inet_timewait_sock *tw = from_timer(tw, t, tw_timer);
146 147

	if (tw->tw_kill)
148
		__NET_INC_STATS(twsk_net(tw), LINUX_MIB_TIMEWAITKILLED);
149
	else
150
		__NET_INC_STATS(twsk_net(tw), LINUX_MIB_TIMEWAITED);
151 152 153 154 155 156 157 158 159 160 161 162 163 164
	inet_twsk_kill(tw);
}

struct inet_timewait_sock *inet_twsk_alloc(const struct sock *sk,
					   struct inet_timewait_death_row *dr,
					   const int state)
{
	struct inet_timewait_sock *tw;

	if (atomic_read(&dr->tw_count) >= dr->sysctl_max_tw_buckets)
		return NULL;

	tw = kmem_cache_alloc(sk->sk_prot_creator->twsk_prot->twsk_slab,
			      GFP_ATOMIC);
165
	if (tw) {
166 167
		const struct inet_sock *inet = inet_sk(sk);

168
		tw->tw_dr	    = dr;
169
		/* Give us an identity. */
170 171
		tw->tw_daddr	    = inet->inet_daddr;
		tw->tw_rcv_saddr    = inet->inet_rcv_saddr;
172
		tw->tw_bound_dev_if = sk->sk_bound_dev_if;
173
		tw->tw_tos	    = inet->tos;
174
		tw->tw_num	    = inet->inet_num;
175 176
		tw->tw_state	    = TCP_TIME_WAIT;
		tw->tw_substate	    = state;
177 178
		tw->tw_sport	    = inet->inet_sport;
		tw->tw_dport	    = inet->inet_dport;
179 180
		tw->tw_family	    = sk->sk_family;
		tw->tw_reuse	    = sk->sk_reuse;
181
		tw->tw_reuseport    = sk->sk_reuseport;
182
		tw->tw_hash	    = sk->sk_hash;
183
		tw->tw_ipv6only	    = 0;
184
		tw->tw_transparent  = inet->transparent;
185
		tw->tw_prot	    = sk->sk_prot_creator;
186
		atomic64_set(&tw->tw_cookie, atomic64_read(&sk->sk_cookie));
187
		twsk_net_set(tw, sock_net(sk));
188
		timer_setup(&tw->tw_timer, tw_timer_handler, TIMER_PINNED);
189 190 191 192 193
		/*
		 * Because we use RCU lookups, we should not set tw_refcnt
		 * to a non null value before everything is setup for this
		 * timewait socket.
		 */
194
		refcount_set(&tw->tw_refcnt, 0);
195

196
		__module_get(tw->tw_prot->owner);
197 198 199 200
	}

	return tw;
}
201 202 203 204 205 206
EXPORT_SYMBOL_GPL(inet_twsk_alloc);

/* These are always called from BH context.  See callers in
 * tcp_input.c to verify this.
 */

207 208 209 210 211
/* This is for handling early-kills of TIME_WAIT sockets.
 * Warning : consume reference.
 * Caller should not access tw anymore.
 */
void inet_twsk_deschedule_put(struct inet_timewait_sock *tw)
212
{
213 214
	if (del_timer_sync(&tw->tw_timer))
		inet_twsk_kill(tw);
215
	inet_twsk_put(tw);
216
}
217
EXPORT_SYMBOL(inet_twsk_deschedule_put);
218

219
void __inet_twsk_schedule(struct inet_timewait_sock *tw, int timeo, bool rearm)
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245
{
	/* timeout := RTO * 3.5
	 *
	 * 3.5 = 1+2+0.5 to wait for two retransmits.
	 *
	 * RATIONALE: if FIN arrived and we entered TIME-WAIT state,
	 * our ACK acking that FIN can be lost. If N subsequent retransmitted
	 * FINs (or previous seqments) are lost (probability of such event
	 * is p^(N+1), where p is probability to lose single packet and
	 * time to detect the loss is about RTO*(2^N - 1) with exponential
	 * backoff). Normal timewait length is calculated so, that we
	 * waited at least for one retransmitted FIN (maximal RTO is 120sec).
	 * [ BTW Linux. following BSD, violates this requirement waiting
	 *   only for 60sec, we should wait at least for 240 secs.
	 *   Well, 240 consumes too much of resources 8)
	 * ]
	 * This interval is not reduced to catch old duplicate and
	 * responces to our wandering segments living for two MSLs.
	 * However, if we use PAWS to detect
	 * old duplicates, we can reduce the interval to bounds required
	 * by RTO, rather than MSL. So, if peer understands PAWS, we
	 * kill tw bucket after 3.5*RTO (it is important that this number
	 * is greater than TS tick!) and detect old duplicates with help
	 * of PAWS.
	 */

246
	tw->tw_kill = timeo <= 4*HZ;
247
	if (!rearm) {
248
		BUG_ON(mod_timer(&tw->tw_timer, jiffies + timeo));
249
		atomic_inc(&tw->tw_dr->tw_count);
250 251
	} else {
		mod_timer_pending(&tw->tw_timer, jiffies + timeo);
252 253
	}
}
254
EXPORT_SYMBOL_GPL(__inet_twsk_schedule);
255

256
void inet_twsk_purge(struct inet_hashinfo *hashinfo, int family)
257 258 259
{
	struct inet_timewait_sock *tw;
	struct sock *sk;
260
	struct hlist_nulls_node *node;
261
	unsigned int slot;
262

263 264 265
	for (slot = 0; slot <= hashinfo->ehash_mask; slot++) {
		struct inet_ehash_bucket *head = &hashinfo->ehash[slot];
restart_rcu:
266
		cond_resched();
267
		rcu_read_lock();
268
restart:
Eric Dumazet's avatar
Eric Dumazet committed
269 270 271
		sk_nulls_for_each_rcu(sk, node, &head->chain) {
			if (sk->sk_state != TCP_TIME_WAIT)
				continue;
272
			tw = inet_twsk(sk);
273
			if ((tw->tw_family != family) ||
274
				refcount_read(&twsk_net(tw)->count))
275 276
				continue;

277
			if (unlikely(!refcount_inc_not_zero(&tw->tw_refcnt)))
278 279
				continue;

280
			if (unlikely((tw->tw_family != family) ||
281
				     refcount_read(&twsk_net(tw)->count))) {
282 283 284 285 286
				inet_twsk_put(tw);
				goto restart;
			}

			rcu_read_unlock();
287
			local_bh_disable();
288
			inet_twsk_deschedule_put(tw);
289
			local_bh_enable();
290
			goto restart_rcu;
291
		}
292 293 294 295 296 297 298
		/* If the nulls value we got at the end of this lookup is
		 * not the expected one, we must restart lookup.
		 * We probably met an item that was moved to another chain.
		 */
		if (get_nulls_value(node) != slot)
			goto restart;
		rcu_read_unlock();
299 300 301
	}
}
EXPORT_SYMBOL_GPL(inet_twsk_purge);