Orange Pi5 kernel

^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   1) // SPDX-License-Identifier: GPL-2.0-only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   3)  * INET		An implementation of the TCP/IP protocol suite for the LINUX
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   4)  *		operating system.  INET is implemented using the  BSD Socket
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   5)  *		interface as the means of communication with the user level.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   6)  *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   7)  *		Generic TIME_WAIT sockets functions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   8)  *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   9)  *		From code orinally in TCP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  10)  */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  11) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  12) #include <linux/kernel.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  13) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  14) #include <linux/module.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  15) #include <net/inet_hashtables.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  16) #include <net/inet_timewait_sock.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  17) #include <net/ip.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  18) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  19) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  20) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  21)  *	inet_twsk_bind_unhash - unhash a timewait socket from bind hash
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  22)  *	@tw: timewait socket
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  23)  *	@hashinfo: hashinfo pointer
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  24)  *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  25)  *	unhash a timewait socket from bind hash, if hashed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  26)  *	bind hash lock must be held by caller.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  27)  *	Returns 1 if caller should call inet_twsk_put() after lock release.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  28)  */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  29) void inet_twsk_bind_unhash(struct inet_timewait_sock *tw,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  30) 			  struct inet_hashinfo *hashinfo)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  31) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  32) 	struct inet_bind_bucket *tb = tw->tw_tb;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  33) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  34) 	if (!tb)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  35) 		return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  36) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  37) 	__hlist_del(&tw->tw_bind_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  38) 	tw->tw_tb = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  39) 	inet_bind_bucket_destroy(hashinfo->bind_bucket_cachep, tb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  40) 	__sock_put((struct sock *)tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  41) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  42) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  43) /* Must be called with locally disabled BHs. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  44) static void inet_twsk_kill(struct inet_timewait_sock *tw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  45) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  46) 	struct inet_hashinfo *hashinfo = tw->tw_dr->hashinfo;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  47) 	spinlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  48) 	struct inet_bind_hashbucket *bhead;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  49) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  50) 	spin_lock(lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  51) 	sk_nulls_del_node_init_rcu((struct sock *)tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  52) 	spin_unlock(lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  53) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  54) 	/* Disassociate with bind bucket. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  55) 	bhead = &hashinfo->bhash[inet_bhashfn(twsk_net(tw), tw->tw_num,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  56) 			hashinfo->bhash_size)];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  57) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  58) 	spin_lock(&bhead->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  59) 	inet_twsk_bind_unhash(tw, hashinfo);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  60) 	spin_unlock(&bhead->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  61) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  62) 	atomic_dec(&tw->tw_dr->tw_count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  63) 	inet_twsk_put(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  64) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  65) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  66) void inet_twsk_free(struct inet_timewait_sock *tw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  67) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  68) 	struct module *owner = tw->tw_prot->owner;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  69) 	twsk_destructor((struct sock *)tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  70) #ifdef SOCK_REFCNT_DEBUG
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  71) 	pr_debug("%s timewait_sock %p released\n", tw->tw_prot->name, tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  72) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  73) 	kmem_cache_free(tw->tw_prot->twsk_prot->twsk_slab, tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  74) 	module_put(owner);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  75) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  76) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  77) void inet_twsk_put(struct inet_timewait_sock *tw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  78) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  79) 	if (refcount_dec_and_test(&tw->tw_refcnt))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  80) 		inet_twsk_free(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  81) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  82) EXPORT_SYMBOL_GPL(inet_twsk_put);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  83) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  84) static void inet_twsk_add_node_rcu(struct inet_timewait_sock *tw,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  85) 				   struct hlist_nulls_head *list)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  86) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  87) 	hlist_nulls_add_head_rcu(&tw->tw_node, list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  88) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  89) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  90) static void inet_twsk_add_bind_node(struct inet_timewait_sock *tw,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  91) 				    struct hlist_head *list)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  92) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  93) 	hlist_add_head(&tw->tw_bind_node, list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  94) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  95) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  96) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  97)  * Enter the time wait state. This is called with locally disabled BH.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  98)  * Essentially we whip up a timewait bucket, copy the relevant info into it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  99)  * from the SK, and mess with hash chains and list linkage.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100)  */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) void inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) 			   struct inet_hashinfo *hashinfo)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) 	const struct inet_sock *inet = inet_sk(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) 	const struct inet_connection_sock *icsk = inet_csk(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) 	struct inet_ehash_bucket *ehead = inet_ehash_bucket(hashinfo, sk->sk_hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) 	spinlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) 	struct inet_bind_hashbucket *bhead;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) 	/* Step 1: Put TW into bind hash. Original socket stays there too.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) 	   Note, that any socket with inet->num != 0 MUST be bound in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) 	   binding cache, even if it is closed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) 	 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) 	bhead = &hashinfo->bhash[inet_bhashfn(twsk_net(tw), inet->inet_num,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) 			hashinfo->bhash_size)];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) 	spin_lock(&bhead->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) 	tw->tw_tb = icsk->icsk_bind_hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) 	WARN_ON(!icsk->icsk_bind_hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) 	inet_twsk_add_bind_node(tw, &tw->tw_tb->owners);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) 	spin_unlock(&bhead->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) 	spin_lock(lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) 	inet_twsk_add_node_rcu(tw, &ehead->chain);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) 	/* Step 3: Remove SK from hash chain */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) 	if (__sk_nulls_del_node_init_rcu(sk))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) 		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) 	spin_unlock(lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) 	/* tw_refcnt is set to 3 because we have :
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) 	 * - one reference for bhash chain.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) 	 * - one reference for ehash chain.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) 	 * - one reference for timer.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) 	 * We can use atomic_set() because prior spin_lock()/spin_unlock()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) 	 * committed into memory all tw fields.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) 	 * Also note that after this point, we lost our implicit reference
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) 	 * so we are not allowed to use tw anymore.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) 	 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) 	refcount_set(&tw->tw_refcnt, 3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) EXPORT_SYMBOL_GPL(inet_twsk_hashdance);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) static void tw_timer_handler(struct timer_list *t)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) 	struct inet_timewait_sock *tw = from_timer(tw, t, tw_timer);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) 	if (tw->tw_kill)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) 		__NET_INC_STATS(twsk_net(tw), LINUX_MIB_TIMEWAITKILLED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) 	else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) 		__NET_INC_STATS(twsk_net(tw), LINUX_MIB_TIMEWAITED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) 	inet_twsk_kill(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) struct inet_timewait_sock *inet_twsk_alloc(const struct sock *sk,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) 					   struct inet_timewait_death_row *dr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) 					   const int state)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) 	struct inet_timewait_sock *tw;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) 	if (atomic_read(&dr->tw_count) >= dr->sysctl_max_tw_buckets)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) 		return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) 	tw = kmem_cache_alloc(sk->sk_prot_creator->twsk_prot->twsk_slab,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) 			      GFP_ATOMIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) 	if (tw) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) 		const struct inet_sock *inet = inet_sk(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) 		tw->tw_dr	    = dr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) 		/* Give us an identity. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) 		tw->tw_daddr	    = inet->inet_daddr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) 		tw->tw_rcv_saddr    = inet->inet_rcv_saddr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) 		tw->tw_bound_dev_if = sk->sk_bound_dev_if;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) 		tw->tw_tos	    = inet->tos;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) 		tw->tw_num	    = inet->inet_num;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) 		tw->tw_state	    = TCP_TIME_WAIT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) 		tw->tw_substate	    = state;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) 		tw->tw_sport	    = inet->inet_sport;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) 		tw->tw_dport	    = inet->inet_dport;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) 		tw->tw_family	    = sk->sk_family;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) 		tw->tw_reuse	    = sk->sk_reuse;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) 		tw->tw_reuseport    = sk->sk_reuseport;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) 		tw->tw_hash	    = sk->sk_hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) 		tw->tw_ipv6only	    = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) 		tw->tw_transparent  = inet->transparent;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) 		tw->tw_prot	    = sk->sk_prot_creator;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) 		atomic64_set(&tw->tw_cookie, atomic64_read(&sk->sk_cookie));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) 		twsk_net_set(tw, sock_net(sk));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) 		timer_setup(&tw->tw_timer, tw_timer_handler, TIMER_PINNED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) 		/*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) 		 * Because we use RCU lookups, we should not set tw_refcnt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) 		 * to a non null value before everything is setup for this
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) 		 * timewait socket.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) 		 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) 		refcount_set(&tw->tw_refcnt, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) 		__module_get(tw->tw_prot->owner);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) 	}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) 	return tw;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) EXPORT_SYMBOL_GPL(inet_twsk_alloc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) /* These are always called from BH context.  See callers in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205)  * tcp_input.c to verify this.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206)  */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) /* This is for handling early-kills of TIME_WAIT sockets.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209)  * Warning : consume reference.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210)  * Caller should not access tw anymore.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211)  */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) void inet_twsk_deschedule_put(struct inet_timewait_sock *tw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) 	if (del_timer_sync(&tw->tw_timer))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) 		inet_twsk_kill(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) 	inet_twsk_put(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) EXPORT_SYMBOL(inet_twsk_deschedule_put);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) void __inet_twsk_schedule(struct inet_timewait_sock *tw, int timeo, bool rearm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) 	/* timeout := RTO * 3.5
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) 	 *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) 	 * 3.5 = 1+2+0.5 to wait for two retransmits.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) 	 *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) 	 * RATIONALE: if FIN arrived and we entered TIME-WAIT state,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) 	 * our ACK acking that FIN can be lost. If N subsequent retransmitted
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) 	 * FINs (or previous seqments) are lost (probability of such event
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) 	 * is p^(N+1), where p is probability to lose single packet and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) 	 * time to detect the loss is about RTO*(2^N - 1) with exponential
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) 	 * backoff). Normal timewait length is calculated so, that we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) 	 * waited at least for one retransmitted FIN (maximal RTO is 120sec).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) 	 * [ BTW Linux. following BSD, violates this requirement waiting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) 	 *   only for 60sec, we should wait at least for 240 secs.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) 	 *   Well, 240 consumes too much of resources 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) 	 * ]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) 	 * This interval is not reduced to catch old duplicate and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) 	 * responces to our wandering segments living for two MSLs.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) 	 * However, if we use PAWS to detect
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) 	 * old duplicates, we can reduce the interval to bounds required
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) 	 * by RTO, rather than MSL. So, if peer understands PAWS, we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) 	 * kill tw bucket after 3.5*RTO (it is important that this number
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) 	 * is greater than TS tick!) and detect old duplicates with help
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) 	 * of PAWS.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) 	 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) 	tw->tw_kill = timeo <= 4*HZ;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) 	if (!rearm) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) 		BUG_ON(mod_timer(&tw->tw_timer, jiffies + timeo));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) 		atomic_inc(&tw->tw_dr->tw_count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) 	} else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) 		mod_timer_pending(&tw->tw_timer, jiffies + timeo);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) 	}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) EXPORT_SYMBOL_GPL(__inet_twsk_schedule);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) void inet_twsk_purge(struct inet_hashinfo *hashinfo, int family)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) 	struct inet_timewait_sock *tw;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) 	struct sock *sk;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) 	struct hlist_nulls_node *node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) 	unsigned int slot;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) 	for (slot = 0; slot <= hashinfo->ehash_mask; slot++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) 		struct inet_ehash_bucket *head = &hashinfo->ehash[slot];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) restart_rcu:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) 		cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) 		rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) restart:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) 		sk_nulls_for_each_rcu(sk, node, &head->chain) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) 			if (sk->sk_state != TCP_TIME_WAIT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) 				continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) 			tw = inet_twsk(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) 			if ((tw->tw_family != family) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) 				refcount_read(&twsk_net(tw)->count))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) 				continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) 			if (unlikely(!refcount_inc_not_zero(&tw->tw_refcnt)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) 				continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) 			if (unlikely((tw->tw_family != family) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) 				     refcount_read(&twsk_net(tw)->count))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) 				inet_twsk_put(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) 				goto restart;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) 			}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) 			rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) 			local_bh_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) 			inet_twsk_deschedule_put(tw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) 			local_bh_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) 			goto restart_rcu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) 		}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) 		/* If the nulls value we got at the end of this lookup is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) 		 * not the expected one, we must restart lookup.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) 		 * We probably met an item that was moved to another chain.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) 		 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) 		if (get_nulls_value(node) != slot)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) 			goto restart;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) 		rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) 	}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) EXPORT_SYMBOL_GPL(inet_twsk_purge);