tcp_ipv4.c

Linux内核源代码 为压缩文件 是<<Linux内核>>一书中的源代码

no_tcp_socket:
	if (len < (th->doff<<2) || tcp_checksum_complete(skb)) {
bad_packet:
		TCP_INC_STATS_BH(TcpInErrs);
	} else {
		tcp_v4_send_reset(skb);
	}

discard_it:
	/* Discard frame. */
	kfree_skb(skb);
  	return 0;

discard_and_relse:
	sock_put(sk);
	goto discard_it;

do_time_wait:
	if (len < (th->doff<<2) || tcp_checksum_complete(skb)) {
		TCP_INC_STATS_BH(TcpInErrs);
		goto discard_and_relse;
	}
	switch(tcp_timewait_state_process((struct tcp_tw_bucket *)sk,
					  skb, th, skb->len)) {
	case TCP_TW_SYN:
	{
		struct sock *sk2;

		sk2 = tcp_v4_lookup_listener(skb->nh.iph->daddr, ntohs(th->dest), tcp_v4_iif(skb));
		if (sk2 != NULL) {
			tcp_tw_deschedule((struct tcp_tw_bucket *)sk);
			tcp_timewait_kill((struct tcp_tw_bucket *)sk);
			tcp_tw_put((struct tcp_tw_bucket *)sk);
			sk = sk2;
			goto process;
		}
		/* Fall through to ACK */
	}
	case TCP_TW_ACK:
		tcp_v4_timewait_ack(sk, skb);
		break;
	case TCP_TW_RST:
		goto no_tcp_socket;
	case TCP_TW_SUCCESS:;
	}
	goto discard_it;
}

/* With per-bucket locks this operation is not-atomic, so that
 * this version is not worse.
 */
static void __tcp_v4_rehash(struct sock *sk)
{
	sk->prot->unhash(sk);
	sk->prot->hash(sk);
}

static int tcp_v4_reselect_saddr(struct sock *sk)
{
	int err;
	struct rtable *rt;
	__u32 old_saddr = sk->saddr;
	__u32 new_saddr;
	__u32 daddr = sk->daddr;

	if(sk->protinfo.af_inet.opt && sk->protinfo.af_inet.opt->srr)
		daddr = sk->protinfo.af_inet.opt->faddr;

	/* Query new route. */
	err = ip_route_connect(&rt, daddr, 0,
			       RT_TOS(sk->protinfo.af_inet.tos)|sk->localroute,
			       sk->bound_dev_if);
	if (err)
		return err;

	__sk_dst_set(sk, &rt->u.dst);
	/* sk->route_caps = rt->u.dst.dev->features; */

	new_saddr = rt->rt_src;

	if (new_saddr == old_saddr)
		return 0;

	if (sysctl_ip_dynaddr > 1) {
		printk(KERN_INFO "tcp_v4_rebuild_header(): shifting sk->saddr "
		       "from %d.%d.%d.%d to %d.%d.%d.%d\n",
		       NIPQUAD(old_saddr), 
		       NIPQUAD(new_saddr));
	}

	sk->saddr = new_saddr;
	sk->rcv_saddr = new_saddr;

	/* XXX The only one ugly spot where we need to
	 * XXX really change the sockets identity after
	 * XXX it has entered the hashes. -DaveM
	 *
	 * Besides that, it does not check for connection
	 * uniqueness. Wait for troubles.
	 */
	__tcp_v4_rehash(sk);
	return 0;
}

int tcp_v4_rebuild_header(struct sock *sk)
{
	struct rtable *rt = (struct rtable *)__sk_dst_check(sk, 0);
	u32 daddr;
	int err;

	/* Route is OK, nothing to do. */
	if (rt != NULL)
		return 0;

	/* Reroute. */
	daddr = sk->daddr;
	if(sk->protinfo.af_inet.opt && sk->protinfo.af_inet.opt->srr)
		daddr = sk->protinfo.af_inet.opt->faddr;

	err = ip_route_output(&rt, daddr, sk->saddr,
			      RT_TOS(sk->protinfo.af_inet.tos) | RTO_CONN | sk->localroute,
			      sk->bound_dev_if);
	if (!err) {
		__sk_dst_set(sk, &rt->u.dst);
		/* sk->route_caps = rt->u.dst.dev->features; */
		return 0;
	}

	/* Routing failed... */
	/* sk->route_caps = 0; */

	if (!sysctl_ip_dynaddr ||
	    sk->state != TCP_SYN_SENT ||
	    (sk->userlocks & SOCK_BINDADDR_LOCK) ||
	    (err = tcp_v4_reselect_saddr(sk)) != 0) {
		sk->err_soft=-err;
		/* sk->error_report(sk); */
	}
	return err;
}

static void v4_addr2sockaddr(struct sock *sk, struct sockaddr * uaddr)
{
	struct sockaddr_in *sin = (struct sockaddr_in *) uaddr;

	sin->sin_family		= AF_INET;
	sin->sin_addr.s_addr	= sk->daddr;
	sin->sin_port		= sk->dport;
}

/* VJ's idea. Save last timestamp seen from this destination
 * and hold it at least for normal timewait interval to use for duplicate
 * segment detection in subsequent connections, before they enter synchronized
 * state.
 */

int tcp_v4_remember_stamp(struct sock *sk)
{
	struct tcp_opt *tp = &sk->tp_pinfo.af_tcp;
	struct rtable *rt = (struct rtable*)__sk_dst_get(sk);
	struct inet_peer *peer = NULL;
	int release_it = 0;

	if (rt == NULL || rt->rt_dst != sk->daddr) {
		peer = inet_getpeer(sk->daddr, 1);
		release_it = 1;
	} else {
		if (rt->peer == NULL)
			rt_bind_peer(rt, 1);
		peer = rt->peer;
	}

	if (peer) {
		if ((s32)(peer->tcp_ts - tp->ts_recent) <= 0 ||
		    (peer->tcp_ts_stamp + TCP_PAWS_MSL < xtime.tv_sec &&
		     peer->tcp_ts_stamp <= tp->ts_recent_stamp)) {
			peer->tcp_ts_stamp = tp->ts_recent_stamp;
			peer->tcp_ts = tp->ts_recent;
		}
		if (release_it)
			inet_putpeer(peer);
		return 1;
	}

	return 0;
}

int tcp_v4_tw_remember_stamp(struct tcp_tw_bucket *tw)
{
	struct inet_peer *peer = NULL;

	peer = inet_getpeer(tw->daddr, 1);

	if (peer) {
		if ((s32)(peer->tcp_ts - tw->ts_recent) <= 0 ||
		    (peer->tcp_ts_stamp + TCP_PAWS_MSL < xtime.tv_sec &&
		     peer->tcp_ts_stamp <= tw->ts_recent_stamp)) {
			peer->tcp_ts_stamp = tw->ts_recent_stamp;
			peer->tcp_ts = tw->ts_recent;
		}
		inet_putpeer(peer);
		return 1;
	}

	return 0;
}

struct tcp_func ipv4_specific = {
	ip_queue_xmit,
	tcp_v4_send_check,
	tcp_v4_rebuild_header,
	tcp_v4_conn_request,
	tcp_v4_syn_recv_sock,
	tcp_v4_hash_connecting,
	tcp_v4_remember_stamp,
	sizeof(struct iphdr),

	ip_setsockopt,
	ip_getsockopt,
	v4_addr2sockaddr,
	sizeof(struct sockaddr_in)
};

/* NOTE: A lot of things set to zero explicitly by call to
 *       sk_alloc() so need not be done here.
 */
static int tcp_v4_init_sock(struct sock *sk)
{
	struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);

	skb_queue_head_init(&tp->out_of_order_queue);
	tcp_init_xmit_timers(sk);
	tcp_prequeue_init(tp);

	tp->rto  = TCP_TIMEOUT_INIT;
	tp->mdev = TCP_TIMEOUT_INIT;
      
	/* So many TCP implementations out there (incorrectly) count the
	 * initial SYN frame in their delayed-ACK and congestion control
	 * algorithms that we must have the following bandaid to talk
	 * efficiently to them.  -DaveM
	 */
	tp->snd_cwnd = 2;

	/* See draft-stevens-tcpca-spec-01 for discussion of the
	 * initialization of these values.
	 */
	tp->snd_ssthresh = 0x7fffffff;	/* Infinity */
	tp->snd_cwnd_clamp = ~0;
	tp->mss_cache = 536;

	tp->reordering = sysctl_tcp_reordering;

	sk->state = TCP_CLOSE;

	sk->write_space = tcp_write_space; 

	sk->tp_pinfo.af_tcp.af_specific = &ipv4_specific;

	sk->sndbuf = sysctl_tcp_wmem[1];
	sk->rcvbuf = sysctl_tcp_rmem[1];

	atomic_inc(&tcp_sockets_allocated);

	return 0;
}

static int tcp_v4_destroy_sock(struct sock *sk)
{
	struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);

	tcp_clear_xmit_timers(sk);

	/* Cleanup up the write buffer. */
  	tcp_writequeue_purge(sk);

	/* Cleans up our, hopefuly empty, out_of_order_queue. */
  	__skb_queue_purge(&tp->out_of_order_queue);

	/* Clean prequeue, it must be empty really */
	__skb_queue_purge(&tp->ucopy.prequeue);

	/* Clean up a referenced TCP bind bucket. */
	if(sk->prev != NULL)
		tcp_put_port(sk);

	atomic_dec(&tcp_sockets_allocated);

	return 0;
}

/* Proc filesystem TCP sock list dumping. */
static void get_openreq(struct sock *sk, struct open_request *req, char *tmpbuf, int i, int uid)
{
	int ttd = req->expires - jiffies;

	sprintf(tmpbuf, "%4d: %08X:%04X %08X:%04X"
		" %02X %08X:%08X %02X:%08X %08X %5d %8d %u %d %p",
		i,
		req->af.v4_req.loc_addr,
		ntohs(sk->sport),
		req->af.v4_req.rmt_addr,
		ntohs(req->rmt_port),
		TCP_SYN_RECV,
		0,0, /* could print option size, but that is af dependent. */
		1,   /* timers active (only the expire timer) */  
		ttd, 
		req->retrans,
		uid,
		0,  /* non standard timer */  
		0, /* open_requests have no inode */
		atomic_read(&sk->refcnt),
		req
		); 
}

static void get_tcp_sock(struct sock *sp, char *tmpbuf, int i)
{
	unsigned int dest, src;
	__u16 destp, srcp;
	int timer_active;
	unsigned long timer_expires;
	struct tcp_opt *tp = &sp->tp_pinfo.af_tcp;

	dest  = sp->daddr;
	src   = sp->rcv_saddr;
	destp = ntohs(sp->dport);
	srcp  = ntohs(sp->sport);
	if (tp->pending == TCP_TIME_RETRANS) {
		timer_active	= 1;
		timer_expires	= tp->timeout;
	} else if (tp->pending == TCP_TIME_PROBE0) {
		timer_active	= 4;
		timer_expires	= tp->timeout;
	} else if (timer_pending(&sp->timer)) {
		timer_active	= 2;
		timer_expires	= sp->timer.expires;
	} else {
		timer_active	= 0;
		timer_expires = jiffies;
	}

	sprintf(tmpbuf, "%4d: %08X:%04X %08X:%04X"
		" %02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %u %u %u %u %d",
		i, src, srcp, dest, destp, sp->state, 
		tp->write_seq-tp->snd_una, tp->rcv_nxt-tp->copied_seq,
		timer_active, timer_expires-jiffies,
		tp->retransmits,
		sock_i_uid(sp),
		tp->probes_out,
		sock_i_ino(sp),
		atomic_read(&sp->refcnt), sp,
		tp->rto, tp->ack.ato, (tp->ack.quick<<1)|tp->ack.pingpong,
		tp->snd_cwnd, tp->snd_ssthresh>=0xFFFF?-1:tp->snd_ssthresh
		);
}

static void get_timewait_sock(struct tcp_tw_bucket *tw, char *tmpbuf, int i)
{
	unsigned int dest, src;
	__u16 destp, srcp;
	int ttd = tw->ttd - jiffies;

	if (ttd < 0)
		ttd = 0;

	dest  = tw->daddr;
	src   = tw->rcv_saddr;
	destp = ntohs(tw->dport);
	srcp  = ntohs(tw->sport);

	sprintf(tmpbuf, "%4d: %08X:%04X %08X:%04X"
		" %02X %08X:%08X %02X:%08X %08X %5d %8d %d %d %p",
		i, src, srcp, dest, destp, tw->substate, 0, 0,
		3, ttd, 0, 0, 0, 0,
		atomic_read(&tw->refcnt), tw);
}

#define TMPSZ 150

int tcp_get_info(char *buffer, char **start, off_t offset, int length)
{
	int len = 0, num = 0, i;
	off_t begin, pos = 0;
	char tmpbuf[TMPSZ+1];

	if (offset < TMPSZ)
		len += sprintf(buffer, "%-*s\n", TMPSZ-1,
			       "  sl  local_address rem_address   st tx_queue "
			       "rx_queue tr tm->when retrnsmt   uid  timeout inode");

	pos = TMPSZ;

	/* First, walk listening socket table. */
	tcp_listen_lock();
	for(i = 0; i < TCP_LHTABLE_SIZE; i++) {
		struct sock *sk = tcp_listening_hash[i];
		struct tcp_listen_opt *lopt;
		int k;

		for (sk = tcp_listening_hash[i]; sk; sk = sk->next, num++) {
			struct open_request *req;
			int uid;
			struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);

			if (!TCP_INET_FAMILY(sk->family))
				goto skip_listen;

			pos += TMPSZ;
			if (pos >= offset) {
				get_tcp_sock(sk, tmpbuf, num);
				len += sprintf(buffer+len, "%-*s\n", TMPSZ-1, tmpbuf);
				if (len >= length) {
					tcp_listen_unlock();
					goto out_no_bh;
				}
			}

skip_listen:
			uid = sock_i_uid(sk);
			read_lock_bh(&tp->syn_wait_lock);
			lopt = tp->listen_opt;
			if (lopt && lopt->qlen != 0) {
				for (k=0; k<TCP_SYNQ_HSIZE; k++) {
					for (req = lopt->syn_table[k]; req; req = req->dl_next, num++) {
						if (!TCP_INET_FAMILY(req->class->family))
							continue;

						pos += TMPSZ;
						if (pos <= offset)
							continue;
						get_openreq(sk, req, tmpbuf, num, uid);
						len += sprintf(buffer+len, "%-*s\n", TMPSZ-1, tmpbuf);
						if(len >= length) {
							read_unlock_bh(&tp->syn_wait_lock);
							tcp_listen_unlock();
							goto out_no_bh;
						}
					}
				}
			}
			read_unlock_bh(&tp->syn_wait_lock);

			/* Completed requests are in normal socket hash table */
		}
	}
	tcp_listen_unlock();

	local_bh_disable();

	/* Next, walk established hash chain. */
	for (i = 0; i < tcp_ehash_size; i++) {
		struct tcp_ehash_bucket *head = &tcp_ehash[i];
		struct sock *sk;
		struct tcp_tw_bucket *tw;

		read_lock(&head->lock);
		for(sk = head->chain; sk; sk = sk->next, num++) {
			if (!TCP_INET_FAMILY(sk->family))
				continue;
			pos += TMPSZ;
			if (pos <= offset)
				continue;
			get_tcp_sock(sk, tmpbuf, num);
			len += sprintf(buffer+len, "%-*s\n", TMPSZ-1, tmpbuf);
			if(len >= length) {
				read_unlock(&head->lock);
				goto out;
			}
		}
		for (tw = (struct tcp_tw_bucket *)tcp_ehash[i+tcp_ehash_size].chain;
		     tw != NULL;
		     tw = (struct tcp_tw_bucket *)tw->next, num++) {
			if (!TCP_INET_FAMILY(tw->family))
				continue;
			pos += TMPSZ;
			if (pos <= offset)
				continue;
			get_timewait_sock(tw, tmpbuf, num);
			len += sprintf(buffer+len, "%-*s\n", TMPSZ-1, tmpbuf);
			if(len >= length) {
				read_unlock(&head->lock);
				goto out;
			}
		}
		read_unlock(&head->lock);
	}

out:
	local_bh_enable();
out_no_bh:

	begin = len - (pos - offset);
	*start = buffer + begin;
	len -= begin;
	if(len > length)
		len = length;
	if (len < 0)
		len = 0; 
	return len;
}

struct proto tcp_prot = {
	name:		"TCP",
	close:		tcp_close,
	connect:	tcp_v4_connect,
	disconnect:	tcp_disconnect,
	accept:		tcp_accept,
	ioctl:		tcp_ioctl,
	init:		tcp_v4_init_sock,
	destroy:	tcp_v4_destroy_sock,
	shutdown:	tcp_shutdown,
	setsockopt:	tcp_setsockopt,
	getsockopt:	tcp_getsockopt,
	sendmsg:	tcp_sendmsg,
	recvmsg:	tcp_recvmsg,
	backlog_rcv:	tcp_v4_do_rcv,
	hash:		tcp_v4_hash,
	unhash:		tcp_unhash,
	get_port:	tcp_v4_get_port,
};



void __init tcp_v4_init(struct net_proto_family *ops)
{
	int err;

	tcp_inode.i_mode = S_IFSOCK;
	tcp_inode.i_sock = 1;
	tcp_inode.i_uid = 0;
	tcp_inode.i_gid = 0;
	init_waitqueue_head(&tcp_inode.i_wait);
	init_waitqueue_head(&tcp_inode.u.socket_i.wait);

	tcp_socket->inode = &tcp_inode;
	tcp_socket->state = SS_UNCONNECTED;
	tcp_socket->type=SOCK_RAW;

	if ((err=ops->create(tcp_socket, IPPROTO_TCP))<0)
		panic("Failed to create the TCP control socket.\n");
	tcp_socket->sk->allocation=GFP_ATOMIC;
	tcp_socket->sk->protinfo.af_inet.ttl = MAXTTL;

	/* Unhash it so that IP input processing does not even
	 * see it, we do not wish this socket to see incoming
	 * packets.
	 */
	tcp_socket->sk->prot->unhash(tcp_socket->sk);
}