af_inet.c

GNU Hurd 源代码

			break;
		schedule();
		current->state = TASK_INTERRUPTIBLE;
	}
	current->state = TASK_RUNNING;
	remove_wait_queue(sk->sleep, &wait);
}

/*
 *	Connect to a remote host. There is regrettably still a little
 *	TCP 'magic' in here.
 */

int inet_stream_connect(struct socket *sock, struct sockaddr * uaddr,
			int addr_len, int flags)
{
	struct sock *sk=sock->sk;
	int err;

	if(sock->state != SS_UNCONNECTED && sock->state != SS_CONNECTING) {
		if(sock->state == SS_CONNECTED)
			return -EISCONN;
		return -EINVAL;
	}

	if(sock->state == SS_CONNECTING) {
		/* Note: tcp_connected contains SYN_RECV, which may cause
		   bogus results here. -AK */
		if(tcp_connected(sk->state)) {
			sock->state = SS_CONNECTED;
			return 0;
		}
		if (sk->zapped || sk->err)
			goto sock_error;
		if (flags & O_NONBLOCK)
			return -EALREADY;
	} else {
		if (sk->prot->connect == NULL)
			return(-EOPNOTSUPP);

		/* We may need to bind the socket. */
		if (inet_autobind(sk) != 0)
			return(-EAGAIN);

		err = sk->prot->connect(sk, uaddr, addr_len);
		/* Note: there is a theoretical race here when an wake up
		   occurred before inet_wait_for_connect is entered. In 2.3
		   the wait queue setup should be moved before the low level
		   connect call. -AK*/
		if (err < 0)
			return(err);
  		sock->state = SS_CONNECTING;
	}

	if (sk->state > TCP_FIN_WAIT2 && sock->state == SS_CONNECTING)
		goto sock_error;

	if (sk->state != TCP_ESTABLISHED && (flags & O_NONBLOCK))
	  	return (-EINPROGRESS);

	if (sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV) {
		inet_wait_for_connect(sk);
		if (signal_pending(current))
			return -ERESTARTSYS;
	}

	sock->state = SS_CONNECTED;
	if ((sk->state != TCP_ESTABLISHED) && sk->err)
		goto sock_error;
	return(0);

sock_error:
	/* This is ugly but needed to fix a race in the ICMP error handler */
	if (sk->zapped && sk->state != TCP_CLOSE) {
		lock_sock(sk);
		tcp_set_state(sk, TCP_CLOSE);
		release_sock(sk);
		sk->zapped = 0;
	}
	sock->state = SS_UNCONNECTED;
	return sock_error(sk);
}

/*
 *	Accept a pending connection. The TCP layer now gives BSD semantics.
 */

int inet_accept(struct socket *sock, struct socket *newsock, int flags)
{
	struct sock *sk1 = sock->sk, *sk2;
	struct sock *newsk = newsock->sk;
	int err = -EINVAL;

	if (sock->state != SS_UNCONNECTED || !(sock->flags & SO_ACCEPTCON))
		goto do_err;

	err = -EOPNOTSUPP;
	if (sk1->prot->accept == NULL)
		goto do_err;

	if((sk2 = sk1->prot->accept(sk1,flags)) == NULL)
		goto do_sk1_err;

	/*
	 *	We've been passed an extra socket.
	 *	We need to free it up because the tcp module creates
	 *	its own when it accepts one.
	 */
	sk2->sleep = newsk->sleep;

	newsock->sk = sk2;
	sk2->socket = newsock;
	newsk->socket = NULL;

	if (flags & O_NONBLOCK)
		goto do_half_success;

	if(sk2->state == TCP_ESTABLISHED)
		goto do_full_success;
	if(sk2->err > 0)
		goto do_connect_err;
	err = -ECONNABORTED;
	if (sk2->state == TCP_CLOSE)
		goto do_bad_connection;
do_full_success:
	destroy_sock(newsk);
	newsock->state = SS_CONNECTED;
	return 0;

do_half_success:
	destroy_sock(newsk);
	return(0);

do_connect_err:
	err = sock_error(sk2);
do_bad_connection:
	sk2->sleep = NULL;
	sk2->socket = NULL;
	destroy_sock(sk2);
	newsock->sk = newsk;
	newsk->socket = newsock;
	return err;

do_sk1_err:
	err = sock_error(sk1);
do_err:
	return err;
}


/*
 *	This does both peername and sockname.
 */

static int inet_getname(struct socket *sock, struct sockaddr *uaddr,
		 int *uaddr_len, int peer)
{
	struct sock *sk		= sock->sk;
	struct sockaddr_in *sin	= (struct sockaddr_in *)uaddr;

	sin->sin_family = AF_INET;
	if (peer) {
		if (!tcp_connected(sk->state))
			return(-ENOTCONN);
		sin->sin_port = sk->dport;
		sin->sin_addr.s_addr = sk->daddr;
	} else {
		__u32 addr = sk->rcv_saddr;
		if (!addr)
			addr = sk->saddr;
		sin->sin_port = sk->sport;
		sin->sin_addr.s_addr = addr;
	}
	*uaddr_len = sizeof(*sin);
	return(0);
}



int inet_recvmsg(struct socket *sock, struct msghdr *msg, int size,
		 int flags, struct scm_cookie *scm)
{
	struct sock *sk = sock->sk;
	int addr_len = 0;
	int err;

	if (sock->flags & SO_ACCEPTCON)
		return(-EINVAL);
	if (sk->prot->recvmsg == NULL)
		return(-EOPNOTSUPP);
	/* We may need to bind the socket. */
	if (inet_autobind(sk) != 0)
		return(-EAGAIN);
	err = sk->prot->recvmsg(sk, msg, size, flags&MSG_DONTWAIT,
				flags&~MSG_DONTWAIT, &addr_len);
	if (err >= 0)
		msg->msg_namelen = addr_len;
	return err;
}


int inet_sendmsg(struct socket *sock, struct msghdr *msg, int size,
		 struct scm_cookie *scm)
{
	struct sock *sk = sock->sk;

	if (sk->shutdown & SEND_SHUTDOWN) {
		if (!(msg->msg_flags&MSG_NOSIGNAL))
			send_sig(SIGPIPE, current, 1);
		return(-EPIPE);
	}
	if (sk->prot->sendmsg == NULL)
		return(-EOPNOTSUPP);
	if(sk->err)
		return sock_error(sk);

	/* We may need to bind the socket. */
	if (inet_autobind(sk) != 0)
		return -EAGAIN;

	return sk->prot->sendmsg(sk, msg, size);
}


int inet_shutdown(struct socket *sock, int how)
{
	struct sock *sk = sock->sk;

	/* This should really check to make sure
	 * the socket is a TCP socket. (WHY AC...)
	 */
	how++; /* maps 0->1 has the advantage of making bit 1 rcvs and
		       1->2 bit 2 snds.
		       2->3 */
	if ((how & ~SHUTDOWN_MASK) || how==0)	/* MAXINT->0 */
		return(-EINVAL);
	if (!sk)
		return(-ENOTCONN);
	if (sock->state == SS_CONNECTING && sk->state == TCP_ESTABLISHED)
		sock->state = SS_CONNECTED;
	if (!tcp_connected(sk->state))
		return(-ENOTCONN);
	sk->shutdown |= how;
	if (sk->prot->shutdown)
		sk->prot->shutdown(sk, how);
	/* Wake up anyone sleeping in poll. */
	sk->state_change(sk);
	return(0);
}


unsigned int inet_poll(struct file * file, struct socket *sock, poll_table *wait)
{
	struct sock *sk = sock->sk;

	if (sk->prot->poll == NULL)
		return(0);
	return sk->prot->poll(file, sock, wait);
}

#ifdef _HURD_
#define inet_ioctl 0
#else

/*
 *	ioctl() calls you can issue on an INET socket. Most of these are
 *	device configuration and stuff and very rarely used. Some ioctls
 *	pass on to the socket itself.
 *
 *	NOTE: I like the idea of a module for the config stuff. ie ifconfig
 *	loads the devconfigure module does its configuring and unloads it.
 *	There's a good 20K of config code hanging around the kernel.
 */

static int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
	struct sock *sk = sock->sk;
	int err;
	int pid;

	switch(cmd)
	{
		case FIOSETOWN:
		case SIOCSPGRP:
			err = get_user(pid, (int *) arg);
			if (err)
				return err;
			if (current->pid != pid && current->pgrp != -pid &&
			    !capable(CAP_NET_ADMIN))
				return -EPERM;
			sk->proc = pid;
			return(0);
		case FIOGETOWN:
		case SIOCGPGRP:
			return put_user(sk->proc, (int *)arg);
		case SIOCGSTAMP:
			if(sk->stamp.tv_sec==0)
				return -ENOENT;
			err = copy_to_user((void *)arg,&sk->stamp,sizeof(struct timeval));
			if (err)
				err = -EFAULT;
			return err;
		case SIOCADDRT:
		case SIOCDELRT:
		case SIOCRTMSG:
			return(ip_rt_ioctl(cmd,(void *) arg));
		case SIOCDARP:
		case SIOCGARP:
		case SIOCSARP:
			return(arp_ioctl(cmd,(void *) arg));
		case SIOCDRARP:
		case SIOCGRARP:
		case SIOCSRARP:
#ifdef CONFIG_KMOD
			if (rarp_ioctl_hook == NULL)
				request_module("rarp");
#endif
			if (rarp_ioctl_hook != NULL)
				return(rarp_ioctl_hook(cmd,(void *) arg));
		case SIOCGIFADDR:
		case SIOCSIFADDR:
		case SIOCGIFBRDADDR:
		case SIOCSIFBRDADDR:
		case SIOCGIFNETMASK:
		case SIOCSIFNETMASK:
		case SIOCGIFDSTADDR:
		case SIOCSIFDSTADDR:
		case SIOCSIFPFLAGS:
		case SIOCGIFPFLAGS:
		case SIOCSIFFLAGS:
			return(devinet_ioctl(cmd,(void *) arg));
		case SIOCGIFBR:
		case SIOCSIFBR:
#ifdef CONFIG_BRIDGE
			return(br_ioctl(cmd,(void *) arg));
#else
			return -ENOPKG;
#endif

		case SIOCADDDLCI:
		case SIOCDELDLCI:
#ifdef CONFIG_DLCI
			return(dlci_ioctl(cmd, (void *) arg));
#endif

#ifdef CONFIG_DLCI_MODULE

#ifdef CONFIG_KMOD
			if (dlci_ioctl_hook == NULL)
				request_module("dlci");
#endif

			if (dlci_ioctl_hook)
				return((*dlci_ioctl_hook)(cmd, (void *) arg));
#endif
			return -ENOPKG;

		default:
			if ((cmd >= SIOCDEVPRIVATE) &&
			    (cmd <= (SIOCDEVPRIVATE + 15)))
				return(dev_ioctl(cmd,(void *) arg));

#ifdef CONFIG_NET_RADIO
			if((cmd >= SIOCIWFIRST) && (cmd <= SIOCIWLAST))
				return(dev_ioctl(cmd,(void *) arg));
#endif

			if (sk->prot->ioctl==NULL || (err=sk->prot->ioctl(sk, cmd, arg))==-ENOIOCTLCMD)
				return(dev_ioctl(cmd,(void *) arg));
			return err;
	}
	/*NOTREACHED*/
	return(0);
}

#endif

struct proto_ops inet_stream_ops = {
	PF_INET,

	sock_no_dup,
	inet_release,
	inet_bind,
	inet_stream_connect,
	sock_no_socketpair,
	inet_accept,
	inet_getname,
	inet_poll,
	inet_ioctl,
	inet_listen,
	inet_shutdown,
	inet_setsockopt,
	inet_getsockopt,
	sock_no_fcntl,
	inet_sendmsg,
	inet_recvmsg
};

struct proto_ops inet_dgram_ops = {
	PF_INET,

	sock_no_dup,
	inet_release,
	inet_bind,
	inet_dgram_connect,
	sock_no_socketpair,
	sock_no_accept,
	inet_getname,
	datagram_poll,
	inet_ioctl,
	sock_no_listen,
	inet_shutdown,
	inet_setsockopt,
	inet_getsockopt,
	sock_no_fcntl,
	inet_sendmsg,
	inet_recvmsg
};

struct net_proto_family inet_family_ops = {
	PF_INET,
	inet_create
};


#ifdef CONFIG_PROC_FS
#ifdef CONFIG_INET_RARP
static struct proc_dir_entry proc_net_rarp = {
	PROC_NET_RARP, 4, "rarp",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	rarp_get_info
};
#endif		/* RARP */
static struct proc_dir_entry proc_net_raw = {
	PROC_NET_RAW, 3, "raw",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	raw_get_info
};
static struct proc_dir_entry proc_net_netstat = {
	PROC_NET_NETSTAT, 7, "netstat",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	netstat_get_info
};
static struct proc_dir_entry proc_net_snmp = {
	PROC_NET_SNMP, 4, "snmp",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	snmp_get_info
};
static struct proc_dir_entry proc_net_sockstat = {
	PROC_NET_SOCKSTAT, 8, "sockstat",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	afinet_get_info
};
static struct proc_dir_entry proc_net_tcp = {
	PROC_NET_TCP, 3, "tcp",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	tcp_get_info
};
static struct proc_dir_entry proc_net_udp = {
	PROC_NET_UDP, 3, "udp",
	S_IFREG | S_IRUGO, 1, 0, 0,
	0, &proc_net_inode_operations,
	udp_get_info
};
#endif		/* CONFIG_PROC_FS */

extern void tcp_init(void);
extern void tcp_v4_init(struct net_proto_family *);


/*
 *	Called by socket.c on kernel startup.
 */

__initfunc(void inet_proto_init(struct net_proto *pro))
{
	struct sk_buff *dummy_skb;
	struct inet_protocol *p;

	printk(KERN_INFO "NET4: Linux TCP/IP 1.0 for NET4.0\n");

	if (sizeof(struct inet_skb_parm) > sizeof(dummy_skb->cb))
	{
		printk(KERN_CRIT "inet_proto_init: panic\n");
		return;
	}

	/*
	 *	Tell SOCKET that we are alive...
	 */

  	(void) sock_register(&inet_family_ops);

	/*
	 *	Add all the protocols.
	 */

	printk(KERN_INFO "IP Protocols: ");
	for(p = inet_protocol_base; p != NULL;)
	{
		struct inet_protocol *tmp = (struct inet_protocol *) p->next;
		inet_add_protocol(p);
		printk("%s%s",p->name,tmp?", ":"\n");
		p = tmp;
	}

	/*
	 *	Set the ARP module up
	 */

	arp_init();

  	/*
  	 *	Set the IP module up
  	 */

	ip_init();

	tcp_v4_init(&inet_family_ops);

	/* Setup TCP slab cache for open requests. */
	tcp_init();


	/*
	 *	Set the ICMP layer up
	 */

	icmp_init(&inet_family_ops);

	/* I wish inet_add_protocol had no constructor hook...
	   I had to move IPIP from net/ipv4/protocol.c :-( --ANK
	 */
#ifdef CONFIG_NET_IPIP
	ipip_init();
#endif
#ifdef CONFIG_NET_IPGRE
	ipgre_init();
#endif

	/*
	 *	Set the firewalling up
	 */
#if defined(CONFIG_IP_FIREWALL)
	ip_fw_init();
#endif

#ifdef CONFIG_IP_MASQUERADE
	ip_masq_init();
#endif

	/*
	 *	Initialise the multicast router
	 */
#if defined(CONFIG_IP_MROUTE)
	ip_mr_init();
#endif

#ifdef CONFIG_INET_RARP
	rarp_ioctl_hook = rarp_ioctl;
#endif
	/*
	 *	Create all the /proc entries.
	 */

#ifdef CONFIG_PROC_FS
#ifdef CONFIG_INET_RARP
	proc_net_register(&proc_net_rarp);
#endif		/* RARP */
	proc_net_register(&proc_net_raw);
	proc_net_register(&proc_net_snmp);
	proc_net_register(&proc_net_netstat);
	proc_net_register(&proc_net_sockstat);
	proc_net_register(&proc_net_tcp);
	proc_net_register(&proc_net_udp);
#endif		/* CONFIG_PROC_FS */
}