af_inet.c

GNU Hurd 源代码

/*
 * 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.
 *
 *		PF_INET protocol family socket handler.
 *
 * Version:	$Id: af_inet.c,v 1.87.2.5 1999/08/08 08:43:10 davem Exp $
 *
 * Authors:	Ross Biro, <[e-mail]>
 *		Fred N. van Kempen, <[e-mail]>
 *		Florian La Roche, <[e-mail]>
 *		Alan Cox, <[e-mail]>
 *
 * Changes (see also sock.c)
 *
 *		A.N.Kuznetsov	:	Socket death error in accept().
 *		John Richardson :	Fix non blocking error in connect()
 *					so sockets that fail to connect
 *					don't return -EINPROGRESS.
 *		Alan Cox	:	Asynchronous I/O support
 *		Alan Cox	:	Keep correct socket pointer on sock structures
 *					when accept() ed
 *		Alan Cox	:	Semantics of SO_LINGER aren't state moved
 *					to close when you look carefully. With
 *					this fixed and the accept bug fixed
 *					some RPC stuff seems happier.
 *		Niibe Yutaka	:	4.4BSD style write async I/O
 *		Alan Cox,
 *		Tony Gale 	:	Fixed reuse semantics.
 *		Alan Cox	:	bind() shouldn't abort existing but dead
 *					sockets. Stops FTP netin:.. I hope.
 *		Alan Cox	:	bind() works correctly for RAW sockets. Note
 *					that FreeBSD at least was broken in this respect
 *					so be careful with compatibility tests...
 *		Alan Cox	:	routing cache support
 *		Alan Cox	:	memzero the socket structure for compactness.
 *		Matt Day	:	nonblock connect error handler
 *		Alan Cox	:	Allow large numbers of pending sockets
 *					(eg for big web sites), but only if
 *					specifically application requested.
 *		Alan Cox	:	New buffering throughout IP. Used dumbly.
 *		Alan Cox	:	New buffering now used smartly.
 *		Alan Cox	:	BSD rather than common sense interpretation of
 *					listen.
 *		Germano Caronni	:	Assorted small races.
 *		Alan Cox	:	sendmsg/recvmsg basic support.
 *		Alan Cox	:	Only sendmsg/recvmsg now supported.
 *		Alan Cox	:	Locked down bind (see security list).
 *		Alan Cox	:	Loosened bind a little.
 *		Mike McLagan	:	ADD/DEL DLCI Ioctls
 *	Willy Konynenberg	:	Transparent proxying support.
 *		David S. Miller	:	New socket lookup architecture.
 *					Some other random speedups.
 *		Cyrus Durgin	:	Cleaned up file for kmod hacks.
 *		Andi Kleen	:	Fix inet_stream_connect TCP race.
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 */

#include <linux/config.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/major.h>
#include <linux/sched.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
#include <linux/init.h>
#include <linux/poll.h>

#include <asm/uaccess.h>
#include <asm/system.h>

#include <linux/inet.h>
#include <linux/netdevice.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/arp.h>
#include <net/rarp.h>
#include <net/route.h>
#include <net/tcp.h>
#include <net/udp.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/raw.h>
#include <net/icmp.h>
#include <net/ipip.h>
#include <net/inet_common.h>
#include <linux/ip_fw.h>
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
#ifdef CONFIG_IP_MASQUERADE
#include <net/ip_masq.h>
#endif
#ifdef CONFIG_BRIDGE
#include <net/br.h>
#endif
#ifdef CONFIG_KMOD
#include <linux/kmod.h>
#endif
#ifdef CONFIG_NET_RADIO
#include <linux/wireless.h>
#endif	/* CONFIG_NET_RADIO */

#define min(a,b)	((a)<(b)?(a):(b))

struct linux_mib net_statistics;

extern int raw_get_info(char *, char **, off_t, int, int);
extern int snmp_get_info(char *, char **, off_t, int, int);
extern int netstat_get_info(char *, char **, off_t, int, int);
extern int afinet_get_info(char *, char **, off_t, int, int);
extern int tcp_get_info(char *, char **, off_t, int, int);
extern int udp_get_info(char *, char **, off_t, int, int);
extern void ip_mc_drop_socket(struct sock *sk);

#ifdef CONFIG_DLCI
extern int dlci_ioctl(unsigned int, void*);
#endif

#ifdef CONFIG_DLCI_MODULE
int (*dlci_ioctl_hook)(unsigned int, void *) = NULL;
#endif

int (*rarp_ioctl_hook)(unsigned int,void*) = NULL;

/*
 *	Destroy an AF_INET socket
 */

static __inline__ void kill_sk_queues(struct sock *sk)
{
	struct sk_buff *skb;

	/* First the read buffer. */
	while((skb = skb_dequeue(&sk->receive_queue)) != NULL)
		kfree_skb(skb);

	/* Next, the error queue. */
	while((skb = skb_dequeue(&sk->error_queue)) != NULL)
		kfree_skb(skb);

  	/* Now the backlog. */
  	while((skb=skb_dequeue(&sk->back_log)) != NULL)
		kfree_skb(skb);
}

static __inline__ void kill_sk_now(struct sock *sk)
{
	/* No longer exists. */
	del_from_prot_sklist(sk);

	/* Remove from protocol hash chains. */
	sk->prot->unhash(sk);

	if(sk->opt)
		kfree(sk->opt);
	dst_release(sk->dst_cache);
	sk_free(sk);
}

static __inline__ void kill_sk_later(struct sock *sk)
{
	/* this should never happen. */
	/* actually it can if an ack has just been sent. */
	/*
	 * It's more normal than that...
	 * It can happen because a skb is still in the device queues
	 * [PR]
	 */

	NETDEBUG(printk(KERN_DEBUG "Socket destroy delayed (r=%d w=%d)\n",
			atomic_read(&sk->rmem_alloc),
			atomic_read(&sk->wmem_alloc)));

	sk->ack_backlog = 0;
	release_sock(sk);
	net_reset_timer(sk, TIME_DESTROY, SOCK_DESTROY_TIME);
}

void destroy_sock(struct sock *sk)
{
	lock_sock(sk);			/* just to be safe. */

  	/* Now we can no longer get new packets or once the
  	 * timers are killed, send them.
  	 */
  	net_delete_timer(sk);

	if (sk->prot->destroy && !sk->destroy)
		sk->prot->destroy(sk);

	sk->destroy = 1;

	kill_sk_queues(sk);

	/* Now if everything is gone we can free the socket
	 * structure, otherwise we need to keep it around until
	 * everything is gone.
	 */
	if (atomic_read(&sk->rmem_alloc) == 0 && atomic_read(&sk->wmem_alloc) == 0)
		kill_sk_now(sk);
	else
		kill_sk_later(sk);
}

/*
 *	The routines beyond this point handle the behaviour of an AF_INET
 *	socket object. Mostly it punts to the subprotocols of IP to do
 *	the work.
 */


/*
 *	Set socket options on an inet socket.
 */

int inet_setsockopt(struct socket *sock, int level, int optname,
		    char *optval, int optlen)
{
	struct sock *sk=sock->sk;
	if (sk->prot->setsockopt==NULL)
		return(-EOPNOTSUPP);
	return sk->prot->setsockopt(sk,level,optname,optval,optlen);
}

/*
 *	Get a socket option on an AF_INET socket.
 *
 *	FIX: POSIX 1003.1g is very ambiguous here. It states that
 *	asynchronous errors should be reported by getsockopt. We assume
 *	this means if you specify SO_ERROR (otherwise whats the point of it).
 */

int inet_getsockopt(struct socket *sock, int level, int optname,
		    char *optval, int *optlen)
{
	struct sock *sk=sock->sk;
	if (sk->prot->getsockopt==NULL)
		return(-EOPNOTSUPP);
	return sk->prot->getsockopt(sk,level,optname,optval,optlen);
}

/*
 *	Automatically bind an unbound socket.
 */

static int inet_autobind(struct sock *sk)
{
	/* We may need to bind the socket. */
	if (sk->num == 0) {
		if (sk->prot->get_port(sk, 0) != 0)
			return(-EAGAIN);
		sk->sport = htons(sk->num);
		sk->prot->hash(sk);
		add_to_prot_sklist(sk);
	}
	return 0;
}

/*
 *	Move a socket into listening state.
 */

int inet_listen(struct socket *sock, int backlog)
{
	struct sock *sk = sock->sk;
	unsigned char old_state;

	if (sock->state != SS_UNCONNECTED || sock->type != SOCK_STREAM)
		return(-EINVAL);

	if ((unsigned) backlog == 0)	/* BSDism */
		backlog = 1;
	if ((unsigned) backlog > SOMAXCONN)
		backlog = SOMAXCONN;
	sk->max_ack_backlog = backlog;

	/* Really, if the socket is already in listen state
	 * we can only allow the backlog to be adjusted.
	 */
	old_state = sk->state;
	if (old_state != TCP_LISTEN) {
		sk->state = TCP_LISTEN;
		sk->ack_backlog = 0;
		if (sk->num == 0) {
			if (sk->prot->get_port(sk, 0) != 0) {
				sk->state = old_state;
				return -EAGAIN;
			}
			sk->sport = htons(sk->num);
			add_to_prot_sklist(sk);
		} else {
			if (sk->prev)
				((struct tcp_bind_bucket*)sk->prev)->fastreuse = 0;
		}

		dst_release(xchg(&sk->dst_cache, NULL));
		sk->prot->hash(sk);
		sk->socket->flags |= SO_ACCEPTCON;
	}
	return 0;
}

/*
 *	Create an inet socket.
 *
 *	FIXME: Gcc would generate much better code if we set the parameters
 *	up in in-memory structure order. Gcc68K even more so
 */

static int inet_create(struct socket *sock, int protocol)
{
	struct sock *sk;
	struct proto *prot;

	/* Compatibility */
	if (sock->type == SOCK_PACKET) {
		static int warned;
		if (net_families[PF_PACKET]==NULL)
		{
#if defined(CONFIG_KMOD) && defined(CONFIG_PACKET_MODULE)
			char module_name[30];
			sprintf(module_name,"net-pf-%d", PF_PACKET);
			request_module(module_name);
			if (net_families[PF_PACKET] == NULL)
#endif
			return -ESOCKTNOSUPPORT;
		}
		if (!warned++)
			printk(KERN_INFO "%s uses obsolete (PF_INET,SOCK_PACKET)\n", current->comm);
		return net_families[PF_PACKET]->create(sock, protocol);
	}

	sock->state = SS_UNCONNECTED;
	sk = sk_alloc(PF_INET, GFP_KERNEL, 1);
	if (sk == NULL)
		goto do_oom;

	switch (sock->type) {
	case SOCK_STREAM:
		if (protocol && protocol != IPPROTO_TCP)
			goto free_and_noproto;
		protocol = IPPROTO_TCP;
		if (ipv4_config.no_pmtu_disc)
			sk->ip_pmtudisc = IP_PMTUDISC_DONT;
		else
			sk->ip_pmtudisc = IP_PMTUDISC_WANT;
		prot = &tcp_prot;
		sock->ops = &inet_stream_ops;
		break;
	case SOCK_SEQPACKET:
		goto free_and_badtype;
	case SOCK_DGRAM:
		if (protocol && protocol != IPPROTO_UDP)
			goto free_and_noproto;
		protocol = IPPROTO_UDP;
		sk->no_check = UDP_NO_CHECK;
		sk->ip_pmtudisc = IP_PMTUDISC_DONT;
		prot=&udp_prot;
		sock->ops = &inet_dgram_ops;
		break;
	case SOCK_RAW:
		if (!capable(CAP_NET_RAW))
			goto free_and_badperm;
		if (!protocol)
			goto free_and_noproto;
		prot = &raw_prot;
		sk->reuse = 1;
		sk->ip_pmtudisc = IP_PMTUDISC_DONT;
		sk->num = protocol;
		sock->ops = &inet_dgram_ops;
		if (protocol == IPPROTO_RAW)
			sk->ip_hdrincl = 1;
		break;
	default:
		goto free_and_badtype;
	}

	sock_init_data(sock,sk);

	sk->destruct = NULL;

	sk->zapped=0;
#ifdef CONFIG_TCP_NAGLE_OFF
	sk->nonagle = 1;
#endif
	sk->family = PF_INET;
	sk->protocol = protocol;

	sk->prot = prot;
	sk->backlog_rcv = prot->backlog_rcv;

	sk->timer.data = (unsigned long)sk;
	sk->timer.function = &net_timer;

	sk->ip_ttl=ip_statistics.IpDefaultTTL;

	sk->ip_mc_loop=1;
	sk->ip_mc_ttl=1;
	sk->ip_mc_index=0;
	sk->ip_mc_list=NULL;

	if (sk->num) {
		/* It assumes that any protocol which allows
		 * the user to assign a number at socket
		 * creation time automatically
		 * shares.
		 */
		sk->sport = htons(sk->num);

		/* Add to protocol hash chains. */
		sk->prot->hash(sk);
		add_to_prot_sklist(sk);
	}

	if (sk->prot->init) {
		int err = sk->prot->init(sk);
		if (err != 0) {
			destroy_sock(sk);
			return(err);
		}
	}
	return(0);

free_and_badtype:
	sk_free(sk);
	return -ESOCKTNOSUPPORT;

free_and_badperm:
	sk_free(sk);
	return -EPERM;

free_and_noproto:
	sk_free(sk);
	return -EPROTONOSUPPORT;

do_oom:
	return -ENOBUFS;
}


/*
 *	The peer socket should always be NULL (or else). When we call this
 *	function we are destroying the object and from then on nobody
 *	should refer to it.
 */

int inet_release(struct socket *sock, struct socket *peersock)
{
	struct sock *sk = sock->sk;

	if (sk) {
		long timeout;

		/* Begin closedown and wake up sleepers. */
		if (sock->state != SS_UNCONNECTED)
			sock->state = SS_DISCONNECTING;
		sk->state_change(sk);

		/* Applications forget to leave groups before exiting */
		ip_mc_drop_socket(sk);

		/* If linger is set, we don't return until the close
		 * is complete.  Otherwise we return immediately. The
		 * actually closing is done the same either way.
		 *
		 * If the close is due to the process exiting, we never
		 * linger..
		 */
		timeout = 0;
		if (sk->linger && !(current->flags & PF_EXITING)) {
			timeout = HZ * sk->lingertime;
			if (!timeout)
				timeout = MAX_SCHEDULE_TIMEOUT;
		}
		sock->sk = NULL;
		sk->socket = NULL;
		sk->prot->close(sk, timeout);
	}
	return(0);
}

static int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
{
	struct sockaddr_in *addr=(struct sockaddr_in *)uaddr;
	struct sock *sk=sock->sk;
	unsigned short snum;
	int chk_addr_ret;

	/* If the socket has its own bind function then use it. (RAW) */
	if(sk->prot->bind)
		return sk->prot->bind(sk, uaddr, addr_len);

	/* Check these errors (active socket, bad address length, double bind). */
	if ((sk->state != TCP_CLOSE)			||
	    (addr_len < sizeof(struct sockaddr_in))	||
	    (sk->num != 0))
		return -EINVAL;

	chk_addr_ret = inet_addr_type(addr->sin_addr.s_addr);
	if (addr->sin_addr.s_addr != 0 && chk_addr_ret != RTN_LOCAL &&
	    chk_addr_ret != RTN_MULTICAST && chk_addr_ret != RTN_BROADCAST) {
#ifdef CONFIG_IP_TRANSPARENT_PROXY
		/* Superuser may bind to any address to allow transparent proxying. */
		if(chk_addr_ret != RTN_UNICAST || !capable(CAP_NET_ADMIN))
#endif
			return -EADDRNOTAVAIL;	/* Source address MUST be ours! */
	}

	/*      We keep a pair of addresses. rcv_saddr is the one
	 *      used by hash lookups, and saddr is used for transmit.
	 *
	 *      In the BSD API these are the same except where it
	 *      would be illegal to use them (multicast/broadcast) in
	 *      which case the sending device address is used.
	 */
	sk->rcv_saddr = sk->saddr = addr->sin_addr.s_addr;
	if(chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST)
		sk->saddr = 0;  /* Use device */

	snum = ntohs(addr->sin_port);
#ifdef CONFIG_IP_MASQUERADE
	/* The kernel masquerader needs some ports. */
	if((snum >= PORT_MASQ_BEGIN) && (snum <= PORT_MASQ_END))
		return -EADDRINUSE;
#endif
	if (snum && snum < PROT_SOCK && !capable(CAP_NET_BIND_SERVICE))
		return(-EACCES);

	/* Make sure we are allowed to bind here. */
	if (sk->prot->get_port(sk, snum) != 0)
		return -EADDRINUSE;

	sk->sport = htons(sk->num);
	sk->daddr = 0;
	sk->dport = 0;
	sk->prot->hash(sk);
	add_to_prot_sklist(sk);
	dst_release(sk->dst_cache);
	sk->dst_cache=NULL;
	return(0);
}

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

	if (inet_autobind(sk) != 0)
		return(-EAGAIN);
	if (sk->prot->connect == NULL)
		return(-EOPNOTSUPP);
	err = sk->prot->connect(sk, (struct sockaddr *)uaddr, addr_len);
	if (err < 0)
		return(err);
	return(0);
}

static void inet_wait_for_connect(struct sock *sk)
{
	struct wait_queue wait = { current, NULL };

	add_wait_queue(sk->sleep, &wait);
	current->state = TASK_INTERRUPTIBLE;
	while (sk->state == TCP_SYN_SENT || sk->state == TCP_SYN_RECV) {
		if (signal_pending(current))
			break;
		if (sk->err)