socket.c

linux 内核源代码

/*
 * net/tipc/socket.c: TIPC socket API
 *
 * Copyright (c) 2001-2007, Ericsson AB
 * Copyright (c) 2004-2007, Wind River Systems
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the names of the copyright holders nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
 *
 * Alternatively, this software may be distributed under the terms of the
 * GNU General Public License ("GPL") version 2 as published by the Free
 * Software Foundation.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/net.h>
#include <linux/socket.h>
#include <linux/errno.h>
#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/poll.h>
#include <linux/fcntl.h>
#include <asm/semaphore.h>
#include <asm/string.h>
#include <asm/atomic.h>
#include <net/sock.h>

#include <linux/tipc.h>
#include <linux/tipc_config.h>
#include <net/tipc/tipc_msg.h>
#include <net/tipc/tipc_port.h>

#include "core.h"

#define SS_LISTENING	-1	/* socket is listening */
#define SS_READY	-2	/* socket is connectionless */

#define OVERLOAD_LIMIT_BASE    5000

struct tipc_sock {
	struct sock sk;
	struct tipc_port *p;
	struct semaphore sem;
};

#define tipc_sk(sk) ((struct tipc_sock*)sk)

static u32 dispatch(struct tipc_port *tport, struct sk_buff *buf);
static void wakeupdispatch(struct tipc_port *tport);

static struct proto_ops packet_ops;
static struct proto_ops stream_ops;
static struct proto_ops msg_ops;

static struct proto tipc_proto;

static int sockets_enabled = 0;

static atomic_t tipc_queue_size = ATOMIC_INIT(0);


/*
 * sock_lock(): Lock a port/socket pair. lock_sock() can
 * not be used here, since the same lock must protect ports
 * with non-socket interfaces.
 * See net.c for description of locking policy.
 */
static void sock_lock(struct tipc_sock* tsock)
{
	spin_lock_bh(tsock->p->lock);
}

/*
 * sock_unlock(): Unlock a port/socket pair
 */
static void sock_unlock(struct tipc_sock* tsock)
{
	spin_unlock_bh(tsock->p->lock);
}

/**
 * pollmask - determine the current set of poll() events for a socket
 * @sock: socket structure
 *
 * TIPC sets the returned events as follows:
 * a) POLLRDNORM and POLLIN are set if the socket's receive queue is non-empty
 *    or if a connection-oriented socket is does not have an active connection
 *    (i.e. a read operation will not block).
 * b) POLLOUT is set except when a socket's connection has been terminated
 *    (i.e. a write operation will not block).
 * c) POLLHUP is set when a socket's connection has been terminated.
 *
 * IMPORTANT: The fact that a read or write operation will not block does NOT
 * imply that the operation will succeed!
 *
 * Returns pollmask value
 */

static u32 pollmask(struct socket *sock)
{
	u32 mask;

	if ((skb_queue_len(&sock->sk->sk_receive_queue) != 0) ||
	    (sock->state == SS_UNCONNECTED) ||
	    (sock->state == SS_DISCONNECTING))
		mask = (POLLRDNORM | POLLIN);
	else
		mask = 0;

	if (sock->state == SS_DISCONNECTING)
		mask |= POLLHUP;
	else
		mask |= POLLOUT;

	return mask;
}


/**
 * advance_queue - discard first buffer in queue
 * @tsock: TIPC socket
 */

static void advance_queue(struct tipc_sock *tsock)
{
	sock_lock(tsock);
	buf_discard(skb_dequeue(&tsock->sk.sk_receive_queue));
	sock_unlock(tsock);
	atomic_dec(&tipc_queue_size);
}

/**
 * tipc_create - create a TIPC socket
 * @sock: pre-allocated socket structure
 * @protocol: protocol indicator (must be 0)
 *
 * This routine creates and attaches a 'struct sock' to the 'struct socket',
 * then create and attaches a TIPC port to the 'struct sock' part.
 *
 * Returns 0 on success, errno otherwise
 */
static int tipc_create(struct net *net, struct socket *sock, int protocol)
{
	struct tipc_sock *tsock;
	struct tipc_port *port;
	struct sock *sk;
	u32 ref;

	if (net != &init_net)
		return -EAFNOSUPPORT;

	if (unlikely(protocol != 0))
		return -EPROTONOSUPPORT;

	ref = tipc_createport_raw(NULL, &dispatch, &wakeupdispatch, TIPC_LOW_IMPORTANCE);
	if (unlikely(!ref))
		return -ENOMEM;

	sock->state = SS_UNCONNECTED;

	switch (sock->type) {
	case SOCK_STREAM:
		sock->ops = &stream_ops;
		break;
	case SOCK_SEQPACKET:
		sock->ops = &packet_ops;
		break;
	case SOCK_DGRAM:
		tipc_set_portunreliable(ref, 1);
		/* fall through */
	case SOCK_RDM:
		tipc_set_portunreturnable(ref, 1);
		sock->ops = &msg_ops;
		sock->state = SS_READY;
		break;
	default:
		tipc_deleteport(ref);
		return -EPROTOTYPE;
	}

	sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto);
	if (!sk) {
		tipc_deleteport(ref);
		return -ENOMEM;
	}

	sock_init_data(sock, sk);
	init_waitqueue_head(sk->sk_sleep);
	sk->sk_rcvtimeo = 8 * HZ;   /* default connect timeout = 8s */

	tsock = tipc_sk(sk);
	port = tipc_get_port(ref);

	tsock->p = port;
	port->usr_handle = tsock;

	init_MUTEX(&tsock->sem);

	dbg("sock_create: %x\n",tsock);

	atomic_inc(&tipc_user_count);

	return 0;
}

/**
 * release - destroy a TIPC socket
 * @sock: socket to destroy
 *
 * This routine cleans up any messages that are still queued on the socket.
 * For DGRAM and RDM socket types, all queued messages are rejected.
 * For SEQPACKET and STREAM socket types, the first message is rejected
 * and any others are discarded.  (If the first message on a STREAM socket
 * is partially-read, it is discarded and the next one is rejected instead.)
 *
 * NOTE: Rejected messages are not necessarily returned to the sender!  They
 * are returned or discarded according to the "destination droppable" setting
 * specified for the message by the sender.
 *
 * Returns 0 on success, errno otherwise
 */

static int release(struct socket *sock)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sock *sk = sock->sk;
	int res = TIPC_OK;
	struct sk_buff *buf;

	dbg("sock_delete: %x\n",tsock);
	if (!tsock)
		return 0;
	down(&tsock->sem);
	if (!sock->sk) {
		up(&tsock->sem);
		return 0;
	}

	/* Reject unreceived messages, unless no longer connected */

	while (sock->state != SS_DISCONNECTING) {
		sock_lock(tsock);
		buf = skb_dequeue(&sk->sk_receive_queue);
		if (!buf)
			tsock->p->usr_handle = NULL;
		sock_unlock(tsock);
		if (!buf)
			break;
		if (TIPC_SKB_CB(buf)->handle != msg_data(buf_msg(buf)))
			buf_discard(buf);
		else
			tipc_reject_msg(buf, TIPC_ERR_NO_PORT);
		atomic_dec(&tipc_queue_size);
	}

	/* Delete TIPC port */

	res = tipc_deleteport(tsock->p->ref);
	sock->sk = NULL;

	/* Discard any remaining messages */

	while ((buf = skb_dequeue(&sk->sk_receive_queue))) {
		buf_discard(buf);
		atomic_dec(&tipc_queue_size);
	}

	up(&tsock->sem);

	sock_put(sk);

	atomic_dec(&tipc_user_count);
	return res;
}

/**
 * bind - associate or disassocate TIPC name(s) with a socket
 * @sock: socket structure
 * @uaddr: socket address describing name(s) and desired operation
 * @uaddr_len: size of socket address data structure
 *
 * Name and name sequence binding is indicated using a positive scope value;
 * a negative scope value unbinds the specified name.  Specifying no name
 * (i.e. a socket address length of 0) unbinds all names from the socket.
 *
 * Returns 0 on success, errno otherwise
 */

static int bind(struct socket *sock, struct sockaddr *uaddr, int uaddr_len)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
	int res;

	if (down_interruptible(&tsock->sem))
		return -ERESTARTSYS;

	if (unlikely(!uaddr_len)) {
		res = tipc_withdraw(tsock->p->ref, 0, NULL);
		goto exit;
	}

	if (uaddr_len < sizeof(struct sockaddr_tipc)) {
		res = -EINVAL;
		goto exit;
	}

	if (addr->family != AF_TIPC) {
		res = -EAFNOSUPPORT;
		goto exit;
	}
	if (addr->addrtype == TIPC_ADDR_NAME)
		addr->addr.nameseq.upper = addr->addr.nameseq.lower;
	else if (addr->addrtype != TIPC_ADDR_NAMESEQ) {
		res = -EAFNOSUPPORT;
		goto exit;
	}

	if (addr->scope > 0)
		res = tipc_publish(tsock->p->ref, addr->scope,
				   &addr->addr.nameseq);
	else
		res = tipc_withdraw(tsock->p->ref, -addr->scope,
				    &addr->addr.nameseq);
exit:
	up(&tsock->sem);
	return res;
}

/**
 * get_name - get port ID of socket or peer socket
 * @sock: socket structure
 * @uaddr: area for returned socket address
 * @uaddr_len: area for returned length of socket address
 * @peer: 0 to obtain socket name, 1 to obtain peer socket name
 *
 * Returns 0 on success, errno otherwise
 */

static int get_name(struct socket *sock, struct sockaddr *uaddr,
		    int *uaddr_len, int peer)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
	u32 res;

	if (down_interruptible(&tsock->sem))
		return -ERESTARTSYS;

	*uaddr_len = sizeof(*addr);
	addr->addrtype = TIPC_ADDR_ID;
	addr->family = AF_TIPC;
	addr->scope = 0;
	if (peer)
		res = tipc_peer(tsock->p->ref, &addr->addr.id);
	else
		res = tipc_ownidentity(tsock->p->ref, &addr->addr.id);
	addr->addr.name.domain = 0;

	up(&tsock->sem);
	return res;
}

/**
 * poll - read and possibly block on pollmask
 * @file: file structure associated with the socket
 * @sock: socket for which to calculate the poll bits
 * @wait: ???
 *
 * Returns the pollmask
 */

static unsigned int poll(struct file *file, struct socket *sock,
			 poll_table *wait)
{
	poll_wait(file, sock->sk->sk_sleep, wait);
	/* NEED LOCK HERE? */
	return pollmask(sock);
}

/**
 * dest_name_check - verify user is permitted to send to specified port name
 * @dest: destination address
 * @m: descriptor for message to be sent
 *
 * Prevents restricted configuration commands from being issued by
 * unauthorized users.
 *
 * Returns 0 if permission is granted, otherwise errno
 */

static int dest_name_check(struct sockaddr_tipc *dest, struct msghdr *m)
{
	struct tipc_cfg_msg_hdr hdr;

	if (likely(dest->addr.name.name.type >= TIPC_RESERVED_TYPES))
		return 0;
	if (likely(dest->addr.name.name.type == TIPC_TOP_SRV))
		return 0;

	if (likely(dest->addr.name.name.type != TIPC_CFG_SRV))
		return -EACCES;

	if (copy_from_user(&hdr, m->msg_iov[0].iov_base, sizeof(hdr)))
		return -EFAULT;
	if ((ntohs(hdr.tcm_type) & 0xC000) && (!capable(CAP_NET_ADMIN)))
		return -EACCES;

	return 0;
}

/**
 * send_msg - send message in connectionless manner
 * @iocb: (unused)
 * @sock: socket structure
 * @m: message to send
 * @total_len: length of message
 *
 * Message must have an destination specified explicitly.
 * Used for SOCK_RDM and SOCK_DGRAM messages,
 * and for 'SYN' messages on SOCK_SEQPACKET and SOCK_STREAM connections.
 * (Note: 'SYN+' is prohibited on SOCK_STREAM.)
 *
 * Returns the number of bytes sent on success, or errno otherwise
 */

static int send_msg(struct kiocb *iocb, struct socket *sock,
		    struct msghdr *m, size_t total_len)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name;
	struct sk_buff *buf;
	int needs_conn;
	int res = -EINVAL;

	if (unlikely(!dest))
		return -EDESTADDRREQ;
	if (unlikely((m->msg_namelen < sizeof(*dest)) ||
		     (dest->family != AF_TIPC)))
		return -EINVAL;

	needs_conn = (sock->state != SS_READY);
	if (unlikely(needs_conn)) {
		if (sock->state == SS_LISTENING)
			return -EPIPE;
		if (sock->state != SS_UNCONNECTED)
			return -EISCONN;
		if ((tsock->p->published) ||
		    ((sock->type == SOCK_STREAM) && (total_len != 0)))
			return -EOPNOTSUPP;
		if (dest->addrtype == TIPC_ADDR_NAME) {
			tsock->p->conn_type = dest->addr.name.name.type;
			tsock->p->conn_instance = dest->addr.name.name.instance;
		}
	}

	if (down_interruptible(&tsock->sem))
		return -ERESTARTSYS;

	if (needs_conn) {

		/* Abort any pending connection attempts (very unlikely) */

		while ((buf = skb_dequeue(&sock->sk->sk_receive_queue))) {
			tipc_reject_msg(buf, TIPC_ERR_NO_PORT);
			atomic_dec(&tipc_queue_size);
		}

		sock->state = SS_CONNECTING;
	}

	do {
		if (dest->addrtype == TIPC_ADDR_NAME) {
			if ((res = dest_name_check(dest, m)))
				goto exit;
			res = tipc_send2name(tsock->p->ref,
					     &dest->addr.name.name,
					     dest->addr.name.domain,
					     m->msg_iovlen,
					     m->msg_iov);
		}
		else if (dest->addrtype == TIPC_ADDR_ID) {
			res = tipc_send2port(tsock->p->ref,
					     &dest->addr.id,
					     m->msg_iovlen,
					     m->msg_iov);
		}
		else if (dest->addrtype == TIPC_ADDR_MCAST) {
			if (needs_conn) {
				res = -EOPNOTSUPP;
				goto exit;
			}
			if ((res = dest_name_check(dest, m)))
				goto exit;
			res = tipc_multicast(tsock->p->ref,
					     &dest->addr.nameseq,
					     0,
					     m->msg_iovlen,
					     m->msg_iov);
		}
		if (likely(res != -ELINKCONG)) {
exit:
			up(&tsock->sem);
			return res;
		}
		if (m->msg_flags & MSG_DONTWAIT) {
			res = -EWOULDBLOCK;
			goto exit;
		}
		if (wait_event_interruptible(*sock->sk->sk_sleep,
					     !tsock->p->congested)) {
		    res = -ERESTARTSYS;
		    goto exit;
		}
	} while (1);
}

/**
 * send_packet - send a connection-oriented message
 * @iocb: (unused)
 * @sock: socket structure
 * @m: message to send
 * @total_len: length of message
 *
 * Used for SOCK_SEQPACKET messages and SOCK_STREAM data.
 *
 * Returns the number of bytes sent on success, or errno otherwise
 */

static int send_packet(struct kiocb *iocb, struct socket *sock,
		       struct msghdr *m, size_t total_len)
{
	struct tipc_sock *tsock = tipc_sk(sock->sk);
	struct sockaddr_tipc *dest = (struct sockaddr_tipc *)m->msg_name;
	int res;

	/* Handle implied connection establishment */

	if (unlikely(dest))
		return send_msg(iocb, sock, m, total_len);

	if (down_interruptible(&tsock->sem)) {
		return -ERESTARTSYS;
	}

	do {
		if (unlikely(sock->state != SS_CONNECTED)) {
			if (sock->state == SS_DISCONNECTING)
				res = -EPIPE;
			else
				res = -ENOTCONN;
			goto exit;
		}

		res = tipc_send(tsock->p->ref, m->msg_iovlen, m->msg_iov);