sock.c

Linux内核源代码 为压缩文件 是<<Linux内核>>一书中的源代码

	}

	return sk;
}

void sk_free(struct sock *sk)
{
#ifdef CONFIG_FILTER
	struct sk_filter *filter;
#endif

	if (sk->destruct)
		sk->destruct(sk);

#ifdef CONFIG_FILTER
	filter = sk->filter;
	if (filter) {
		sk_filter_release(sk, filter);
		sk->filter = NULL;
	}
#endif

	if (atomic_read(&sk->omem_alloc))
		printk(KERN_DEBUG "sk_free: optmem leakage (%d bytes) detected.\n", atomic_read(&sk->omem_alloc));

	kmem_cache_free(sk_cachep, sk);
}

void __init sk_init(void)
{
	sk_cachep = kmem_cache_create("sock", sizeof(struct sock), 0,
				      SLAB_HWCACHE_ALIGN, 0, 0);
	if (!sk_cachep)
		printk(KERN_CRIT "sk_init: Cannot create sock SLAB cache!");

	if (num_physpages <= 4096) {
		sysctl_wmem_max = 32767;
		sysctl_rmem_max = 32767;
		sysctl_wmem_default = 32767;
		sysctl_wmem_default = 32767;
	} else if (num_physpages >= 131072) {
		sysctl_wmem_max = 131071;
		sysctl_rmem_max = 131071;
	}
}

/*
 *	Simple resource managers for sockets.
 */


/* 
 * Write buffer destructor automatically called from kfree_skb. 
 */
void sock_wfree(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;

	/* In case it might be waiting for more memory. */
	atomic_sub(skb->truesize, &sk->wmem_alloc);
	sk->write_space(sk);
	sock_put(sk);
}

/* 
 * Read buffer destructor automatically called from kfree_skb. 
 */
void sock_rfree(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;

	atomic_sub(skb->truesize, &sk->rmem_alloc);
}

/*
 * Allocate a skb from the socket's send buffer.
 */
struct sk_buff *sock_wmalloc(struct sock *sk, unsigned long size, int force, int priority)
{
	if (force || atomic_read(&sk->wmem_alloc) < sk->sndbuf) {
		struct sk_buff * skb = alloc_skb(size, priority);
		if (skb) {
			skb_set_owner_w(skb, sk);
			return skb;
		}
	}
	return NULL;
}

/*
 * Allocate a skb from the socket's receive buffer.
 */ 
struct sk_buff *sock_rmalloc(struct sock *sk, unsigned long size, int force, int priority)
{
	if (force || atomic_read(&sk->rmem_alloc) < sk->rcvbuf) {
		struct sk_buff *skb = alloc_skb(size, priority);
		if (skb) {
			skb_set_owner_r(skb, sk);
			return skb;
		}
	}
	return NULL;
}

/* 
 * Allocate a memory block from the socket's option memory buffer.
 */ 
void *sock_kmalloc(struct sock *sk, int size, int priority)
{
	if ((unsigned)size <= sysctl_optmem_max &&
	    atomic_read(&sk->omem_alloc)+size < sysctl_optmem_max) {
		void *mem;
		/* First do the add, to avoid the race if kmalloc
 		 * might sleep.
		 */
		atomic_add(size, &sk->omem_alloc);
		mem = kmalloc(size, priority);
		if (mem)
			return mem;
		atomic_sub(size, &sk->omem_alloc);
	}
	return NULL;
}

/*
 * Free an option memory block.
 */
void sock_kfree_s(struct sock *sk, void *mem, int size)
{
	kfree(mem);
	atomic_sub(size, &sk->omem_alloc);
}

/* It is almost wait_for_tcp_memory minus release_sock/lock_sock.
   I think, these locks should be removed for datagram sockets.
 */
static long sock_wait_for_wmem(struct sock * sk, long timeo)
{
	DECLARE_WAITQUEUE(wait, current);

	clear_bit(SOCK_ASYNC_NOSPACE, &sk->socket->flags);
	add_wait_queue(sk->sleep, &wait);
	for (;;) {
		if (signal_pending(current))
			break;
		set_bit(SOCK_NOSPACE, &sk->socket->flags);
		set_current_state(TASK_INTERRUPTIBLE);
		if (atomic_read(&sk->wmem_alloc) < sk->sndbuf)
			break;
		if (sk->shutdown & SEND_SHUTDOWN)
			break;
		if (sk->err)
			break;
		timeo = schedule_timeout(timeo);
	}
	__set_current_state(TASK_RUNNING);
	remove_wait_queue(sk->sleep, &wait);
	return timeo;
}


/*
 *	Generic send/receive buffer handlers
 */

struct sk_buff *sock_alloc_send_skb(struct sock *sk, unsigned long size, 
			unsigned long fallback, int noblock, int *errcode)
{
	int err;
	struct sk_buff *skb;
	long timeo;

	timeo = sock_sndtimeo(sk, noblock);

	while (1) {
		unsigned long try_size = size;

		err = sock_error(sk);
		if (err != 0)
			goto failure;

		/*
		 *	We should send SIGPIPE in these cases according to
		 *	1003.1g draft 6.4. If we (the user) did a shutdown()
		 *	call however we should not. 
		 *
		 *	Note: This routine isnt just used for datagrams and
		 *	anyway some datagram protocols have a notion of
		 *	close down.
		 */

		err = -EPIPE;
		if (sk->shutdown&SEND_SHUTDOWN)
			goto failure;

		if (atomic_read(&sk->wmem_alloc) < sk->sndbuf) {
			if (fallback) {
				/* The buffer get won't block, or use the atomic queue.
			 	* It does produce annoying no free page messages still.
			 	*/
				skb = alloc_skb(size, GFP_BUFFER);
				if (skb)
					break;
				try_size = fallback;
			}
			skb = alloc_skb(try_size, sk->allocation);
			if (skb)
				break;
			err = -ENOBUFS;
			goto failure;
		}

		/*
		 *	This means we have too many buffers for this socket already.
		 */

		set_bit(SOCK_ASYNC_NOSPACE, &sk->socket->flags);
		set_bit(SOCK_NOSPACE, &sk->socket->flags);
		err = -EAGAIN;
		if (!timeo)
			goto failure;
		if (signal_pending(current))
			goto interrupted;
		timeo = sock_wait_for_wmem(sk, timeo);
	}

	skb_set_owner_w(skb, sk);
	return skb;

interrupted:
	err = sock_intr_errno(timeo);
failure:
	*errcode = err;
	return NULL;
}

void __lock_sock(struct sock *sk)
{
	DECLARE_WAITQUEUE(wait, current);

	add_wait_queue_exclusive(&sk->lock.wq, &wait);
	for(;;) {
		current->state = TASK_UNINTERRUPTIBLE;
		spin_unlock_bh(&sk->lock.slock);
		schedule();
		spin_lock_bh(&sk->lock.slock);
		if(!sk->lock.users)
			break;
	}
	current->state = TASK_RUNNING;
	remove_wait_queue(&sk->lock.wq, &wait);
}

void __release_sock(struct sock *sk)
{
	struct sk_buff *skb = sk->backlog.head;

	do {
		sk->backlog.head = sk->backlog.tail = NULL;
		bh_unlock_sock(sk);

		do {
			struct sk_buff *next = skb->next;

			skb->next = NULL;
			sk->backlog_rcv(sk, skb);
			skb = next;
		} while (skb != NULL);

		bh_lock_sock(sk);
	} while((skb = sk->backlog.head) != NULL);
}

/*
 *	Generic socket manager library. Most simpler socket families
 *	use this to manage their socket lists. At some point we should
 *	hash these. By making this generic we get the lot hashed for free.
 *
 *	It is broken by design. All the protocols using it must be fixed. --ANK
 */

rwlock_t net_big_sklist_lock = RW_LOCK_UNLOCKED;
 
void sklist_remove_socket(struct sock **list, struct sock *sk)
{
	struct sock *s;

	write_lock_bh(&net_big_sklist_lock);

	while ((s = *list) != NULL) {
		if (s == sk) {
			*list = s->next;
			break;
		}
		list = &s->next;
	}

	write_unlock_bh(&net_big_sklist_lock);
	if (s)
		sock_put(s);
}

void sklist_insert_socket(struct sock **list, struct sock *sk)
{
	write_lock_bh(&net_big_sklist_lock);
	sk->next= *list;
	*list=sk;
	sock_hold(sk);
	write_unlock_bh(&net_big_sklist_lock);
}

/*
 *	This is only called from user mode. Thus it protects itself against
 *	interrupt users but doesn't worry about being called during work.
 *	Once it is removed from the queue no interrupt or bottom half will
 *	touch it and we are (fairly 8-) ) safe.
 */

void sklist_destroy_socket(struct sock **list, struct sock *sk);

/*
 *	Handler for deferred kills.
 */

static void sklist_destroy_timer(unsigned long data)
{
	struct sock *sk=(struct sock *)data;
	sklist_destroy_socket(NULL,sk);
}

/*
 *	Destroy a socket. We pass NULL for a list if we know the
 *	socket is not on a list.
 */
 
void sklist_destroy_socket(struct sock **list,struct sock *sk)
{
	struct sk_buff *skb;
	if(list)
		sklist_remove_socket(list, sk);

	while((skb=skb_dequeue(&sk->receive_queue))!=NULL)
	{
		kfree_skb(skb);
	}

	if(atomic_read(&sk->wmem_alloc) == 0 &&
	   atomic_read(&sk->rmem_alloc) == 0 &&
	   sk->dead)
	{
		sock_put(sk);
	}
	else
	{
		/*
		 *	Someone is using our buffers still.. defer
		 */
		init_timer(&sk->timer);
		sk->timer.expires=jiffies+SOCK_DESTROY_TIME;
		sk->timer.function=sklist_destroy_timer;
		sk->timer.data = (unsigned long)sk;
		add_timer(&sk->timer);
	}
}

/*
 * Set of default routines for initialising struct proto_ops when
 * the protocol does not support a particular function. In certain
 * cases where it makes no sense for a protocol to have a "do nothing"
 * function, some default processing is provided.
 */

int sock_no_release(struct socket *sock)
{
	return 0;
}

int sock_no_bind(struct socket *sock, struct sockaddr *saddr, int len)
{
	return -EOPNOTSUPP;
}

int sock_no_connect(struct socket *sock, struct sockaddr *saddr, 
		    int len, int flags)
{
	return -EOPNOTSUPP;
}

int sock_no_socketpair(struct socket *sock1, struct socket *sock2)
{
	return -EOPNOTSUPP;
}

int sock_no_accept(struct socket *sock, struct socket *newsock, int flags)
{
	return -EOPNOTSUPP;
}

int sock_no_getname(struct socket *sock, struct sockaddr *saddr, 
		    int *len, int peer)
{
	return -EOPNOTSUPP;
}

unsigned int sock_no_poll(struct file * file, struct socket *sock, poll_table *pt)
{
	return 0;
}

int sock_no_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
	return -EOPNOTSUPP;
}

int sock_no_listen(struct socket *sock, int backlog)
{
	return -EOPNOTSUPP;
}

int sock_no_shutdown(struct socket *sock, int how)
{
	return -EOPNOTSUPP;
}

int sock_no_setsockopt(struct socket *sock, int level, int optname,
		    char *optval, int optlen)
{
	return -EOPNOTSUPP;
}

int sock_no_getsockopt(struct socket *sock, int level, int optname,
		    char *optval, int *optlen)
{
	return -EOPNOTSUPP;
}

/* 
 * Note: if you add something that sleeps here then change sock_fcntl()
 *       to do proper fd locking.
 */
int sock_no_fcntl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
	struct sock *sk = sock->sk;

	switch(cmd)
	{
		case F_SETOWN:
			/*
			 * This is a little restrictive, but it's the only
			 * way to make sure that you can't send a sigurg to
			 * another process.
			 */
			if (current->pgrp != -arg &&
				current->pid != arg &&
				!capable(CAP_KILL)) return(-EPERM);
			sk->proc = arg;
			return(0);
		case F_GETOWN:
			return(sk->proc);
		default:
			return(-EINVAL);
	}
}

int sock_no_sendmsg(struct socket *sock, struct msghdr *m, int flags,
		    struct scm_cookie *scm)
{
	return -EOPNOTSUPP;
}

int sock_no_recvmsg(struct socket *sock, struct msghdr *m, int len, int flags,
		    struct scm_cookie *scm)
{
	return -EOPNOTSUPP;
}

int sock_no_mmap(struct file *file, struct socket *sock, struct vm_area_struct *vma)
{
	/* Mirror missing mmap method error code */
	return -ENODEV;
}

/*
 *	Default Socket Callbacks
 */

void sock_def_wakeup(struct sock *sk)
{
	read_lock(&sk->callback_lock);
	if (sk->sleep && waitqueue_active(sk->sleep))
		wake_up_interruptible_all(sk->sleep);
	read_unlock(&sk->callback_lock);
}

void sock_def_error_report(struct sock *sk)
{
	read_lock(&sk->callback_lock);
	if (sk->sleep && waitqueue_active(sk->sleep))
		wake_up_interruptible(sk->sleep);
	sk_wake_async(sk,0,POLL_ERR); 
	read_unlock(&sk->callback_lock);
}

void sock_def_readable(struct sock *sk, int len)
{
	read_lock(&sk->callback_lock);
	if (sk->sleep && waitqueue_active(sk->sleep))
		wake_up_interruptible(sk->sleep);
	sk_wake_async(sk,1,POLL_IN);
	read_unlock(&sk->callback_lock);
}

void sock_def_write_space(struct sock *sk)
{
	read_lock(&sk->callback_lock);

	/* Do not wake up a writer until he can make "significant"
	 * progress.  --DaveM
	 */
	if((atomic_read(&sk->wmem_alloc) << 1) <= sk->sndbuf) {
		if (sk->sleep && waitqueue_active(sk->sleep))
			wake_up_interruptible(sk->sleep);

		/* Should agree with poll, otherwise some programs break */
		if (sock_writeable(sk))
			sk_wake_async(sk, 2, POLL_OUT);
	}

	read_unlock(&sk->callback_lock);
}

void sock_def_destruct(struct sock *sk)
{
	if (sk->protinfo.destruct_hook)
		kfree(sk->protinfo.destruct_hook);
}

void sock_init_data(struct socket *sock, struct sock *sk)
{
	skb_queue_head_init(&sk->receive_queue);
	skb_queue_head_init(&sk->write_queue);
	skb_queue_head_init(&sk->error_queue);

	init_timer(&sk->timer);
	
	sk->allocation	=	GFP_KERNEL;
	sk->rcvbuf	=	sysctl_rmem_default;
	sk->sndbuf	=	sysctl_wmem_default;
	sk->state 	= 	TCP_CLOSE;
	sk->zapped	=	1;
	sk->socket	=	sock;

	if(sock)
	{
		sk->type	=	sock->type;
		sk->sleep	=	&sock->wait;
		sock->sk	=	sk;
	} else
		sk->sleep	=	NULL;

	sk->dst_lock		=	RW_LOCK_UNLOCKED;
	sk->callback_lock	=	RW_LOCK_UNLOCKED;

	sk->state_change	=	sock_def_wakeup;
	sk->data_ready		=	sock_def_readable;
	sk->write_space		=	sock_def_write_space;
	sk->error_report	=	sock_def_error_report;
	sk->destruct            =       sock_def_destruct;

	sk->peercred.pid 	=	0;
	sk->peercred.uid	=	-1;
	sk->peercred.gid	=	-1;
	sk->rcvlowat		=	1;
	sk->rcvtimeo		=	MAX_SCHEDULE_TIMEOUT;
	sk->sndtimeo		=	MAX_SCHEDULE_TIMEOUT;

	atomic_set(&sk->refcnt, 1);
}