tcp.c

《嵌入式系统设计与实例开发实验教材二源码》Linux内核移植与编译实验

			if (answ && !skb_queue_empty(&sk->receive_queue))
				answ -= ((struct sk_buff*)sk->receive_queue.prev)->h.th->fin;
		} else
			answ = tp->urg_seq - tp->copied_seq;
		release_sock(sk);
		break;
	case SIOCATMARK:
		{
			answ = tp->urg_data && tp->urg_seq == tp->copied_seq;
			break;
		}
	case SIOCOUTQ:
		if (sk->state == TCP_LISTEN)
			return(-EINVAL);

		if ((1<<sk->state) & (TCPF_SYN_SENT|TCPF_SYN_RECV))
			answ = 0;
		else
			answ = tp->write_seq - tp->snd_una;
		break;
	default:
		return(-ENOIOCTLCMD);
	};

	return put_user(answ, (int *)arg);
}


int tcp_listen_start(struct sock *sk)
{
	struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);
	struct tcp_listen_opt *lopt;

	sk->max_ack_backlog = 0;
	sk->ack_backlog = 0;
	tp->accept_queue = tp->accept_queue_tail = NULL;
	tp->syn_wait_lock = RW_LOCK_UNLOCKED;
	tcp_delack_init(tp);

	lopt = kmalloc(sizeof(struct tcp_listen_opt), GFP_KERNEL);
	if (!lopt)
		return -ENOMEM;

	memset(lopt, 0, sizeof(struct tcp_listen_opt));
	for (lopt->max_qlen_log = 6; ; lopt->max_qlen_log++)
		if ((1<<lopt->max_qlen_log) >= sysctl_max_syn_backlog)
			break;

	write_lock_bh(&tp->syn_wait_lock);
	tp->listen_opt = lopt;
	write_unlock_bh(&tp->syn_wait_lock);

	/* There is race window here: we announce ourselves listening,
	 * but this transition is still not validated by get_port().
	 * It is OK, because this socket enters to hash table only
	 * after validation is complete.
	 */
	sk->state = TCP_LISTEN;
	if (sk->prot->get_port(sk, sk->num) == 0) {
		sk->sport = htons(sk->num);

		sk_dst_reset(sk);
		sk->prot->hash(sk);

		return 0;
	}

	sk->state = TCP_CLOSE;
	write_lock_bh(&tp->syn_wait_lock);
	tp->listen_opt = NULL;
	write_unlock_bh(&tp->syn_wait_lock);
	kfree(lopt);
	return -EADDRINUSE;
}

/*
 *	This routine closes sockets which have been at least partially
 *	opened, but not yet accepted.
 */

static void tcp_listen_stop (struct sock *sk)
{
	struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);
	struct tcp_listen_opt *lopt = tp->listen_opt;
	struct open_request *acc_req = tp->accept_queue;
	struct open_request *req;
	int i;

	tcp_delete_keepalive_timer(sk);

	/* make all the listen_opt local to us */
	write_lock_bh(&tp->syn_wait_lock);
	tp->listen_opt =NULL;
	write_unlock_bh(&tp->syn_wait_lock);
	tp->accept_queue = tp->accept_queue_tail = NULL;

	if (lopt->qlen) {
		for (i=0; i<TCP_SYNQ_HSIZE; i++) {
			while ((req = lopt->syn_table[i]) != NULL) {
				lopt->syn_table[i] = req->dl_next;
				lopt->qlen--;
				tcp_openreq_free(req);

		/* Following specs, it would be better either to send FIN
		 * (and enter FIN-WAIT-1, it is normal close)
		 * or to send active reset (abort). 
		 * Certainly, it is pretty dangerous while synflood, but it is
		 * bad justification for our negligence 8)
		 * To be honest, we are not able to make either
		 * of the variants now.			--ANK
		 */
			}
		}
	}
	BUG_TRAP(lopt->qlen == 0);

	kfree(lopt);

	while ((req=acc_req) != NULL) {
		struct sock *child = req->sk;

		acc_req = req->dl_next;

		local_bh_disable();
		bh_lock_sock(child);
		BUG_TRAP(child->lock.users==0);
		sock_hold(child);

		tcp_disconnect(child, O_NONBLOCK);

		sock_orphan(child);

		atomic_inc(&tcp_orphan_count);

		tcp_destroy_sock(child);

		bh_unlock_sock(child);
		local_bh_enable();
		sock_put(child);

		tcp_acceptq_removed(sk);
		tcp_openreq_fastfree(req);
	}
	BUG_TRAP(sk->ack_backlog == 0);
}

/*
 *	Wait for a socket to get into the connected state
 *
 *	Note: Must be called with the socket locked.
 */
static int wait_for_tcp_connect(struct sock * sk, int flags, long *timeo_p)
{
	struct task_struct *tsk = current;
	DECLARE_WAITQUEUE(wait, tsk);

	while((1 << sk->state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT)) {
		if(sk->err)
			return sock_error(sk);
		if((1 << sk->state) &
		   ~(TCPF_SYN_SENT | TCPF_SYN_RECV))
			return -EPIPE;
		if(!*timeo_p)
			return -EAGAIN;
		if(signal_pending(tsk))
			return sock_intr_errno(*timeo_p);

		__set_task_state(tsk, TASK_INTERRUPTIBLE);
		add_wait_queue(sk->sleep, &wait);
		sk->tp_pinfo.af_tcp.write_pending++;

		release_sock(sk);
		*timeo_p = schedule_timeout(*timeo_p);
		lock_sock(sk);

		__set_task_state(tsk, TASK_RUNNING);
		remove_wait_queue(sk->sleep, &wait);
		sk->tp_pinfo.af_tcp.write_pending--;
	}
	return 0;
}

static inline int tcp_memory_free(struct sock *sk)
{
	return sk->wmem_queued < sk->sndbuf;
}

/*
 *	Wait for more memory for a socket
 */
static int wait_for_tcp_memory(struct sock * sk, long *timeo)
{
	int err = 0;
	long vm_wait = 0;
	long current_timeo = *timeo;
	DECLARE_WAITQUEUE(wait, current);

	if (tcp_memory_free(sk))
		current_timeo = vm_wait = (net_random()%(HZ/5))+2;

	add_wait_queue(sk->sleep, &wait);
	for (;;) {
		set_bit(SOCK_ASYNC_NOSPACE, &sk->socket->flags);

		set_current_state(TASK_INTERRUPTIBLE);

		if (sk->err || (sk->shutdown & SEND_SHUTDOWN))
			goto do_error;
		if (!*timeo)
			goto do_nonblock;
		if (signal_pending(current))
			goto do_interrupted;
		clear_bit(SOCK_ASYNC_NOSPACE, &sk->socket->flags);
		if (tcp_memory_free(sk) && !vm_wait)
			break;

		set_bit(SOCK_NOSPACE, &sk->socket->flags);
		sk->tp_pinfo.af_tcp.write_pending++;
		release_sock(sk);
		if (!tcp_memory_free(sk) || vm_wait)
			current_timeo = schedule_timeout(current_timeo);
		lock_sock(sk);
		sk->tp_pinfo.af_tcp.write_pending--;

		if (vm_wait) {
			vm_wait -= current_timeo;
			current_timeo = *timeo;
			if (current_timeo != MAX_SCHEDULE_TIMEOUT &&
			    (current_timeo -= vm_wait) < 0)
				current_timeo = 0;
			vm_wait = 0;
		}
		*timeo = current_timeo;
	}
out:
	current->state = TASK_RUNNING;
	remove_wait_queue(sk->sleep, &wait);
	return err;

do_error:
	err = -EPIPE;
	goto out;
do_nonblock:
	err = -EAGAIN;
	goto out;
do_interrupted:
	err = sock_intr_errno(*timeo);
	goto out;
}

ssize_t do_tcp_sendpages(struct sock *sk, struct page **pages, int poffset, size_t psize, int flags);

static inline int
can_coalesce(struct sk_buff *skb, int i, struct page *page, int off)
{
	if (i) {
		skb_frag_t *frag = &skb_shinfo(skb)->frags[i-1];
		return page == frag->page &&
			off == frag->page_offset+frag->size;
	}
	return 0;
}

static inline void
fill_page_desc(struct sk_buff *skb, int i, struct page *page, int off, int size)
{
	skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
	frag->page = page;
	frag->page_offset = off;
	frag->size = size;
	skb_shinfo(skb)->nr_frags = i+1;
}

static inline void tcp_mark_push(struct tcp_opt *tp, struct sk_buff *skb)
{
	TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH;
	tp->pushed_seq = tp->write_seq;
}

static inline int forced_push(struct tcp_opt *tp)
{
	return after(tp->write_seq, tp->pushed_seq + (tp->max_window>>1));
}

static inline void
skb_entail(struct sock *sk, struct tcp_opt *tp, struct sk_buff *skb)
{
	skb->csum = 0;
	TCP_SKB_CB(skb)->seq = tp->write_seq;
	TCP_SKB_CB(skb)->end_seq = tp->write_seq;
	TCP_SKB_CB(skb)->flags = TCPCB_FLAG_ACK;
	TCP_SKB_CB(skb)->sacked = 0;
	__skb_queue_tail(&sk->write_queue, skb);
	tcp_charge_skb(sk, skb);
	if (tp->send_head == NULL)
		tp->send_head = skb;
}

static inline void
tcp_mark_urg(struct tcp_opt *tp, int flags, struct sk_buff *skb)
{
	if (flags & MSG_OOB) {
		tp->urg_mode = 1;
		tp->snd_up = tp->write_seq;
		TCP_SKB_CB(skb)->sacked |= TCPCB_URG;
	}
}

static inline void
tcp_push(struct sock *sk, struct tcp_opt *tp, int flags, int mss_now, int nonagle)
{
	if (tp->send_head) {
		struct sk_buff *skb = sk->write_queue.prev;
		if (!(flags&MSG_MORE) || forced_push(tp))
			tcp_mark_push(tp, skb);
		tcp_mark_urg(tp, flags, skb);
		__tcp_push_pending_frames(sk, tp, mss_now, (flags&MSG_MORE) ? 2 : nonagle);
	}
}

static int tcp_error(struct sock *sk, int flags, int err)
{
	if (err == -EPIPE)
		err = sock_error(sk) ? : -EPIPE;
	if (err == -EPIPE && !(flags&MSG_NOSIGNAL))
		send_sig(SIGPIPE, current, 0);
	return err;
}

ssize_t do_tcp_sendpages(struct sock *sk, struct page **pages, int poffset, size_t psize, int flags)
{
	struct tcp_opt *tp = &(sk->tp_pinfo.af_tcp);
	int mss_now;
	int err;
	ssize_t copied;
	long timeo = sock_sndtimeo(sk, flags&MSG_DONTWAIT);

	/* Wait for a connection to finish. */
	if ((1 << sk->state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT))
		if((err = wait_for_tcp_connect(sk, 0, &timeo)) != 0)
			goto out_err;

	clear_bit(SOCK_ASYNC_NOSPACE, &sk->socket->flags);

	mss_now = tcp_current_mss(sk);
	copied = 0;

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

	while (psize > 0) {
		struct sk_buff *skb = sk->write_queue.prev;
		int offset, size, copy, i;
		struct page *page;

		page = pages[poffset/PAGE_SIZE];
		offset = poffset % PAGE_SIZE;
		size = min_t(size_t, psize, PAGE_SIZE-offset);

		if (tp->send_head==NULL || (copy = mss_now - skb->len) <= 0) {
new_segment:
			if (!tcp_memory_free(sk))
				goto wait_for_sndbuf;

			skb = tcp_alloc_pskb(sk, 0, tp->mss_cache, sk->allocation);
			if (skb == NULL)
				goto wait_for_memory;

			skb_entail(sk, tp, skb);
			copy = mss_now;
		}

		if (copy > size)
			copy = size;

		i = skb_shinfo(skb)->nr_frags;
		if (can_coalesce(skb, i, page, offset)) {
			skb_shinfo(skb)->frags[i-1].size += copy;
		} else if (i < MAX_SKB_FRAGS) {
			get_page(page);
			fill_page_desc(skb, i, page, offset, copy);
		} else {
			tcp_mark_push(tp, skb);
			goto new_segment;
		}

		skb->len += copy;
		skb->data_len += copy;
		skb->ip_summed = CHECKSUM_HW;
		tp->write_seq += copy;
		TCP_SKB_CB(skb)->end_seq += copy;

		if (!copied)
			TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_PSH;

		copied += copy;
		poffset += copy;
		if (!(psize -= copy))
			goto out;

		if (skb->len != mss_now || (flags&MSG_OOB))
			continue;

		if (forced_push(tp)) {
			tcp_mark_push(tp, skb);
			__tcp_push_pending_frames(sk, tp, mss_now, 1);
		} else if (skb == tp->send_head)
			tcp_push_one(sk, mss_now);
		continue;

wait_for_sndbuf:
		set_bit(SOCK_NOSPACE, &sk->socket->flags);
wait_for_memory:
		if (copied)
			tcp_push(sk, tp, flags&~MSG_MORE, mss_now, 1);

		if ((err = wait_for_tcp_memory(sk, &timeo)) != 0)
			goto do_error;

		mss_now = tcp_current_mss(sk);
	}

out:
	if (copied)
		tcp_push(sk, tp, flags, mss_now, tp->nonagle);
	return copied;

do_error:
	if (copied)
		goto out;
out_err:
	return tcp_error(sk, flags, err);
}

ssize_t tcp_sendpage(struct socket *sock, struct page *page, int offset, size_t size, int flags)
{
	ssize_t res;
	struct sock *sk = sock->sk;

#define TCP_ZC_CSUM_FLAGS (NETIF_F_IP_CSUM|NETIF_F_NO_CSUM|NETIF_F_HW_CSUM)

	if (!(sk->route_caps & NETIF_F_SG) || 
	    !(sk->route_caps & TCP_ZC_CSUM_FLAGS))
		return sock_no_sendpage(sock, page, offset, size, flags);

#undef TCP_ZC_CSUM_FLAGS

	lock_sock(sk);
	TCP_CHECK_TIMER(sk);
	res = do_tcp_sendpages(sk, &page, offset, size, flags);
	TCP_CHECK_TIMER(sk);
	release_sock(sk);
	return res;
}

#define TCP_PAGE(sk)	(sk->tp_pinfo.af_tcp.sndmsg_page)
#define TCP_OFF(sk)	(sk->tp_pinfo.af_tcp.sndmsg_off)

static inline int
tcp_copy_to_page(struct sock *sk, char *from, struct sk_buff *skb,
		 struct page *page, int off, int copy)
{
	int err = 0;
	unsigned int csum;

	csum = csum_and_copy_from_user(from, page_address(page)+off,
				       copy, 0, &err);
	if (!err) {
		if (skb->ip_summed == CHECKSUM_NONE)
			skb->csum = csum_block_add(skb->csum, csum, skb->len);
		skb->len += copy;
		skb->data_len += copy;
		skb->truesize += copy;
		sk->wmem_queued += copy;
		sk->forward_alloc -= copy;
	}
	return err;
}

static inline int
skb_add_data(struct sk_buff *skb, char *from, int copy)
{
	int err = 0;
	unsigned int csum;
	int off = skb->len;

	csum = csum_and_copy_from_user(from, skb_put(skb, copy),
				       copy, 0, &err);
	if (!err) {
		skb->csum = csum_block_add(skb->csum, csum, off);
		return 0;
	}

	__skb_trim(skb, off);
	return -EFAULT;