udp.c

GNU Hurd 源代码

	/*
	 *	Get and verify the address.
	 */

	if (msg->msg_name) {
		struct sockaddr_in * usin = (struct sockaddr_in*)msg->msg_name;
		if (msg->msg_namelen < sizeof(*usin))
			return(-EINVAL);
		if (usin->sin_family != AF_INET) {
			static int complained;
			if (!complained++)
				printk(KERN_WARNING "%s forgot to set AF_INET in udp sendmsg. Fix it!\n", current->comm);
			if (usin->sin_family)
				return -EINVAL;
		}
		ufh.daddr = usin->sin_addr.s_addr;
		ufh.uh.dest = usin->sin_port;
		if (ufh.uh.dest == 0)
			return -EINVAL;
	} else {
		if (sk->state != TCP_ESTABLISHED)
			return -ENOTCONN;
		ufh.daddr = sk->daddr;
		ufh.uh.dest = sk->dport;
		/* Open fast path for connected socket.
		   Route will not be used, if at least one option is set.
		 */
		connected = 1;
  	}
#ifdef CONFIG_IP_TRANSPARENT_PROXY
	if (msg->msg_flags&MSG_PROXY) {
		/*
		 * We map the first 8 bytes of a second sockaddr_in
		 * into the last 8 (unused) bytes of a sockaddr_in.
		 */
		struct sockaddr_in *from = (struct sockaddr_in *)msg->msg_name;
		from = (struct sockaddr_in *)&from->sin_zero;
		if (from->sin_family != AF_INET)
			return -EINVAL;
		ipc.addr = from->sin_addr.s_addr;
		ufh.uh.source = from->sin_port;
		if (ipc.addr == 0)
			ipc.addr = sk->saddr;
		connected = 0;
	} else
#endif
	{
		ipc.addr = sk->saddr;
		ufh.uh.source = sk->sport;
	}

	ipc.opt = NULL;
	ipc.oif = sk->bound_dev_if;
	if (msg->msg_controllen) {
		err = ip_cmsg_send(msg, &ipc);
		if (err)
			return err;
		if (ipc.opt)
			free = 1;
		connected = 0;
	}
	if (!ipc.opt)
		ipc.opt = sk->opt;

	ufh.saddr = ipc.addr;
	ipc.addr = daddr = ufh.daddr;

	if (ipc.opt && ipc.opt->srr) {
		if (!daddr)
			return -EINVAL;
		daddr = ipc.opt->faddr;
		connected = 0;
	}
	tos = RT_TOS(sk->ip_tos);
	if (sk->localroute || (msg->msg_flags&MSG_DONTROUTE) ||
	    (ipc.opt && ipc.opt->is_strictroute)) {
		tos |= RTO_ONLINK;
		connected = 0;
	}

	if (MULTICAST(daddr)) {
		if (!ipc.oif)
			ipc.oif = sk->ip_mc_index;
		if (!ufh.saddr)
			ufh.saddr = sk->ip_mc_addr;
		connected = 0;
	}

	if (connected && sk->dst_cache) {
		rt = (struct rtable*)sk->dst_cache;
		if (rt->u.dst.obsolete) {
			sk->dst_cache = NULL;
			dst_release(&rt->u.dst);
			rt = NULL;
		} else
			dst_clone(&rt->u.dst);
	}

	if (rt == NULL) {
		err = ip_route_output(&rt, daddr, ufh.saddr,
#ifdef CONFIG_IP_TRANSPARENT_PROXY
			(msg->msg_flags&MSG_PROXY ? RTO_TPROXY : 0) |
#endif
			 tos, ipc.oif);
		if (err)
			goto out;

		err = -EACCES;
		if (rt->rt_flags&RTCF_BROADCAST && !sk->broadcast)
			goto out;
		if (connected && sk->dst_cache == NULL)
			sk->dst_cache = dst_clone(&rt->u.dst);
	}

	ufh.saddr = rt->rt_src;
	if (!ipc.addr)
		ufh.daddr = ipc.addr = rt->rt_dst;
	ufh.uh.len = htons(ulen);
	ufh.uh.check = 0;
	ufh.iov = msg->msg_iov;
	ufh.wcheck = 0;

	/* RFC1122: OK.  Provides the checksumming facility (MUST) as per */
	/* 4.1.3.4. It's configurable by the application via setsockopt() */
	/* (MAY) and it defaults to on (MUST). */

	err = ip_build_xmit(sk,sk->no_check ? udp_getfrag_nosum : udp_getfrag,
			    &ufh, ulen, &ipc, rt, msg->msg_flags);

out:
	ip_rt_put(rt);
	if (free)
		kfree(ipc.opt);
	if (!err) {
		udp_statistics.UdpOutDatagrams++;
		return len;
	}
	return err;
}

#ifdef _HURD_

#define udp_ioctl 0

#else

/*
 *	IOCTL requests applicable to the UDP protocol
 */

int udp_ioctl(struct sock *sk, int cmd, unsigned long arg)
{
	switch(cmd)
	{
		case TIOCOUTQ:
		{
			unsigned long amount;

			amount = sock_wspace(sk);
			return put_user(amount, (int *)arg);
		}

		case TIOCINQ:
		{
			struct sk_buff *skb;
			unsigned long amount;

			amount = 0;
			/* N.B. Is this interrupt safe??
			   -> Yes. Interrupts do not remove skbs. --ANK (980725)
			 */
			skb = skb_peek(&sk->receive_queue);
			if (skb != NULL) {
				/*
				 * We will only return the amount
				 * of this packet since that is all
				 * that will be read.
				 */
				amount = skb->len - sizeof(struct udphdr);
			}
			return put_user(amount, (int *)arg);
		}

		default:
			return(-ENOIOCTLCMD);
	}
	return(0);
}

#endif

#ifndef HAVE_CSUM_COPY_USER
#undef CONFIG_UDP_DELAY_CSUM
#endif

/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

int udp_recvmsg(struct sock *sk, struct msghdr *msg, int len,
		int noblock, int flags, int *addr_len)
{
  	struct sockaddr_in *sin = (struct sockaddr_in *)msg->msg_name;
  	struct sk_buff *skb;
  	int copied, err;

	if (flags & MSG_ERRQUEUE)
		return ip_recv_error(sk, msg, len);

	/*
	 *	From here the generic datagram does a lot of the work. Come
	 *	the finished NET3, it will do _ALL_ the work!
	 */

	skb = skb_recv_datagram(sk, flags, noblock, &err);
	if (!skb)
		goto out;

  	copied = skb->len - sizeof(struct udphdr);
	if (copied > len) {
		copied = len;
		msg->msg_flags |= MSG_TRUNC;
	}

#ifndef CONFIG_UDP_DELAY_CSUM
	err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov,
					copied);
#else
	if (skb->ip_summed==CHECKSUM_UNNECESSARY) {
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov,
					      copied);
	} else if (copied > msg->msg_iov[0].iov_len || (msg->msg_flags&MSG_TRUNC)) {
		if ((unsigned short)csum_fold(csum_partial(skb->h.raw, skb->len, skb->csum)))
			goto csum_copy_err;
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov,
					      copied);
	} else {
		unsigned int csum;

		err = 0;
		csum = csum_partial(skb->h.raw, sizeof(struct udphdr), skb->csum);
		csum = csum_and_copy_to_user((char*)&skb->h.uh[1], msg->msg_iov[0].iov_base,
					     copied, csum, &err);
		if (err)
			goto out_free;
		if ((unsigned short)csum_fold(csum))
			goto csum_copy_err;
	}
#endif
	if (err)
		goto out_free;
	sk->stamp=skb->stamp;

	/* Copy the address. */
	if (sin)
	{
		/*
		 *	Check any passed addresses
		 */
		if (addr_len)
			*addr_len=sizeof(*sin);

		sin->sin_family = AF_INET;
		sin->sin_port = skb->h.uh->source;
		sin->sin_addr.s_addr = skb->nh.iph->saddr;
#ifdef CONFIG_IP_TRANSPARENT_PROXY
		if (flags&MSG_PROXY)
		{
			/*
			 * We map the first 8 bytes of a second sockaddr_in
			 * into the last 8 (unused) bytes of a sockaddr_in.
			 * This _is_ ugly, but it's the only way to do it
			 * easily,  without adding system calls.
			 */
			struct sockaddr_in *sinto =
				(struct sockaddr_in *) sin->sin_zero;

			sinto->sin_family = AF_INET;
			sinto->sin_port = skb->h.uh->dest;
			sinto->sin_addr.s_addr = skb->nh.iph->daddr;
		}
#endif
  	}
	if (sk->ip_cmsg_flags)
		ip_cmsg_recv(msg, skb);
	err = copied;

out_free:
  	skb_free_datagram(sk, skb);
out:
  	return err;

#ifdef CONFIG_UDP_DELAY_CSUM
csum_copy_err:
	udp_statistics.UdpInErrors++;
	skb_free_datagram(sk, skb);

	/*
	 * Error for blocking case is chosen to masquerade
   	 * as some normal condition.
	 */
	return (flags&MSG_DONTWAIT) ? -EAGAIN : -EHOSTUNREACH;
#endif
}

int udp_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
{
	struct sockaddr_in *usin = (struct sockaddr_in *) uaddr;
	struct rtable *rt;
	int err;


	if (addr_len < sizeof(*usin))
	  	return(-EINVAL);

	/*
	 *	1003.1g - break association.
	 */

	if (usin->sin_family==AF_UNSPEC)
	{
		sk->saddr=INADDR_ANY;
		sk->rcv_saddr=INADDR_ANY;
		sk->daddr=INADDR_ANY;
		sk->state = TCP_CLOSE;
		if(uh_cache_sk == sk)
			uh_cache_sk = NULL;
		return 0;
	}

	if (usin->sin_family && usin->sin_family != AF_INET)
	  	return(-EAFNOSUPPORT);

	dst_release(xchg(&sk->dst_cache, NULL));

	err = ip_route_connect(&rt, usin->sin_addr.s_addr, sk->saddr,
			       sk->ip_tos|sk->localroute, sk->bound_dev_if);
	if (err)
		return err;
	if ((rt->rt_flags&RTCF_BROADCAST) && !sk->broadcast) {
		ip_rt_put(rt);
		return -EACCES;
	}
  	if(!sk->saddr)
	  	sk->saddr = rt->rt_src;		/* Update source address */
	if(!sk->rcv_saddr)
		sk->rcv_saddr = rt->rt_src;
	sk->daddr = rt->rt_dst;
	sk->dport = usin->sin_port;
	sk->state = TCP_ESTABLISHED;

	if(uh_cache_sk == sk)
		uh_cache_sk = NULL;

	sk->dst_cache = &rt->u.dst;
	return(0);
}


static void udp_close(struct sock *sk, long timeout)
{
	/* See for explanation: raw_close in ipv4/raw.c */
	sk->state = TCP_CLOSE;
	udp_v4_unhash(sk);
	sk->dead = 1;
	destroy_sock(sk);
}

static int udp_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
{
	/*
	 *	Charge it to the socket, dropping if the queue is full.
	 */

#if defined(CONFIG_FILTER) && defined(CONFIG_UDP_DELAY_CSUM)
	if (sk->filter && skb->ip_summed != CHECKSUM_UNNECESSARY) {
		if ((unsigned short)csum_fold(csum_partial(skb->h.raw, skb->len, skb->csum))) {
			udp_statistics.UdpInErrors++;
			ip_statistics.IpInDiscards++;
			ip_statistics.IpInDelivers--;
			kfree_skb(skb);
			return -1;
		}
		skb->ip_summed = CHECKSUM_UNNECESSARY;
	}
#endif

	if (sock_queue_rcv_skb(sk,skb)<0) {
		udp_statistics.UdpInErrors++;
		ip_statistics.IpInDiscards++;
		ip_statistics.IpInDelivers--;
		kfree_skb(skb);
		return -1;
	}
	udp_statistics.UdpInDatagrams++;
	return 0;
}


static inline void udp_deliver(struct sock *sk, struct sk_buff *skb)
{
	udp_queue_rcv_skb(sk, skb);
}

/*
 *	Multicasts and broadcasts go to each listener.
 *
 *	Note: called only from the BH handler context,
 *	so we don't need to lock the hashes.
 */
static int udp_v4_mcast_deliver(struct sk_buff *skb, struct udphdr *uh,
				 u32 saddr, u32 daddr)
{
	struct sock *sk;
	int dif;

	sk = udp_hash[ntohs(uh->dest) & (UDP_HTABLE_SIZE - 1)];
	dif = skb->dev->ifindex;
	sk = udp_v4_mcast_next(sk, uh->dest, saddr, uh->source, daddr, dif);
	if (sk) {
		struct sock *sknext = NULL;

		do {
			struct sk_buff *skb1 = skb;

			sknext = udp_v4_mcast_next(sk->next, uh->dest, saddr,
						   uh->source, daddr, dif);
			if(sknext)
				skb1 = skb_clone(skb, GFP_ATOMIC);

			if(skb1)
				udp_deliver(sk, skb1);
			sk = sknext;
		} while(sknext);
	} else
		kfree_skb(skb);
	return 0;
}

#ifdef CONFIG_IP_TRANSPARENT_PROXY
/*
 *	Check whether a received UDP packet might be for one of our
 *	sockets.
 */

int udp_chkaddr(struct sk_buff *skb)
{
	struct iphdr *iph = skb->nh.iph;
	struct udphdr *uh = (struct udphdr *)(skb->nh.raw + iph->ihl*4);
	struct sock *sk;

	sk = udp_v4_lookup(iph->saddr, uh->source, iph->daddr, uh->dest, skb->dev->ifindex);
	if (!sk)
		return 0;

	/* 0 means accept all LOCAL addresses here, not all the world... */
	if (sk->rcv_saddr == 0)
		return 0;

	return 1;
}
#endif

/*
 *	All we need to do is get the socket, and then do a checksum.
 */

int udp_rcv(struct sk_buff *skb, unsigned short len)
{
  	struct sock *sk;
  	struct udphdr *uh;
	unsigned short ulen;
	struct rtable *rt = (struct rtable*)skb->dst;
	u32 saddr = skb->nh.iph->saddr;
	u32 daddr = skb->nh.iph->daddr;

	/*
	 * First time through the loop.. Do all the setup stuff
	 * (including finding out the socket we go to etc)
	 */

	/*
	 *	Get the header.
	 */

  	uh = skb->h.uh;
	__skb_pull(skb, skb->h.raw - skb->data);

  	ip_statistics.IpInDelivers++;

	/*
	 *	Validate the packet and the UDP length.
	 */

	ulen = ntohs(uh->len);

	if (ulen > len || ulen < sizeof(*uh)) {
		NETDEBUG(printk(KERN_DEBUG "UDP: short packet: %d/%d\n", ulen, len));
		udp_statistics.UdpInErrors++;
		kfree_skb(skb);
		return(0);
	}
	skb_trim(skb, ulen);

#ifndef CONFIG_UDP_DELAY_CSUM
	if (uh->check &&
	    (((skb->ip_summed==CHECKSUM_HW)&&udp_check(uh,ulen,saddr,daddr,skb->csum)) ||
	     ((skb->ip_summed==CHECKSUM_NONE) &&
	      (udp_check(uh,ulen,saddr,daddr, csum_partial((char*)uh, ulen, 0))))))
		goto csum_error;
#else
	if (uh->check==0)
		skb->ip_summed = CHECKSUM_UNNECESSARY;
	else if (skb->ip_summed==CHECKSUM_HW) {
		if (udp_check(uh,ulen,saddr,daddr,skb->csum))
			goto csum_error;
		skb->ip_summed = CHECKSUM_UNNECESSARY;
	} else if (skb->ip_summed != CHECKSUM_UNNECESSARY)
		skb->csum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0);
#endif

	if(rt->rt_flags & (RTCF_BROADCAST|RTCF_MULTICAST))
		return udp_v4_mcast_deliver(skb, uh, saddr, daddr);

#ifdef CONFIG_IP_TRANSPARENT_PROXY
	if (IPCB(skb)->redirport)
		sk = udp_v4_proxy_lookup(uh->dest, saddr, uh->source,
					 daddr, skb->dev, IPCB(skb)->redirport,
					 skb->dev->ifindex);
	else
#endif
	sk = udp_v4_lookup(saddr, uh->source, daddr, uh->dest, skb->dev->ifindex);

	if (sk == NULL) {
#ifdef CONFIG_UDP_DELAY_CSUM
		if (skb->ip_summed != CHECKSUM_UNNECESSARY &&
		    (unsigned short)csum_fold(csum_partial((char*)uh, ulen, skb->csum)))
			goto csum_error;
#endif
  		udp_statistics.UdpNoPorts++;
		icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);

		/*
		 * Hmm.  We got an UDP broadcast to a port to which we
		 * don't wanna listen.  Ignore it.
		 */
		kfree_skb(skb);
		return(0);
  	}
	udp_deliver(sk, skb);
	return 0;

csum_error:
	/*
	 * RFC1122: OK.  Discards the bad packet silently (as far as
	 * the network is concerned, anyway) as per 4.1.3.4 (MUST).
	 */
	NETDEBUG(printk(KERN_DEBUG "UDP: bad checksum. From %d.%d.%d.%d:%d to %d.%d.%d.%d:%d ulen %d\n",
			NIPQUAD(saddr),
			ntohs(uh->source),
			NIPQUAD(daddr),
			ntohs(uh->dest),
			ulen));
	udp_statistics.UdpInErrors++;
	kfree_skb(skb);
	return(0);
}

struct proto udp_prot = {
	(struct sock *)&udp_prot,	/* sklist_next */
	(struct sock *)&udp_prot,	/* sklist_prev */
	udp_close,			/* close */
	udp_connect,			/* connect */
	NULL,				/* accept */
	NULL,				/* retransmit */
	NULL,				/* write_wakeup */
	NULL,				/* read_wakeup */
	datagram_poll,			/* poll */
	udp_ioctl,			/* ioctl */
	NULL,				/* init */
	NULL,				/* destroy */
	NULL,				/* shutdown */
	ip_setsockopt,			/* setsockopt */
	ip_getsockopt,			/* getsockopt */
	udp_sendmsg,			/* sendmsg */
	udp_recvmsg,			/* recvmsg */
	NULL,				/* bind */
	udp_queue_rcv_skb,		/* backlog_rcv */
	udp_v4_hash,			/* hash */
	udp_v4_unhash,			/* unhash */
	udp_v4_get_port,		/* good_socknum */
	128,				/* max_header */
	0,				/* retransmits */
 	"UDP",				/* name */
	0,				/* inuse */
	0				/* highestinuse */
};