raw.c

嵌入式系统设计与实例开发实验教材二源码 多线程应用程序设计 串行端口程序设计 AD接口实验 CAN总线通信实验 GPS通信实验 Linux内核移植与编译实验 IC卡读写实验 SD驱动使

/*
 * 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.
 *
 *		RAW - implementation of IP "raw" sockets.
 *
 * Version:	$Id: raw.c,v 1.63 2001/07/10 04:29:01 davem Exp $
 *
 * Authors:	Ross Biro, <[e-mail]>
 *		Fred N. van Kempen, <[e-mail]>
 *
 * Fixes:
 *		Alan Cox	:	verify_area() fixed up
 *		Alan Cox	:	ICMP error handling
 *		Alan Cox	:	EMSGSIZE if you send too big a packet
 *		Alan Cox	: 	Now uses generic datagrams and shared
 *					skbuff library. No more peek crashes,
 *					no more backlogs
 *		Alan Cox	:	Checks sk->broadcast.
 *		Alan Cox	:	Uses skb_free_datagram/skb_copy_datagram
 *		Alan Cox	:	Raw passes ip options too
 *		Alan Cox	:	Setsocketopt added
 *		Alan Cox	:	Fixed error return for broadcasts
 *		Alan Cox	:	Removed wake_up calls
 *		Alan Cox	:	Use ttl/tos
 *		Alan Cox	:	Cleaned up old debugging
 *		Alan Cox	:	Use new kernel side addresses
 *	Arnt Gulbrandsen	:	Fixed MSG_DONTROUTE in raw sockets.
 *		Alan Cox	:	BSD style RAW socket demultiplexing.
 *		Alan Cox	:	Beginnings of mrouted support.
 *		Alan Cox	:	Added IP_HDRINCL option.
 *		Alan Cox	:	Skip broadcast check if BSDism set.
 *		David S. Miller	:	New socket lookup architecture.
 *
 *		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 <asm/system.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
#include <linux/types.h>
#include <linux/sched.h>
#include <linux/errno.h>
#include <linux/timer.h>
#include <linux/mm.h>
#include <linux/kernel.h>
#include <linux/fcntl.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/mroute.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/icmp.h>
#include <net/udp.h>
#include <net/raw.h>
#include <net/inet_common.h>
#include <net/checksum.h>

struct sock *raw_v4_htable[RAWV4_HTABLE_SIZE];
rwlock_t raw_v4_lock = RW_LOCK_UNLOCKED;

static void raw_v4_hash(struct sock *sk)
{
	struct sock **skp = &raw_v4_htable[sk->num & (RAWV4_HTABLE_SIZE - 1)];

	write_lock_bh(&raw_v4_lock);
	if ((sk->next = *skp) != NULL)
		(*skp)->pprev = &sk->next;
	*skp = sk;
	sk->pprev = skp;
	sock_prot_inc_use(sk->prot);
 	sock_hold(sk);
	write_unlock_bh(&raw_v4_lock);
}

static void raw_v4_unhash(struct sock *sk)
{
 	write_lock_bh(&raw_v4_lock);
	if (sk->pprev) {
		if (sk->next)
			sk->next->pprev = sk->pprev;
		*sk->pprev = sk->next;
		sk->pprev = NULL;
		sock_prot_dec_use(sk->prot);
		__sock_put(sk);
	}
	write_unlock_bh(&raw_v4_lock);
}

struct sock *__raw_v4_lookup(struct sock *sk, unsigned short num,
			     unsigned long raddr, unsigned long laddr,
			     int dif)
{
	struct sock *s = sk;

	for (s = sk; s; s = s->next) {
		if (s->num == num 				&&
		    !(s->daddr && s->daddr != raddr) 		&&
		    !(s->rcv_saddr && s->rcv_saddr != laddr)	&&
		    !(s->bound_dev_if && s->bound_dev_if != dif))
			break; /* gotcha */
	}
	return s;
}

/*
 *	0 - deliver
 *	1 - block
 */
static __inline__ int icmp_filter(struct sock *sk, struct sk_buff *skb)
{
	int type;

	type = skb->h.icmph->type;
	if (type < 32) {
		__u32 data = sk->tp_pinfo.tp_raw4.filter.data;

		return ((1 << type) & data) != 0;
	}

	/* Do not block unknown ICMP types */
	return 0;
}

/* IP input processing comes here for RAW socket delivery.
 * This is fun as to avoid copies we want to make no surplus
 * copies.
 *
 * RFC 1122: SHOULD pass TOS value up to the transport layer.
 * -> It does. And not only TOS, but all IP header.
 */
struct sock *raw_v4_input(struct sk_buff *skb, struct iphdr *iph, int hash)
{
	struct sock *sk;

	read_lock(&raw_v4_lock);
	if ((sk = raw_v4_htable[hash]) == NULL)
		goto out;
	sk = __raw_v4_lookup(sk, iph->protocol,
			     iph->saddr, iph->daddr,
			     skb->dev->ifindex);

	while (sk) {
		struct sock *sknext = __raw_v4_lookup(sk->next, iph->protocol,
						      iph->saddr, iph->daddr,
						      skb->dev->ifindex);
		if (iph->protocol != IPPROTO_ICMP ||
		    !icmp_filter(sk, skb)) {
			struct sk_buff *clone;

			if (!sknext)
				break;
			clone = skb_clone(skb, GFP_ATOMIC);
			/* Not releasing hash table! */
			if (clone)
				raw_rcv(sk, clone);
		}
		sk = sknext;
	}
out:
	if (sk)
		sock_hold(sk);
	read_unlock(&raw_v4_lock);

	return sk;
}

void raw_err (struct sock *sk, struct sk_buff *skb, u32 info)
{
	int type = skb->h.icmph->type;
	int code = skb->h.icmph->code;
	int err = 0;
	int harderr = 0;

	/* Report error on raw socket, if:
	   1. User requested ip_recverr.
	   2. Socket is connected (otherwise the error indication
	      is useless without ip_recverr and error is hard.
	 */
	if (!sk->protinfo.af_inet.recverr && sk->state != TCP_ESTABLISHED)
		return;

	switch (type) {
	default:
	case ICMP_TIME_EXCEEDED:
		err = EHOSTUNREACH;
		break;
	case ICMP_SOURCE_QUENCH:
		return;
	case ICMP_PARAMETERPROB:
		err = EPROTO;
		harderr = 1;
		break;
	case ICMP_DEST_UNREACH:
		err = EHOSTUNREACH;
		if (code > NR_ICMP_UNREACH)
			break;
		err = icmp_err_convert[code].errno;
		harderr = icmp_err_convert[code].fatal;
		if (code == ICMP_FRAG_NEEDED) {
			harderr = sk->protinfo.af_inet.pmtudisc !=
					IP_PMTUDISC_DONT;
			err = EMSGSIZE;
		}
	}

	if (sk->protinfo.af_inet.recverr) {
		struct iphdr *iph = (struct iphdr*)skb->data;
		u8 *payload = skb->data + (iph->ihl << 2);

		if (sk->protinfo.af_inet.hdrincl)
			payload = skb->data;
		ip_icmp_error(sk, skb, err, 0, info, payload);
	}

	if (sk->protinfo.af_inet.recverr || harderr) {
		sk->err = err;
		sk->error_report(sk);
	}
}

static int raw_rcv_skb(struct sock * sk, struct sk_buff * skb)
{
	/* Charge it to the socket. */
	
	if (sock_queue_rcv_skb(sk, skb) < 0) {
		IP_INC_STATS(IpInDiscards);
		kfree_skb(skb);
		return NET_RX_DROP;
	}

	IP_INC_STATS(IpInDelivers);
	return NET_RX_SUCCESS;
}

int raw_rcv(struct sock *sk, struct sk_buff *skb)
{
	skb_push(skb, skb->data - skb->nh.raw);

	raw_rcv_skb(sk, skb);
	return 0;
}

struct rawfakehdr 
{
	struct	iovec *iov;
	u32	saddr;
	struct	dst_entry *dst;
};

/*
 *	Send a RAW IP packet.
 */

/*
 *	Callback support is trivial for SOCK_RAW
 */
  
static int raw_getfrag(const void *p, char *to, unsigned int offset,
			unsigned int fraglen)
{
	struct rawfakehdr *rfh = (struct rawfakehdr *) p;
	return memcpy_fromiovecend(to, rfh->iov, offset, fraglen);
}

/*
 *	IPPROTO_RAW needs extra work.
 */
 
static int raw_getrawfrag(const void *p, char *to, unsigned int offset,
				unsigned int fraglen)
{
	struct rawfakehdr *rfh = (struct rawfakehdr *) p;

	if (memcpy_fromiovecend(to, rfh->iov, offset, fraglen))
		return -EFAULT;

	if (!offset) {
		struct iphdr *iph = (struct iphdr *)to;
		if (!iph->saddr)
			iph->saddr = rfh->saddr;
		iph->check   = 0;
		iph->tot_len = htons(fraglen); /* This is right as you can't
						  frag RAW packets */
		/*
	 	 *	Deliberate breach of modularity to keep 
	 	 *	ip_build_xmit clean (well less messy).
		 */
		if (!iph->id)
			ip_select_ident(iph, rfh->dst, NULL);
		iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl);
	}
	return 0;
}

static int raw_sendmsg(struct sock *sk, struct msghdr *msg, int len)
{
	struct ipcm_cookie ipc;
	struct rawfakehdr rfh;
	struct rtable *rt = NULL;
	int free = 0;
	u32 daddr;
	u8  tos;
	int err;

	/* This check is ONLY to check for arithmetic overflow
	   on integer(!) len. Not more! Real check will be made
	   in ip_build_xmit --ANK

	   BTW socket.c -> af_*.c -> ... make multiple
	   invalid conversions size_t -> int. We MUST repair it f.e.
	   by replacing all of them with size_t and revise all
	   the places sort of len += sizeof(struct iphdr)
	   If len was ULONG_MAX-10 it would be cathastrophe  --ANK
	 */

	err = -EMSGSIZE;
	if (len < 0 || len > 0xFFFF)
		goto out;

	/*
	 *	Check the flags.
	 */

	err = -EOPNOTSUPP;
	if (msg->msg_flags & MSG_OOB)	/* Mirror BSD error message */
		goto out;               /* compatibility */
			 
	/*
	 *	Get and verify the address. 
	 */

	if (msg->msg_namelen) {
		struct sockaddr_in *usin = (struct sockaddr_in*)msg->msg_name;
		err = -EINVAL;
		if (msg->msg_namelen < sizeof(*usin))
			goto out;
		if (usin->sin_family != AF_INET) {
			static int complained;
			if (!complained++)
				printk(KERN_INFO "%s forgot to set AF_INET in "