af_packet.c

嵌入式系统设计与实验教材二源码linux内核移植与编译

/*
 * 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.
 *
 *		PACKET - implements raw packet sockets.
 *
 * Version:	$Id: af_packet.c,v 1.58 2001/11/28 21:02:10 davem Exp $
 *
 * Authors:	Ross Biro, <[e-mail]>
 *		Fred N. van Kempen, <[e-mail]>
 *		Alan Cox, <[e-mail]>
 *
 * Fixes:	
 *		Alan Cox	:	verify_area() now used correctly
 *		Alan Cox	:	new skbuff lists, look ma no backlogs!
 *		Alan Cox	:	tidied skbuff lists.
 *		Alan Cox	:	Now uses generic datagram routines I
 *					added. Also fixed the peek/read crash
 *					from all old Linux datagram code.
 *		Alan Cox	:	Uses the improved datagram code.
 *		Alan Cox	:	Added NULL's for socket options.
 *		Alan Cox	:	Re-commented the code.
 *		Alan Cox	:	Use new kernel side addressing
 *		Rob Janssen	:	Correct MTU usage.
 *		Dave Platt	:	Counter leaks caused by incorrect
 *					interrupt locking and some slightly
 *					dubious gcc output. Can you read
 *					compiler: it said _VOLATILE_
 *	Richard Kooijman	:	Timestamp fixes.
 *		Alan Cox	:	New buffers. Use sk->mac.raw.
 *		Alan Cox	:	sendmsg/recvmsg support.
 *		Alan Cox	:	Protocol setting support
 *	Alexey Kuznetsov	:	Untied from IPv4 stack.
 *	Cyrus Durgin		:	Fixed kerneld for kmod.
 *	Michal Ostrowski        :       Module initialization cleanup.
 *
 *		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 <linux/types.h>
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/fcntl.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/if_packet.h>
#include <linux/wireless.h>
#include <linux/kmod.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <linux/errno.h>
#include <linux/timer.h>
#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
#include <linux/proc_fs.h>
#include <linux/poll.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/if_bridge.h>

#ifdef CONFIG_NET_DIVERT
#include <linux/divert.h>
#endif /* CONFIG_NET_DIVERT */

#ifdef CONFIG_INET
#include <net/inet_common.h>
#endif

#ifdef CONFIG_DLCI
extern int dlci_ioctl(unsigned int, void*);
#endif

#define CONFIG_SOCK_PACKET	1

/*
   Proposed replacement for SIOC{ADD,DEL}MULTI and
   IFF_PROMISC, IFF_ALLMULTI flags.

   It is more expensive, but I believe,
   it is really correct solution: reentereble, safe and fault tolerant.

   IFF_PROMISC/IFF_ALLMULTI/SIOC{ADD/DEL}MULTI are faked by keeping
   reference count and global flag, so that real status is
   (gflag|(count != 0)), so that we can use obsolete faulty interface
   not harming clever users.
 */
#define CONFIG_PACKET_MULTICAST	1

/*
   Assumptions:
   - if device has no dev->hard_header routine, it adds and removes ll header
     inside itself. In this case ll header is invisible outside of device,
     but higher levels still should reserve dev->hard_header_len.
     Some devices are enough clever to reallocate skb, when header
     will not fit to reserved space (tunnel), another ones are silly
     (PPP).
   - packet socket receives packets with pulled ll header,
     so that SOCK_RAW should push it back.

On receive:
-----------

Incoming, dev->hard_header!=NULL
   mac.raw -> ll header
   data    -> data

Outgoing, dev->hard_header!=NULL
   mac.raw -> ll header
   data    -> ll header

Incoming, dev->hard_header==NULL
   mac.raw -> UNKNOWN position. It is very likely, that it points to ll header.
              PPP makes it, that is wrong, because introduce assymetry
	      between rx and tx paths.
   data    -> data

Outgoing, dev->hard_header==NULL
   mac.raw -> data. ll header is still not built!
   data    -> data

Resume
  If dev->hard_header==NULL we are unlikely to restore sensible ll header.


On transmit:
------------

dev->hard_header != NULL
   mac.raw -> ll header
   data    -> ll header

dev->hard_header == NULL (ll header is added by device, we cannot control it)
   mac.raw -> data
   data -> data

   We should set nh.raw on output to correct posistion,
   packet classifier depends on it.
 */

/* List of all packet sockets. */
static struct sock * packet_sklist;
static rwlock_t packet_sklist_lock = RW_LOCK_UNLOCKED;

atomic_t packet_socks_nr;


/* Private packet socket structures. */

#ifdef CONFIG_PACKET_MULTICAST
struct packet_mclist
{
	struct packet_mclist	*next;
	int			ifindex;
	int			count;
	unsigned short		type;
	unsigned short		alen;
	unsigned char		addr[8];
};
#endif
#ifdef CONFIG_PACKET_MMAP
static int packet_set_ring(struct sock *sk, struct tpacket_req *req, int closing);
#endif

static void packet_flush_mclist(struct sock *sk);

struct packet_opt
{
	struct packet_type	prot_hook;
	spinlock_t		bind_lock;
	char			running;	/* prot_hook is attached*/
	int			ifindex;	/* bound device		*/
	struct tpacket_stats	stats;
#ifdef CONFIG_PACKET_MULTICAST
	struct packet_mclist	*mclist;
#endif
#ifdef CONFIG_PACKET_MMAP
	atomic_t		mapped;
	unsigned long		*pg_vec;
	unsigned int		pg_vec_order;
	unsigned int		pg_vec_pages;
	unsigned int		pg_vec_len;

	struct tpacket_hdr	**iovec;
	unsigned int		frame_size;
	unsigned int		iovmax;
	unsigned int		head;
	int			copy_thresh;
#endif
};

void packet_sock_destruct(struct sock *sk)
{
	BUG_TRAP(atomic_read(&sk->rmem_alloc)==0);
	BUG_TRAP(atomic_read(&sk->wmem_alloc)==0);

	if (!sk->dead) {
		printk("Attempt to release alive packet socket: %p\n", sk);
		return;
	}

	if (sk->protinfo.destruct_hook)
		kfree(sk->protinfo.destruct_hook);
	atomic_dec(&packet_socks_nr);
#ifdef PACKET_REFCNT_DEBUG
	printk(KERN_DEBUG "PACKET socket %p is free, %d are alive\n", sk, atomic_read(&packet_socks_nr));
#endif
	MOD_DEC_USE_COUNT;
}


extern struct proto_ops packet_ops;

#ifdef CONFIG_SOCK_PACKET
extern struct proto_ops packet_ops_spkt;

static int packet_rcv_spkt(struct sk_buff *skb, struct net_device *dev,  struct packet_type *pt)
{
	struct sock *sk;
	struct sockaddr_pkt *spkt;

	/*
	 *	When we registered the protocol we saved the socket in the data
	 *	field for just this event.
	 */

	sk = (struct sock *) pt->data;
	
	/*
	 *	Yank back the headers [hope the device set this
	 *	right or kerboom...]
	 *
	 *	Incoming packets have ll header pulled,
	 *	push it back.
	 *
	 *	For outgoing ones skb->data == skb->mac.raw
	 *	so that this procedure is noop.
	 */

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto out;

	if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
		goto oom;

	spkt = (struct sockaddr_pkt*)skb->cb;

	skb_push(skb, skb->data-skb->mac.raw);

	/*
	 *	The SOCK_PACKET socket receives _all_ frames.
	 */

	spkt->spkt_family = dev->type;
	strncpy(spkt->spkt_device, dev->name, sizeof(spkt->spkt_device));
	spkt->spkt_protocol = skb->protocol;

	/*
	 *	Charge the memory to the socket. This is done specifically
	 *	to prevent sockets using all the memory up.
	 */

	if (sock_queue_rcv_skb(sk,skb) == 0)
		return 0;

out:
	kfree_skb(skb);
oom:
	return 0;
}


/*
 *	Output a raw packet to a device layer. This bypasses all the other
 *	protocol layers and you must therefore supply it with a complete frame
 */
 
static int packet_sendmsg_spkt(struct socket *sock, struct msghdr *msg, int len,
			       struct scm_cookie *scm)
{
	struct sock *sk = sock->sk;
	struct sockaddr_pkt *saddr=(struct sockaddr_pkt *)msg->msg_name;
	struct sk_buff *skb;
	struct net_device *dev;
	unsigned short proto=0;
	int err;
	
	/*
	 *	Get and verify the address. 
	 */

	if (saddr)
	{
		if (msg->msg_namelen < sizeof(struct sockaddr))
			return(-EINVAL);
		if (msg->msg_namelen==sizeof(struct sockaddr_pkt))
			proto=saddr->spkt_protocol;
	}
	else
		return(-ENOTCONN);	/* SOCK_PACKET must be sent giving an address */

	/*
	 *	Find the device first to size check it 
	 */

	saddr->spkt_device[13] = 0;
	dev = dev_get_by_name(saddr->spkt_device);
	err = -ENODEV;
	if (dev == NULL)
		goto out_unlock;
	
	/*
	 *	You may not queue a frame bigger than the mtu. This is the lowest level
	 *	raw protocol and you must do your own fragmentation at this level.
	 */
	 
	err = -EMSGSIZE;
 	if(len>dev->mtu+dev->hard_header_len)
		goto out_unlock;

	err = -ENOBUFS;
	skb = sock_wmalloc(sk, len+dev->hard_header_len+15, 0, GFP_KERNEL);

	/*
	 *	If the write buffer is full, then tough. At this level the user gets to
	 *	deal with the problem - do your own algorithmic backoffs. That's far
	 *	more flexible.
	 */
	 
	if (skb == NULL) 
		goto out_unlock;

	/*
	 *	Fill it in 
	 */
	 
	/* FIXME: Save some space for broken drivers that write a
	 * hard header at transmission time by themselves. PPP is the
	 * notable one here. This should really be fixed at the driver level.
	 */
	skb_reserve(skb,(dev->hard_header_len+15)&~15);
	skb->nh.raw = skb->data;

	/* Try to align data part correctly */
	if (dev->hard_header) {
		skb->data -= dev->hard_header_len;
		skb->tail -= dev->hard_header_len;
	}

	/* Returns -EFAULT on error */
	err = memcpy_fromiovec(skb_put(skb,len), msg->msg_iov, len);
	skb->protocol = proto;
	skb->dev = dev;
	skb->priority = sk->priority;
	if (err)
		goto out_free;

	err = -ENETDOWN;
	if (!(dev->flags & IFF_UP))
		goto out_free;

	/*
	 *	Now send it
	 */

	dev_queue_xmit(skb);
	dev_put(dev);
	return(len);

out_free:
	kfree_skb(skb);
out_unlock:
	if (dev)
		dev_put(dev);
	return err;
}
#endif

/*
   This function makes lazy skb cloning in hope that most of packets
   are discarded by BPF.

   Note tricky part: we DO mangle shared skb! skb->data, skb->len
   and skb->cb are mangled. It works because (and until) packets
   falling here are owned by current CPU. Output packets are cloned
   by dev_queue_xmit_nit(), input packets are processed by net_bh
   sequencially, so that if we return skb to original state on exit,
   we will not harm anyone.
 */

static int packet_rcv(struct sk_buff *skb, struct net_device *dev,  struct packet_type *pt)
{
	struct sock *sk;
	struct sockaddr_ll *sll;
	struct packet_opt *po;
	u8 * skb_head = skb->data;
	int skb_len = skb->len;
#ifdef CONFIG_FILTER
	unsigned snaplen;
#endif

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto drop;

	sk = (struct sock *) pt->data;
	po = sk->protinfo.af_packet;

	skb->dev = dev;

	if (dev->hard_header) {
		/* The device has an explicit notion of ll header,
		   exported to higher levels.

		   Otherwise, the device hides datails of it frame
		   structure, so that corresponding packet head
		   never delivered to user.
		 */
		if (sk->type != SOCK_DGRAM)
			skb_push(skb, skb->data - skb->mac.raw);
		else if (skb->pkt_type == PACKET_OUTGOING) {
			/* Special case: outgoing packets have ll header at head */
			skb_pull(skb, skb->nh.raw - skb->data);
		}
	}

#ifdef CONFIG_FILTER
	snaplen = skb->len;

	if (sk->filter) {
		unsigned res = snaplen;
		struct sk_filter *filter;

		bh_lock_sock(sk);
		if ((filter = sk->filter) != NULL)
			res = sk_run_filter(skb, sk->filter->insns, sk->filter->len);
		bh_unlock_sock(sk);

		if (res == 0)
			goto drop_n_restore;
		if (snaplen > res)
			snaplen = res;
	}
#endif /* CONFIG_FILTER */

	if (atomic_read(&sk->rmem_alloc) + skb->truesize >= (unsigned)sk->rcvbuf)
		goto drop_n_acct;

	if (skb_shared(skb)) {
		struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC);
		if (nskb == NULL)
			goto drop_n_acct;

		if (skb_head != skb->data) {
			skb->data = skb_head;
			skb->len = skb_len;
		}
		kfree_skb(skb);
		skb = nskb;
	}

	sll = (struct sockaddr_ll*)skb->cb;
	sll->sll_family = AF_PACKET;
	sll->sll_hatype = dev->type;
	sll->sll_protocol = skb->protocol;
	sll->sll_pkttype = skb->pkt_type;
	sll->sll_ifindex = dev->ifindex;
	sll->sll_halen = 0;

	if (dev->hard_header_parse)
		sll->sll_halen = dev->hard_header_parse(skb, sll->sll_addr);

#ifdef CONFIG_FILTER
	if (pskb_trim(skb, snaplen))
		goto drop_n_acct;
#endif

	skb_set_owner_r(skb, sk);
	skb->dev = NULL;
	spin_lock(&sk->receive_queue.lock);
	po->stats.tp_packets++;
	__skb_queue_tail(&sk->receive_queue, skb);
	spin_unlock(&sk->receive_queue.lock);
	sk->data_ready(sk,skb->len);
	return 0;

drop_n_acct:
	spin_lock(&sk->receive_queue.lock);
	po->stats.tp_drops++;
	spin_unlock(&sk->receive_queue.lock);

#ifdef CONFIG_FILTER
drop_n_restore:
#endif
	if (skb_head != skb->data && skb_shared(skb)) {
		skb->data = skb_head;
		skb->len = skb_len;
	}
drop:
	kfree_skb(skb);
	return 0;
}

#ifdef CONFIG_PACKET_MMAP
static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev,  struct packet_type *pt)
{
	struct sock *sk;
	struct packet_opt *po;
	struct sockaddr_ll *sll;
	struct tpacket_hdr *h;
	u8 * skb_head = skb->data;
	int skb_len = skb->len;
	unsigned snaplen;
	unsigned long status = TP_STATUS_LOSING|TP_STATUS_USER;
	unsigned short macoff, netoff;
	struct sk_buff *copy_skb = NULL;

	if (skb->pkt_type == PACKET_LOOPBACK)
		goto drop;

	sk = (struct sock *) pt->data;
	po = sk->protinfo.af_packet;

	if (dev->hard_header) {
		if (sk->type != SOCK_DGRAM)
			skb_push(skb, skb->data - skb->mac.raw);
		else if (skb->pkt_type == PACKET_OUTGOING) {
			/* Special case: outgoing packets have ll header at head */
			skb_pull(skb, skb->nh.raw - skb->data);
			if (skb->ip_summed == CHECKSUM_HW)
				status |= TP_STATUS_CSUMNOTREADY;
		}
	}

	snaplen = skb->len;

#ifdef CONFIG_FILTER
	if (sk->filter) {
		unsigned res = snaplen;
		struct sk_filter *filter;

		bh_lock_sock(sk);
		if ((filter = sk->filter) != NULL)
			res = sk_run_filter(skb, sk->filter->insns, sk->filter->len);
		bh_unlock_sock(sk);

		if (res == 0)
			goto drop_n_restore;
		if (snaplen > res)
			snaplen = res;
	}
#endif

	if (sk->type == SOCK_DGRAM) {
		macoff = netoff = TPACKET_ALIGN(TPACKET_HDRLEN) + 16;
	} else {
		unsigned maclen = skb->nh.raw - skb->data;
		netoff = TPACKET_ALIGN(TPACKET_HDRLEN + (maclen < 16 ? 16 : maclen));
		macoff = netoff - maclen;
	}

	if (macoff + snaplen > po->frame_size) {
		if (po->copy_thresh &&
		    atomic_read(&sk->rmem_alloc) + skb->truesize < (unsigned)sk->rcvbuf) {
			if (skb_shared(skb)) {
				copy_skb = skb_clone(skb, GFP_ATOMIC);
			} else {
				copy_skb = skb_get(skb);
				skb_head = skb->data;
			}
			if (copy_skb)
				skb_set_owner_r(copy_skb, sk);
		}
		snaplen = po->frame_size - macoff;
		if ((int)snaplen < 0)
			snaplen = 0;
	}
	if (snaplen > skb->len-skb->data_len)
		snaplen = skb->len-skb->data_len;

	spin_lock(&sk->receive_queue.lock);
	h = po->iovec[po->head];

	if (h->tp_status)
		goto ring_is_full;
	po->head = po->head != po->iovmax ? po->head+1 : 0;
	po->stats.tp_packets++;
	if (copy_skb) {
		status |= TP_STATUS_COPY;
		__skb_queue_tail(&sk->receive_queue, copy_skb);
	}
	if (!po->stats.tp_drops)
		status &= ~TP_STATUS_LOSING;
	spin_unlock(&sk->receive_queue.lock);

	memcpy((u8*)h + macoff, skb->data, snaplen);

	h->tp_len = skb->len;
	h->tp_snaplen = snaplen;
	h->tp_mac = macoff;
	h->tp_net = netoff;
	h->tp_sec = skb->stamp.tv_sec;
	h->tp_usec = skb->stamp.tv_usec;

	sll = (struct sockaddr_ll*)((u8*)h + TPACKET_ALIGN(sizeof(*h)));
	sll->sll_halen = 0;
	if (dev->hard_header_parse)
		sll->sll_halen = dev->hard_header_parse(skb, sll->sll_addr);
	sll->sll_family = AF_PACKET;
	sll->sll_hatype = dev->type;
	sll->sll_protocol = skb->protocol;
	sll->sll_pkttype = skb->pkt_type;
	sll->sll_ifindex = dev->ifindex;

	h->tp_status = status;
	mb();

	{
		struct page *p_start, *p_end;
		u8 *h_end = (u8 *)h + macoff + snaplen - 1;

		p_start = virt_to_page(h);
		p_end = virt_to_page(h_end);
		while (p_start <= p_end) {
			flush_dcache_page(p_start);
			p_start++;
		}