bpf.c

早期freebsd实现

/*
 * Copyright (c) 1990, 1991, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from the Stanford/CMU enet packet filter,
 * (net/enet.c) distributed as part of 4.3BSD, and code contributed
 * to Berkeley by Steven McCanne and Van Jacobson both of Lawrence 
 * Berkeley Laboratory.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *      @(#)bpf.c	8.2 (Berkeley) 3/28/94
 *
 * static char rcsid[] =
 * "$Header: bpf.c,v 1.33 91/10/27 21:21:58 mccanne Exp $";
 */

#include "bpfilter.h"

#if NBPFILTER > 0

#ifndef __GNUC__
#define inline
#else
#define inline __inline
#endif

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/mbuf.h>
#include <sys/buf.h>
#include <sys/time.h>
#include <sys/proc.h>
#include <sys/user.h>
#include <sys/ioctl.h>
#include <sys/map.h>

#include <sys/file.h>
#if defined(sparc) && BSD < 199103
#include <sys/stream.h>
#endif
#include <sys/tty.h>
#include <sys/uio.h>

#include <sys/protosw.h>
#include <sys/socket.h>
#include <net/if.h>

#include <net/bpf.h>
#include <net/bpfdesc.h>

#include <sys/errno.h>

#include <netinet/in.h>
#include <netinet/if_ether.h>
#include <sys/kernel.h>

/*
 * Older BSDs don't have kernel malloc.
 */
#if BSD < 199103
extern bcopy();
static caddr_t bpf_alloc();
#include <net/bpf_compat.h>
#define BPF_BUFSIZE (MCLBYTES-8)
#define UIOMOVE(cp, len, code, uio) uiomove(cp, len, code, uio)
#else
#define BPF_BUFSIZE 4096
#define UIOMOVE(cp, len, code, uio) uiomove(cp, len, uio)
#endif

#define PRINET  26			/* interruptible */

/*
 * The default read buffer size is patchable.
 */
int bpf_bufsize = BPF_BUFSIZE;

/*
 *  bpf_iflist is the list of interfaces; each corresponds to an ifnet
 *  bpf_dtab holds the descriptors, indexed by minor device #
 */
struct bpf_if	*bpf_iflist;
struct bpf_d	bpf_dtab[NBPFILTER];

#if BSD >= 199207
/*
 * bpfilterattach() is called at boot time in new systems.  We do
 * nothing here since old systems will not call this.
 */
/* ARGSUSED */
void
bpfilterattach(n)
	int n;
{
}
#endif

static int	bpf_allocbufs __P((struct bpf_d *));
static int	bpf_allocbufs __P((struct bpf_d *));
static void	bpf_freed __P((struct bpf_d *));
static void	bpf_freed __P((struct bpf_d *));
static void	bpf_ifname __P((struct ifnet *, struct ifreq *));
static void	bpf_ifname __P((struct ifnet *, struct ifreq *));
static void	bpf_mcopy __P((const void *, void *, u_int));
static int	bpf_movein __P((struct uio *, int,
		    struct mbuf **, struct sockaddr *, int *));
static int	bpf_setif __P((struct bpf_d *, struct ifreq *));
static int	bpf_setif __P((struct bpf_d *, struct ifreq *));
static inline void
		bpf_wakeup __P((struct bpf_d *));
static void	catchpacket __P((struct bpf_d *, u_char *, u_int,
		    u_int, void (*)(const void *, void *, u_int)));
static void	reset_d __P((struct bpf_d *));

static int
bpf_movein(uio, linktype, mp, sockp, datlen)
	register struct uio *uio;
	int linktype, *datlen;
	register struct mbuf **mp;
	register struct sockaddr *sockp;
{
	struct mbuf *m;
	int error;
	int len;
	int hlen;

	/*
	 * Build a sockaddr based on the data link layer type.
	 * We do this at this level because the ethernet header
	 * is copied directly into the data field of the sockaddr.
	 * In the case of SLIP, there is no header and the packet
	 * is forwarded as is.
	 * Also, we are careful to leave room at the front of the mbuf
	 * for the link level header.
	 */
	switch (linktype) {

	case DLT_SLIP:
		sockp->sa_family = AF_INET;
		hlen = 0;
		break;

	case DLT_EN10MB:
		sockp->sa_family = AF_UNSPEC;
		/* XXX Would MAXLINKHDR be better? */
		hlen = sizeof(struct ether_header);
		break;

	case DLT_FDDI:
		sockp->sa_family = AF_UNSPEC;
		/* XXX 4(FORMAC)+6(dst)+6(src)+3(LLC)+5(SNAP) */
		hlen = 24;
		break;

	case DLT_NULL:
		sockp->sa_family = AF_UNSPEC;
		hlen = 0;
		break;

	default:
		return (EIO);
	}

	len = uio->uio_resid;
	*datlen = len - hlen;
	if ((unsigned)len > MCLBYTES)
		return (EIO);

	MGET(m, M_WAIT, MT_DATA);
	if (m == 0)
		return (ENOBUFS);
	if (len > MLEN) {
#if BSD >= 199103
		MCLGET(m, M_WAIT);
		if ((m->m_flags & M_EXT) == 0) {
#else
		MCLGET(m);
		if (m->m_len != MCLBYTES) {
#endif
			error = ENOBUFS;
			goto bad;
		}
	}
	m->m_len = len;
	*mp = m;
	/*
	 * Make room for link header.
	 */
	if (hlen != 0) {
		m->m_len -= hlen;
#if BSD >= 199103
		m->m_data += hlen; /* XXX */
#else
		m->m_off += hlen;
#endif
		error = UIOMOVE((caddr_t)sockp->sa_data, hlen, UIO_WRITE, uio);
		if (error)
			goto bad;
	}
	error = UIOMOVE(mtod(m, caddr_t), len - hlen, UIO_WRITE, uio);
	if (!error)
		return (0);
 bad:
	m_freem(m);
	return (error);
}

/*
 * Attach file to the bpf interface, i.e. make d listen on bp.
 * Must be called at splimp.
 */
static void
bpf_attachd(d, bp)
	struct bpf_d *d;
	struct bpf_if *bp;
{
	/*
	 * Point d at bp, and add d to the interface's list of listeners.
	 * Finally, point the driver's bpf cookie at the interface so
	 * it will divert packets to bpf.
	 */
	d->bd_bif = bp;
	d->bd_next = bp->bif_dlist;
	bp->bif_dlist = d;

	*bp->bif_driverp = bp;
}

/*
 * Detach a file from its interface.
 */
static void
bpf_detachd(d)
	struct bpf_d *d;
{
	struct bpf_d **p;
	struct bpf_if *bp;

	bp = d->bd_bif;
	/*
	 * Check if this descriptor had requested promiscuous mode.
	 * If so, turn it off.
	 */
	if (d->bd_promisc) {
		d->bd_promisc = 0;
		if (ifpromisc(bp->bif_ifp, 0))
			/*
			 * Something is really wrong if we were able to put
			 * the driver into promiscuous mode, but can't
			 * take it out.
			 */
			panic("bpf: ifpromisc failed");
	}
	/* Remove d from the interface's descriptor list. */
	p = &bp->bif_dlist;
	while (*p != d) {
		p = &(*p)->bd_next;
		if (*p == 0)
			panic("bpf_detachd: descriptor not in list");
	}
	*p = (*p)->bd_next;
	if (bp->bif_dlist == 0)
		/*
		 * Let the driver know that there are no more listeners.
		 */
		*d->bd_bif->bif_driverp = 0;
	d->bd_bif = 0;
}


/*
 * Mark a descriptor free by making it point to itself.
 * This is probably cheaper than marking with a constant since
 * the address should be in a register anyway.
 */
#define D_ISFREE(d) ((d) == (d)->bd_next)
#define D_MARKFREE(d) ((d)->bd_next = (d))
#define D_MARKUSED(d) ((d)->bd_next = 0)

/*
 * Open ethernet device.  Returns ENXIO for illegal minor device number,
 * EBUSY if file is open by another process.
 */
/* ARGSUSED */
int
bpfopen(dev, flag)
	dev_t dev;
	int flag;
{
	register struct bpf_d *d;

	if (minor(dev) >= NBPFILTER)
		return (ENXIO);
	/*
	 * Each minor can be opened by only one process.  If the requested
	 * minor is in use, return EBUSY.
	 */
	d = &bpf_dtab[minor(dev)];
	if (!D_ISFREE(d))
		return (EBUSY);

	/* Mark "free" and do most initialization. */
	bzero((char *)d, sizeof(*d));
	d->bd_bufsize = bpf_bufsize;

	return (0);
}

/*
 * Close the descriptor by detaching it from its interface,
 * deallocating its buffers, and marking it free.
 */
/* ARGSUSED */
int
bpfclose(dev, flag)
	dev_t dev;
	int flag;
{
	register struct bpf_d *d = &bpf_dtab[minor(dev)];
	register int s;

	s = splimp();
	if (d->bd_bif)
		bpf_detachd(d);
	splx(s);
	bpf_freed(d);

	return (0);
}

/*
 * Support for SunOS, which does not have tsleep.
 */
#if BSD < 199103
static
bpf_timeout(arg)
	caddr_t arg;
{
	struct bpf_d *d = (struct bpf_d *)arg;
	d->bd_timedout = 1;
	wakeup(arg);
}

#define BPF_SLEEP(chan, pri, s, t) bpf_sleep((struct bpf_d *)chan)

int
bpf_sleep(d)
	register struct bpf_d *d;
{
	register int rto = d->bd_rtout;
	register int st;

	if (rto != 0) {
		d->bd_timedout = 0;
		timeout(bpf_timeout, (caddr_t)d, rto);
	}
	st = sleep((caddr_t)d, PRINET|PCATCH);
	if (rto != 0) {
		if (d->bd_timedout == 0)
			untimeout(bpf_timeout, (caddr_t)d);
		else if (st == 0)
			return EWOULDBLOCK;
	}
	return (st != 0) ? EINTR : 0;
}
#else
#define BPF_SLEEP tsleep
#endif

/*
 * Rotate the packet buffers in descriptor d.  Move the store buffer
 * into the hold slot, and the free buffer into the store slot.
 * Zero the length of the new store buffer.
 */
#define ROTATE_BUFFERS(d) \
	(d)->bd_hbuf = (d)->bd_sbuf; \
	(d)->bd_hlen = (d)->bd_slen; \
	(d)->bd_sbuf = (d)->bd_fbuf; \
	(d)->bd_slen = 0; \
	(d)->bd_fbuf = 0;
/*
 *  bpfread - read next chunk of packets from buffers
 */
int
bpfread(dev, uio)
	dev_t dev;
	register struct uio *uio;
{
	register struct bpf_d *d = &bpf_dtab[minor(dev)];
	int error;
	int s;

	/*
	 * Restrict application to use a buffer the same size as
	 * as kernel buffers.
	 */
	if (uio->uio_resid != d->bd_bufsize)
		return (EINVAL);

	s = splimp();
	/*
	 * If the hold buffer is empty, then do a timed sleep, which
	 * ends when the timeout expires or when enough packets
	 * have arrived to fill the store buffer.
	 */
	while (d->bd_hbuf == 0) {
		if (d->bd_immediate && d->bd_slen != 0) {
			/*
			 * A packet(s) either arrived since the previous
			 * read or arrived while we were asleep.
			 * Rotate the buffers and return what's here.
			 */
			ROTATE_BUFFERS(d);
			break;
		}
		error = BPF_SLEEP((caddr_t)d, PRINET|PCATCH, "bpf",
				  d->bd_rtout);
		if (error == EINTR || error == ERESTART) {
			splx(s);
			return (error);
		}
		if (error == EWOULDBLOCK) {
			/*
			 * On a timeout, return what's in the buffer,
			 * which may be nothing.  If there is something
			 * in the store buffer, we can rotate the buffers.
			 */
			if (d->bd_hbuf)
				/*
				 * We filled up the buffer in between
				 * getting the timeout and arriving
				 * here, so we don't need to rotate.
				 */
				break;

			if (d->bd_slen == 0) {
				splx(s);
				return (0);
			}
			ROTATE_BUFFERS(d);
			break;
		}
	}
	/*
	 * At this point, we know we have something in the hold slot.
	 */
	splx(s);

	/*
	 * Move data from hold buffer into user space.
	 * We know the entire buffer is transferred since
	 * we checked above that the read buffer is bpf_bufsize bytes.
	 */
	error = UIOMOVE(d->bd_hbuf, d->bd_hlen, UIO_READ, uio);

	s = splimp();
	d->bd_fbuf = d->bd_hbuf;
	d->bd_hbuf = 0;
	d->bd_hlen = 0;
	splx(s);

	return (error);
}


/*
 * If there are processes sleeping on this descriptor, wake them up.
 */
static inline void
bpf_wakeup(d)
	register struct bpf_d *d;
{
	wakeup((caddr_t)d);
#if BSD >= 199103
	selwakeup(&d->bd_sel);
	/* XXX */
	d->bd_sel.si_pid = 0;
#else
	if (d->bd_selproc) {
		selwakeup(d->bd_selproc, (int)d->bd_selcoll);
		d->bd_selcoll = 0;
		d->bd_selproc = 0;
	}
#endif
}

int
bpfwrite(dev, uio)
	dev_t dev;
	struct uio *uio;
{
	register struct bpf_d *d = &bpf_dtab[minor(dev)];
	struct ifnet *ifp;
	struct mbuf *m;
	int error, s;
	static struct sockaddr dst;
	int datlen;

	if (d->bd_bif == 0)
		return (ENXIO);

	ifp = d->bd_bif->bif_ifp;

	if (uio->uio_resid == 0)
		return (0);

	error = bpf_movein(uio, (int)d->bd_bif->bif_dlt, &m, &dst, &datlen);
	if (error)
		return (error);

	if (datlen > ifp->if_mtu)
		return (EMSGSIZE);

	s = splnet();
#if BSD >= 199103
	error = (*ifp->if_output)(ifp, m, &dst, (struct rtentry *)0);
#else
	error = (*ifp->if_output)(ifp, m, &dst);
#endif
	splx(s);
	/*
	 * The driver frees the mbuf.
	 */
	return (error);
}

/*
 * Reset a descriptor by flushing its packet buffer and clearing the
 * receive and drop counts.  Should be called at splimp.
 */
static void
reset_d(d)
	struct bpf_d *d;
{
	if (d->bd_hbuf) {
		/* Free the hold buffer. */
		d->bd_fbuf = d->bd_hbuf;
		d->bd_hbuf = 0;
	}
	d->bd_slen = 0;
	d->bd_hlen = 0;
	d->bd_rcount = 0;
	d->bd_dcount = 0;
}

/*
 *  FIONREAD		Check for read packet available.
 *  SIOCGIFADDR		Get interface address - convenient hook to driver.
 *  BIOCGBLEN		Get buffer len [for read()].
 *  BIOCSETF		Set ethernet read filter.
 *  BIOCFLUSH		Flush read packet buffer.
 *  BIOCPROMISC		Put interface into promiscuous mode.
 *  BIOCGDLT		Get link layer type.
 *  BIOCGETIF		Get interface name.
 *  BIOCSETIF		Set interface.
 *  BIOCSRTIMEOUT	Set read timeout.
 *  BIOCGRTIMEOUT	Get read timeout.
 *  BIOCGSTATS		Get packet stats.
 *  BIOCIMMEDIATE	Set immediate mode.
 *  BIOCVERSION		Get filter language version.
 */
/* ARGSUSED */
int
bpfioctl(dev, cmd, addr, flag)
	dev_t dev;
	int cmd;
	caddr_t addr;
	int flag;
{
	register struct bpf_d *d = &bpf_dtab[minor(dev)];
	int s, error = 0;

	switch (cmd) {

	default:
		error = EINVAL;
		break;

	/*
	 * Check for read packet available.
	 */
	case FIONREAD:
		{
			int n;

			s = splimp();
			n = d->bd_slen;
			if (d->bd_hbuf)
				n += d->bd_hlen;
			splx(s);

			*(int *)addr = n;
			break;
		}

	case SIOCGIFADDR:
		{
			struct ifnet *ifp;

			if (d->bd_bif == 0)
				error = EINVAL;
			else {
				ifp = d->bd_bif->bif_ifp;
				error = (*ifp->if_ioctl)(ifp, cmd, addr);
			}
			break;
		}

	/*
	 * Get buffer len [for read()].
	 */
	case BIOCGBLEN:
		*(u_int *)addr = d->bd_bufsize;
		break;

	/*
	 * Set buffer length.
	 */
	case BIOCSBLEN:
#if BSD < 199103
		error = EINVAL;
#else
		if (d->bd_bif != 0)
			error = EINVAL;
		else {
			register u_int size = *(u_int *)addr;

			if (size > BPF_MAXBUFSIZE)
				*(u_int *)addr = size = BPF_MAXBUFSIZE;
			else if (size < BPF_MINBUFSIZE)
				*(u_int *)addr = size = BPF_MINBUFSIZE;
			d->bd_bufsize = size;