NetBSD/sys/netns/ns_output.c
2004-04-19 00:10:48 +00:00

167 lines
4.3 KiB
C

/* $NetBSD: ns_output.c,v 1.13 2004/04/19 00:10:48 matt Exp $ */
/*
* Copyright (c) 1984, 1985, 1986, 1987, 1993
* The Regents of the University of California. All rights reserved.
*
* 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. 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.
*
* @(#)ns_output.c 8.1 (Berkeley) 6/10/93
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: ns_output.c,v 1.13 2004/04/19 00:10:48 matt Exp $");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/errno.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <net/if.h>
#include <net/route.h>
#include <netns/ns.h>
#include <netns/ns_if.h>
#include <netns/ns_var.h>
#include <netns/idp.h>
#include <netns/idp_var.h>
#include <machine/stdarg.h>
int ns_hold_output = 0;
int ns_copy_output = 0;
int ns_output_cnt = 0;
struct mbuf *ns_lastout;
int
ns_output(struct mbuf *m0, ...)
{
struct route *ro;
int flags;
struct idp *idp = mtod(m0, struct idp *);
struct ifnet *ifp = 0;
int error = 0;
struct route idproute;
struct sockaddr_ns *dst;
va_list ap;
va_start(ap, m0);
ro = va_arg(ap, struct route *);
flags = va_arg(ap, int);
va_end(ap);
if (ns_hold_output) {
if (ns_lastout) {
(void)m_free(ns_lastout);
}
ns_lastout = m_copy(m0, 0, (int)M_COPYALL);
}
/*
* Route packet.
*/
if (ro == 0) {
ro = &idproute;
bzero((caddr_t)ro, sizeof (*ro));
}
dst = satosns(&ro->ro_dst);
if (ro->ro_rt == 0) {
dst->sns_family = AF_NS;
dst->sns_len = sizeof (*dst);
dst->sns_addr = idp->idp_dna;
dst->sns_addr.x_port = 0;
/*
* If routing to interface only,
* short circuit routing lookup.
*/
if (flags & NS_ROUTETOIF) {
struct ns_ifaddr *ia = ns_iaonnetof(&idp->idp_dna);
if (ia == 0) {
error = ENETUNREACH;
goto bad;
}
ifp = ia->ia_ifp;
goto gotif;
}
rtalloc(ro);
} else if ((ro->ro_rt->rt_flags & RTF_UP) == 0) {
/*
* The old route has gone away; try for a new one.
*/
rtfree(ro->ro_rt);
ro->ro_rt = NULL;
rtalloc(ro);
}
if (ro->ro_rt == 0 || (ifp = ro->ro_rt->rt_ifp) == 0) {
error = ENETUNREACH;
goto bad;
}
ro->ro_rt->rt_use++;
if (ro->ro_rt->rt_flags & (RTF_GATEWAY|RTF_HOST))
dst = satosns(ro->ro_rt->rt_gateway);
gotif:
/*
* Look for multicast addresses and
* and verify user is allowed to send
* such a packet.
*/
if (dst->sns_addr.x_host.c_host[0]&1) {
if ((ifp->if_flags & IFF_BROADCAST) == 0) {
error = EADDRNOTAVAIL;
goto bad;
}
if ((flags & NS_ALLOWBROADCAST) == 0) {
error = EACCES;
goto bad;
}
}
if (ntohs(idp->idp_len) <= ifp->if_mtu) {
ns_output_cnt++;
if (ns_copy_output) {
ns_watch_output(m0, ifp);
}
error = (*ifp->if_output)(ifp, m0, snstosa(dst), ro->ro_rt);
goto done;
} else error = EMSGSIZE;
bad:
if (ns_copy_output) {
ns_watch_output(m0, ifp);
}
m_freem(m0);
done:
if (ro == &idproute && (flags & NS_ROUTETOIF) == 0 && ro->ro_rt) {
RTFREE(ro->ro_rt);
ro->ro_rt = 0;
}
return (error);
}