d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
fad994205a
freebsd-skq/sys/netinet/ip_fw_pfil.c
oleg 7eef73ab3f 1) dummynet_io() declaration has changed. 
2) Alter packet flow inside dummynet: allow certain packets to bypass
dummynet scheduler. Benefits are:

- lower latency: if packet flow does not exceed pipe bandwidth, packets
  will not be (up to tick) delayed (due to dummynet's scheduler granularity).
- lower overhead: if packet avoids dummynet scheduler it shouldn't reenter ip
  stack later. Such packets can be fastforwarded.
- recursion (which can lead to kernel stack exhaution) eliminated. This fix
  long existed panic, which can be triggered this way:
  	kldload dummynet
	sysctl net.inet.ip.fw.one_pass=0
	ipfw pipe 1 config bw 0
	for i in `jot 30`; do ipfw add 1 pipe 1 icmp from any to any; done
	ping -c 1 localhost

3) Three new sysctl nodes are added:
net.inet.ip.dummynet.io_pkt -		packets passed to dummynet
net.inet.ip.dummynet.io_pkt_fast - 	packets avoided dummynet scheduler
net.inet.ip.dummynet.io_pkt_drop -	packets dropped by dummynet

P.S. Above comments are true only for layer 3 packets. Layer 2 packet flow
     is not changed yet.

MFC after:	3 month
2007-11-06 23:01:42 +00:00

571 lines
12 KiB
C
Raw Blame History

/*-
* Copyright (c) 2004 Andre Oppermann, Internet Business Solutions AG
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#if !defined(KLD_MODULE)
#include "opt_ipfw.h"
#include "opt_ipdn.h"
#include "opt_inet.h"
#ifndef INET
#error IPFIREWALL requires INET.
#endif /* INET */
#endif /* KLD_MODULE */
#include "opt_inet6.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <sys/kernel.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/sysctl.h>
#include <sys/ucred.h>
#include <net/if.h>
#include <net/route.h>
#include <net/pfil.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/in_var.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/ip_fw.h>
#include <netinet/ip_divert.h>
#include <netinet/ip_dummynet.h>
#include <netgraph/ng_ipfw.h>
#include <machine/in_cksum.h>
int fw_enable = 1;
#ifdef INET6
int fw6_enable = 1;
#endif
int ipfw_chg_hook(SYSCTL_HANDLER_ARGS);
/* Dummynet hooks. */
ip_dn_ruledel_t *ip_dn_ruledel_ptr = NULL;
/* Divert hooks. */
ip_divert_packet_t *ip_divert_ptr = NULL;
/* ng_ipfw hooks. */
ng_ipfw_input_t *ng_ipfw_input_p = NULL;
/* Forward declarations. */
static int ipfw_divert(struct mbuf **, int, int);
#define DIV_DIR_IN 1
#define DIV_DIR_OUT 0
int
ipfw_check_in(void *arg, struct mbuf **m0, struct ifnet *ifp, int dir,
struct inpcb *inp)
{
struct ip_fw_args args;
struct ng_ipfw_tag *ng_tag;
struct m_tag *dn_tag;
int ipfw = 0;
int divert;
int tee;
#ifdef IPFIREWALL_FORWARD
struct m_tag *fwd_tag;
#endif
KASSERT(dir == PFIL_IN, ("ipfw_check_in wrong direction!"));
bzero(&args, sizeof(args));
ng_tag = (struct ng_ipfw_tag *)m_tag_locate(*m0, NGM_IPFW_COOKIE, 0,
NULL);
if (ng_tag != NULL) {
KASSERT(ng_tag->dir == NG_IPFW_IN,
("ng_ipfw tag with wrong direction"));
args.rule = ng_tag->rule;
m_tag_delete(*m0, (struct m_tag *)ng_tag);
}
again:
dn_tag = m_tag_find(*m0, PACKET_TAG_DUMMYNET, NULL);
if (dn_tag != NULL){
struct dn_pkt_tag *dt;
dt = (struct dn_pkt_tag *)(dn_tag+1);
args.rule = dt->rule;
m_tag_delete(*m0, dn_tag);
}
args.m = *m0;
args.inp = inp;
ipfw = ipfw_chk(&args);
*m0 = args.m;
tee = 0;
KASSERT(*m0 != NULL || ipfw == IP_FW_DENY, ("%s: m0 is NULL",
__func__));
switch (ipfw) {
case IP_FW_PASS:
if (args.next_hop == NULL)
goto pass;
#ifdef IPFIREWALL_FORWARD
fwd_tag = m_tag_get(PACKET_TAG_IPFORWARD,
sizeof(struct sockaddr_in), M_NOWAIT);
if (fwd_tag == NULL)
goto drop;
bcopy(args.next_hop, (fwd_tag+1), sizeof(struct sockaddr_in));
m_tag_prepend(*m0, fwd_tag);
if (in_localip(args.next_hop->sin_addr))
(*m0)->m_flags |= M_FASTFWD_OURS;
goto pass;
#endif
break; /* not reached */
case IP_FW_DENY:
goto drop;
break; /* not reached */
case IP_FW_DUMMYNET:
if (!DUMMYNET_LOADED)
goto drop;
if (mtod(*m0, struct ip *)->ip_v == 4)
ip_dn_io_ptr(m0, DN_TO_IP_IN, &args);
else if (mtod(*m0, struct ip *)->ip_v == 6)
ip_dn_io_ptr(m0, DN_TO_IP6_IN, &args);
if (*m0 != NULL)
goto again;
return 0; /* packet consumed */
case IP_FW_TEE:
tee = 1;
/* fall through */
case IP_FW_DIVERT:
divert = ipfw_divert(m0, DIV_DIR_IN, tee);
if (divert) {
*m0 = NULL;
return 0; /* packet consumed */
} else {
args.rule = NULL;
goto again; /* continue with packet */
}
case IP_FW_NGTEE:
if (!NG_IPFW_LOADED)
goto drop;
(void)ng_ipfw_input_p(m0, NG_IPFW_IN, &args, 1);
goto again; /* continue with packet */
case IP_FW_NETGRAPH:
if (!NG_IPFW_LOADED)
goto drop;
return ng_ipfw_input_p(m0, NG_IPFW_IN, &args, 0);
case IP_FW_NAT:
goto again; /* continue with packet */
default:
KASSERT(0, ("%s: unknown retval", __func__));
}
drop:
if (*m0)
m_freem(*m0);
*m0 = NULL;
return (EACCES);
pass:
return 0; /* not filtered */
}
int
ipfw_check_out(void *arg, struct mbuf **m0, struct ifnet *ifp, int dir,
struct inpcb *inp)
{
struct ip_fw_args args;
struct ng_ipfw_tag *ng_tag;
struct m_tag *dn_tag;
int ipfw = 0;
int divert;
int tee;
#ifdef IPFIREWALL_FORWARD
struct m_tag *fwd_tag;
#endif
KASSERT(dir == PFIL_OUT, ("ipfw_check_out wrong direction!"));
bzero(&args, sizeof(args));
ng_tag = (struct ng_ipfw_tag *)m_tag_locate(*m0, NGM_IPFW_COOKIE, 0,
NULL);
if (ng_tag != NULL) {
KASSERT(ng_tag->dir == NG_IPFW_OUT,
("ng_ipfw tag with wrong direction"));
args.rule = ng_tag->rule;
m_tag_delete(*m0, (struct m_tag *)ng_tag);
}
again:
dn_tag = m_tag_find(*m0, PACKET_TAG_DUMMYNET, NULL);
if (dn_tag != NULL) {
struct dn_pkt_tag *dt;
dt = (struct dn_pkt_tag *)(dn_tag+1);
args.rule = dt->rule;
m_tag_delete(*m0, dn_tag);
}
args.m = *m0;
args.oif = ifp;
args.inp = inp;
ipfw = ipfw_chk(&args);
*m0 = args.m;
tee = 0;
KASSERT(*m0 != NULL || ipfw == IP_FW_DENY, ("%s: m0 is NULL",
__func__));
switch (ipfw) {
case IP_FW_PASS:
if (args.next_hop == NULL)
goto pass;
#ifdef IPFIREWALL_FORWARD
/* Overwrite existing tag. */
fwd_tag = m_tag_find(*m0, PACKET_TAG_IPFORWARD, NULL);
if (fwd_tag == NULL) {
fwd_tag = m_tag_get(PACKET_TAG_IPFORWARD,
sizeof(struct sockaddr_in), M_NOWAIT);
if (fwd_tag == NULL)
goto drop;
} else
m_tag_unlink(*m0, fwd_tag);
bcopy(args.next_hop, (fwd_tag+1), sizeof(struct sockaddr_in));
m_tag_prepend(*m0, fwd_tag);
if (in_localip(args.next_hop->sin_addr))
(*m0)->m_flags |= M_FASTFWD_OURS;
goto pass;
#endif
break; /* not reached */
case IP_FW_DENY:
goto drop;
break; /* not reached */
case IP_FW_DUMMYNET:
if (!DUMMYNET_LOADED)
break;
if (mtod(*m0, struct ip *)->ip_v == 4)
ip_dn_io_ptr(m0, DN_TO_IP_OUT, &args);
else if (mtod(*m0, struct ip *)->ip_v == 6)
ip_dn_io_ptr(m0, DN_TO_IP6_OUT, &args);
if (*m0 != NULL)
goto again;
return 0; /* packet consumed */
break;
case IP_FW_TEE:
tee = 1;
/* fall through */
case IP_FW_DIVERT:
divert = ipfw_divert(m0, DIV_DIR_OUT, tee);
if (divert) {
*m0 = NULL;
return 0; /* packet consumed */
} else {
args.rule = NULL;
goto again; /* continue with packet */
}
case IP_FW_NGTEE:
if (!NG_IPFW_LOADED)
goto drop;
(void)ng_ipfw_input_p(m0, NG_IPFW_OUT, &args, 1);
goto again; /* continue with packet */
case IP_FW_NETGRAPH:
if (!NG_IPFW_LOADED)
goto drop;
return ng_ipfw_input_p(m0, NG_IPFW_OUT, &args, 0);
case IP_FW_NAT:
goto again; /* continue with packet */
default:
KASSERT(0, ("%s: unknown retval", __func__));
}
drop:
if (*m0)
m_freem(*m0);
*m0 = NULL;
return (EACCES);
pass:
return 0; /* not filtered */
}
static int
ipfw_divert(struct mbuf **m, int incoming, int tee)
{
/*
* ipfw_chk() has already tagged the packet with the divert tag.
* If tee is set, copy packet and return original.
* If not tee, consume packet and send it to divert socket.
*/
struct mbuf *clone, *reass;
struct ip *ip;
int hlen;
reass = NULL;
/* Is divert module loaded? */
if (ip_divert_ptr == NULL)
goto nodivert;
/* Cloning needed for tee? */
if (tee)
clone = m_dup(*m, M_DONTWAIT);
else
clone = *m;
/* In case m_dup was unable to allocate mbufs. */
if (clone == NULL)
goto teeout;
/*
* Divert listeners can only handle non-fragmented packets.
* However when tee is set we will *not* de-fragment the packets;
* Doing do would put the reassembly into double-jeopardy. On top
* of that someone doing a tee will probably want to get the packet
* in its original form.
*/
ip = mtod(clone, struct ip *);
if (!tee && ip->ip_off & (IP_MF | IP_OFFMASK)) {
/* Reassemble packet. */
reass = ip_reass(clone);
/*
* IP header checksum fixup after reassembly and leave header
* in network byte order.
*/
if (reass != NULL) {
ip = mtod(reass, struct ip *);
hlen = ip->ip_hl << 2;
ip->ip_len = htons(ip->ip_len);
ip->ip_off = htons(ip->ip_off);
ip->ip_sum = 0;
if (hlen == sizeof(struct ip))
ip->ip_sum = in_cksum_hdr(ip);
else
ip->ip_sum = in_cksum(reass, hlen);
clone = reass;
} else
clone = NULL;
} else {
/* Convert header to network byte order. */
ip->ip_len = htons(ip->ip_len);
ip->ip_off = htons(ip->ip_off);
}
/* Do the dirty job... */
if (clone && ip_divert_ptr != NULL)
ip_divert_ptr(clone, incoming);
teeout:
/*
* For tee we leave the divert tag attached to original packet.
* It will then continue rule evaluation after the tee rule.
*/
if (tee)
return 0;
/* Packet diverted and consumed */
return 1;
nodivert:
m_freem(*m);
return 1;
}
static int
ipfw_hook(void)
{
struct pfil_head *pfh_inet;
pfh_inet = pfil_head_get(PFIL_TYPE_AF, AF_INET);
if (pfh_inet == NULL)
return ENOENT;
pfil_add_hook(ipfw_check_in, NULL, PFIL_IN | PFIL_WAITOK, pfh_inet);
pfil_add_hook(ipfw_check_out, NULL, PFIL_OUT | PFIL_WAITOK, pfh_inet);
return 0;
}
static int
ipfw_unhook(void)
{
struct pfil_head *pfh_inet;
pfh_inet = pfil_head_get(PFIL_TYPE_AF, AF_INET);
if (pfh_inet == NULL)
return ENOENT;
pfil_remove_hook(ipfw_check_in, NULL, PFIL_IN | PFIL_WAITOK, pfh_inet);
pfil_remove_hook(ipfw_check_out, NULL, PFIL_OUT | PFIL_WAITOK, pfh_inet);
return 0;
}
#ifdef INET6
static int
ipfw6_hook(void)
{
struct pfil_head *pfh_inet6;
pfh_inet6 = pfil_head_get(PFIL_TYPE_AF, AF_INET6);
if (pfh_inet6 == NULL)
return ENOENT;
pfil_add_hook(ipfw_check_in, NULL, PFIL_IN | PFIL_WAITOK, pfh_inet6);
pfil_add_hook(ipfw_check_out, NULL, PFIL_OUT | PFIL_WAITOK, pfh_inet6);
return 0;
}
static int
ipfw6_unhook(void)
{
struct pfil_head *pfh_inet6;
pfh_inet6 = pfil_head_get(PFIL_TYPE_AF, AF_INET6);
if (pfh_inet6 == NULL)
return ENOENT;
pfil_remove_hook(ipfw_check_in, NULL, PFIL_IN | PFIL_WAITOK, pfh_inet6);
pfil_remove_hook(ipfw_check_out, NULL, PFIL_OUT | PFIL_WAITOK, pfh_inet6);
return 0;
}
#endif /* INET6 */
int
ipfw_chg_hook(SYSCTL_HANDLER_ARGS)
{
int enable = *(int *)arg1;
int error;
error = sysctl_handle_int(oidp, &enable, 0, req);
if (error)
return (error);
enable = (enable) ? 1 : 0;
if (enable == *(int *)arg1)
return (0);
if (arg1 == &fw_enable) {
if (enable)
error = ipfw_hook();
else
error = ipfw_unhook();
}
#ifdef INET6
if (arg1 == &fw6_enable) {
if (enable)
error = ipfw6_hook();
else
error = ipfw6_unhook();
}
#endif
if (error)
return (error);
*(int *)arg1 = enable;
return (0);
}
static int
ipfw_modevent(module_t mod, int type, void *unused)
{
int err = 0;
switch (type) {
case MOD_LOAD:
if ((err = ipfw_init()) != 0) {
printf("ipfw_init() error\n");
break;
}
if ((err = ipfw_hook()) != 0) {
printf("ipfw_hook() error\n");
break;
}
#ifdef INET6
if ((err = ipfw6_hook()) != 0) {
printf("ipfw_hook() error\n");
break;
}
#endif
break;
case MOD_UNLOAD:
if ((err = ipfw_unhook()) > 0)
break;
#ifdef INET6
if ((err = ipfw6_unhook()) > 0)
break;
#endif
ipfw_destroy();
break;
default:
return EOPNOTSUPP;
break;
}
return err;
}
static moduledata_t ipfwmod = {
"ipfw",
ipfw_modevent,
0
};
DECLARE_MODULE(ipfw, ipfwmod, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY);
MODULE_VERSION(ipfw, 2);
Reference in New Issue View Git Blame Copy Permalink