freebsd-skq/sys/netgraph/ng_bridge.c
Andre Oppermann 9b932e9e04 Convert ipfw to use PFIL_HOOKS. This is change is transparent to userland
and preserves the ipfw ABI.  The ipfw core packet inspection and filtering
functions have not been changed, only how ipfw is invoked is different.

However there are many changes how ipfw is and its add-on's are handled:

 In general ipfw is now called through the PFIL_HOOKS and most associated
 magic, that was in ip_input() or ip_output() previously, is now done in
 ipfw_check_[in|out]() in the ipfw PFIL handler.

 IPDIVERT is entirely handled within the ipfw PFIL handlers.  A packet to
 be diverted is checked if it is fragmented, if yes, ip_reass() gets in for
 reassembly.  If not, or all fragments arrived and the packet is complete,
 divert_packet is called directly.  For 'tee' no reassembly attempt is made
 and a copy of the packet is sent to the divert socket unmodified.  The
 original packet continues its way through ip_input/output().

 ipfw 'forward' is done via m_tag's.  The ipfw PFIL handlers tag the packet
 with the new destination sockaddr_in.  A check if the new destination is a
 local IP address is made and the m_flags are set appropriately.  ip_input()
 and ip_output() have some more work to do here.  For ip_input() the m_flags
 are checked and a packet for us is directly sent to the 'ours' section for
 further processing.  Destination changes on the input path are only tagged
 and the 'srcrt' flag to ip_forward() is set to disable destination checks
 and ICMP replies at this stage.  The tag is going to be handled on output.
 ip_output() again checks for m_flags and the 'ours' tag.  If found, the
 packet will be dropped back to the IP netisr where it is going to be picked
 up by ip_input() again and the directly sent to the 'ours' section.  When
 only the destination changes, the route's 'dst' is overwritten with the
 new destination from the forward m_tag.  Then it jumps back at the route
 lookup again and skips the firewall check because it has been marked with
 M_SKIP_FIREWALL.  ipfw 'forward' has to be compiled into the kernel with
 'option IPFIREWALL_FORWARD' to enable it.

 DUMMYNET is entirely handled within the ipfw PFIL handlers.  A packet for
 a dummynet pipe or queue is directly sent to dummynet_io().  Dummynet will
 then inject it back into ip_input/ip_output() after it has served its time.
 Dummynet packets are tagged and will continue from the next rule when they
 hit the ipfw PFIL handlers again after re-injection.

 BRIDGING and IPFW_ETHER are not changed yet and use ipfw_chk() directly as
 they did before.  Later this will be changed to dedicated ETHER PFIL_HOOKS.

More detailed changes to the code:

 conf/files
	Add netinet/ip_fw_pfil.c.

 conf/options
	Add IPFIREWALL_FORWARD option.

 modules/ipfw/Makefile
	Add ip_fw_pfil.c.

 net/bridge.c
	Disable PFIL_HOOKS if ipfw for bridging is active.  Bridging ipfw
	is still directly invoked to handle layer2 headers and packets would
	get a double ipfw when run through PFIL_HOOKS as well.

 netinet/ip_divert.c
	Removed divert_clone() function.  It is no longer used.

 netinet/ip_dummynet.[ch]
	Neither the route 'ro' nor the destination 'dst' need to be stored
	while in dummynet transit.  Structure members and associated macros
	are removed.

 netinet/ip_fastfwd.c
	Removed all direct ipfw handling code and replace it with the new
	'ipfw forward' handling code.

 netinet/ip_fw.h
	Removed 'ro' and 'dst' from struct ip_fw_args.

 netinet/ip_fw2.c
	(Re)moved some global variables and the module handling.

 netinet/ip_fw_pfil.c
	New file containing the ipfw PFIL handlers and module initialization.

 netinet/ip_input.c
	Removed all direct ipfw handling code and replace it with the new
	'ipfw forward' handling code.  ip_forward() does not longer require
	the 'next_hop' struct sockaddr_in argument.  Disable early checks
	if 'srcrt' is set.

 netinet/ip_output.c
	Removed all direct ipfw handling code and replace it with the new
	'ipfw forward' handling code.

 netinet/ip_var.h
	Add ip_reass() as general function.  (Used from ipfw PFIL handlers
	for IPDIVERT.)

 netinet/raw_ip.c
	Directly check if ipfw and dummynet control pointers are active.

 netinet/tcp_input.c
	Rework the 'ipfw forward' to local code to work with the new way of
	forward tags.

 netinet/tcp_sack.c
	Remove include 'opt_ipfw.h' which is not needed here.

 sys/mbuf.h
	Remove m_claim_next() macro which was exclusively for ipfw 'forward'
	and is no longer needed.

Approved by:	re (scottl)
2004-08-17 22:05:54 +00:00

1054 lines
29 KiB
C

/*
* ng_bridge.c
*
* Copyright (c) 2000 Whistle Communications, Inc.
* All rights reserved.
*
* Subject to the following obligations and disclaimer of warranty, use and
* redistribution of this software, in source or object code forms, with or
* without modifications are expressly permitted by Whistle Communications;
* provided, however, that:
* 1. Any and all reproductions of the source or object code must include the
* copyright notice above and the following disclaimer of warranties; and
* 2. No rights are granted, in any manner or form, to use Whistle
* Communications, Inc. trademarks, including the mark "WHISTLE
* COMMUNICATIONS" on advertising, endorsements, or otherwise except as
* such appears in the above copyright notice or in the software.
*
* THIS SOFTWARE IS BEING PROVIDED BY WHISTLE COMMUNICATIONS "AS IS", AND
* TO THE MAXIMUM EXTENT PERMITTED BY LAW, WHISTLE COMMUNICATIONS MAKES NO
* REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED, REGARDING THIS SOFTWARE,
* INCLUDING WITHOUT LIMITATION, ANY AND ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT.
* WHISTLE COMMUNICATIONS DOES NOT WARRANT, GUARANTEE, OR MAKE ANY
* REPRESENTATIONS REGARDING THE USE OF, OR THE RESULTS OF THE USE OF THIS
* SOFTWARE IN TERMS OF ITS CORRECTNESS, ACCURACY, RELIABILITY OR OTHERWISE.
* IN NO EVENT SHALL WHISTLE COMMUNICATIONS BE LIABLE FOR ANY DAMAGES
* RESULTING FROM OR ARISING OUT OF ANY USE OF THIS SOFTWARE, INCLUDING
* WITHOUT LIMITATION, ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY,
* PUNITIVE, OR CONSEQUENTIAL DAMAGES, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES, LOSS OF USE, DATA OR PROFITS, HOWEVER CAUSED AND UNDER 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 WHISTLE COMMUNICATIONS IS ADVISED OF THE POSSIBILITY
* OF SUCH DAMAGE.
*
* Author: Archie Cobbs <archie@freebsd.org>
*
* $FreeBSD$
*/
/*
* ng_bridge(4) netgraph node type
*
* The node performs standard intelligent Ethernet bridging over
* each of its connected hooks, or links. A simple loop detection
* algorithm is included which disables a link for priv->conf.loopTimeout
* seconds when a host is seen to have jumped from one link to
* another within priv->conf.minStableAge seconds.
*
* We keep a hashtable that maps Ethernet addresses to host info,
* which is contained in struct ng_bridge_host's. These structures
* tell us on which link the host may be found. A host's entry will
* expire after priv->conf.maxStaleness seconds.
*
* This node is optimzed for stable networks, where machines jump
* from one port to the other only rarely.
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/errno.h>
#include <sys/syslog.h>
#include <sys/socket.h>
#include <sys/ctype.h>
#include <net/if.h>
#include <net/ethernet.h>
#include <netinet/in.h>
#include <netinet/ip_fw.h>
#include <netgraph/ng_message.h>
#include <netgraph/netgraph.h>
#include <netgraph/ng_parse.h>
#include <netgraph/ng_bridge.h>
#ifdef NG_SEPARATE_MALLOC
MALLOC_DEFINE(M_NETGRAPH_BRIDGE, "netgraph_bridge", "netgraph bridge node ");
#else
#define M_NETGRAPH_BRIDGE M_NETGRAPH
#endif
/* Per-link private data */
struct ng_bridge_link {
hook_p hook; /* netgraph hook */
u_int16_t loopCount; /* loop ignore timer */
struct ng_bridge_link_stats stats; /* link stats */
};
/* Per-node private data */
struct ng_bridge_private {
struct ng_bridge_bucket *tab; /* hash table bucket array */
struct ng_bridge_link *links[NG_BRIDGE_MAX_LINKS];
struct ng_bridge_config conf; /* node configuration */
node_p node; /* netgraph node */
u_int numHosts; /* num entries in table */
u_int numBuckets; /* num buckets in table */
u_int hashMask; /* numBuckets - 1 */
int numLinks; /* num connected links */
struct callout timer; /* one second periodic timer */
};
typedef struct ng_bridge_private *priv_p;
/* Information about a host, stored in a hash table entry */
struct ng_bridge_hent {
struct ng_bridge_host host; /* actual host info */
SLIST_ENTRY(ng_bridge_hent) next; /* next entry in bucket */
};
/* Hash table bucket declaration */
SLIST_HEAD(ng_bridge_bucket, ng_bridge_hent);
/* Netgraph node methods */
static ng_constructor_t ng_bridge_constructor;
static ng_rcvmsg_t ng_bridge_rcvmsg;
static ng_shutdown_t ng_bridge_shutdown;
static ng_newhook_t ng_bridge_newhook;
static ng_rcvdata_t ng_bridge_rcvdata;
static ng_disconnect_t ng_bridge_disconnect;
/* Other internal functions */
static struct ng_bridge_host *ng_bridge_get(priv_p priv, const u_char *addr);
static int ng_bridge_put(priv_p priv, const u_char *addr, int linkNum);
static void ng_bridge_rehash(priv_p priv);
static void ng_bridge_remove_hosts(priv_p priv, int linkNum);
static void ng_bridge_timeout(void *arg);
static const char *ng_bridge_nodename(node_p node);
/* Ethernet broadcast */
static const u_char ng_bridge_bcast_addr[ETHER_ADDR_LEN] =
{ 0xff, 0xff, 0xff, 0xff, 0xff, 0xff };
/* Store each hook's link number in the private field */
#define LINK_NUM(hook) (*(u_int16_t *)(&(hook)->private))
/* Compare Ethernet addresses using 32 and 16 bit words instead of bytewise */
#define ETHER_EQUAL(a,b) (((const u_int32_t *)(a))[0] \
== ((const u_int32_t *)(b))[0] \
&& ((const u_int16_t *)(a))[2] \
== ((const u_int16_t *)(b))[2])
/* Minimum and maximum number of hash buckets. Must be a power of two. */
#define MIN_BUCKETS (1 << 5) /* 32 */
#define MAX_BUCKETS (1 << 14) /* 16384 */
/* Configuration default values */
#define DEFAULT_LOOP_TIMEOUT 60
#define DEFAULT_MAX_STALENESS (15 * 60) /* same as ARP timeout */
#define DEFAULT_MIN_STABLE_AGE 1
/******************************************************************
NETGRAPH PARSE TYPES
******************************************************************/
/*
* How to determine the length of the table returned by NGM_BRIDGE_GET_TABLE
*/
static int
ng_bridge_getTableLength(const struct ng_parse_type *type,
const u_char *start, const u_char *buf)
{
const struct ng_bridge_host_ary *const hary
= (const struct ng_bridge_host_ary *)(buf - sizeof(u_int32_t));
return hary->numHosts;
}
/* Parse type for struct ng_bridge_host_ary */
static const struct ng_parse_struct_field ng_bridge_host_type_fields[]
= NG_BRIDGE_HOST_TYPE_INFO(&ng_parse_enaddr_type);
static const struct ng_parse_type ng_bridge_host_type = {
&ng_parse_struct_type,
&ng_bridge_host_type_fields
};
static const struct ng_parse_array_info ng_bridge_hary_type_info = {
&ng_bridge_host_type,
ng_bridge_getTableLength
};
static const struct ng_parse_type ng_bridge_hary_type = {
&ng_parse_array_type,
&ng_bridge_hary_type_info
};
static const struct ng_parse_struct_field ng_bridge_host_ary_type_fields[]
= NG_BRIDGE_HOST_ARY_TYPE_INFO(&ng_bridge_hary_type);
static const struct ng_parse_type ng_bridge_host_ary_type = {
&ng_parse_struct_type,
&ng_bridge_host_ary_type_fields
};
/* Parse type for struct ng_bridge_config */
static const struct ng_parse_fixedarray_info ng_bridge_ipfwary_type_info = {
&ng_parse_uint8_type,
NG_BRIDGE_MAX_LINKS
};
static const struct ng_parse_type ng_bridge_ipfwary_type = {
&ng_parse_fixedarray_type,
&ng_bridge_ipfwary_type_info
};
static const struct ng_parse_struct_field ng_bridge_config_type_fields[]
= NG_BRIDGE_CONFIG_TYPE_INFO(&ng_bridge_ipfwary_type);
static const struct ng_parse_type ng_bridge_config_type = {
&ng_parse_struct_type,
&ng_bridge_config_type_fields
};
/* Parse type for struct ng_bridge_link_stat */
static const struct ng_parse_struct_field ng_bridge_stats_type_fields[]
= NG_BRIDGE_STATS_TYPE_INFO;
static const struct ng_parse_type ng_bridge_stats_type = {
&ng_parse_struct_type,
&ng_bridge_stats_type_fields
};
/* List of commands and how to convert arguments to/from ASCII */
static const struct ng_cmdlist ng_bridge_cmdlist[] = {
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_SET_CONFIG,
"setconfig",
&ng_bridge_config_type,
NULL
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_GET_CONFIG,
"getconfig",
NULL,
&ng_bridge_config_type
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_RESET,
"reset",
NULL,
NULL
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_GET_STATS,
"getstats",
&ng_parse_uint32_type,
&ng_bridge_stats_type
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_CLR_STATS,
"clrstats",
&ng_parse_uint32_type,
NULL
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_GETCLR_STATS,
"getclrstats",
&ng_parse_uint32_type,
&ng_bridge_stats_type
},
{
NGM_BRIDGE_COOKIE,
NGM_BRIDGE_GET_TABLE,
"gettable",
NULL,
&ng_bridge_host_ary_type
},
{ 0 }
};
/* Node type descriptor */
static struct ng_type ng_bridge_typestruct = {
.version = NG_ABI_VERSION,
.name = NG_BRIDGE_NODE_TYPE,
.constructor = ng_bridge_constructor,
.rcvmsg = ng_bridge_rcvmsg,
.shutdown = ng_bridge_shutdown,
.newhook = ng_bridge_newhook,
.rcvdata = ng_bridge_rcvdata,
.disconnect = ng_bridge_disconnect,
.cmdlist = ng_bridge_cmdlist,
};
NETGRAPH_INIT(bridge, &ng_bridge_typestruct);
/******************************************************************
NETGRAPH NODE METHODS
******************************************************************/
/*
* Node constructor
*/
static int
ng_bridge_constructor(node_p node)
{
priv_p priv;
/* Allocate and initialize private info */
MALLOC(priv, priv_p, sizeof(*priv), M_NETGRAPH_BRIDGE, M_NOWAIT | M_ZERO);
if (priv == NULL)
return (ENOMEM);
callout_init(&priv->timer, 0);
/* Allocate and initialize hash table, etc. */
MALLOC(priv->tab, struct ng_bridge_bucket *,
MIN_BUCKETS * sizeof(*priv->tab), M_NETGRAPH_BRIDGE, M_NOWAIT | M_ZERO);
if (priv->tab == NULL) {
FREE(priv, M_NETGRAPH_BRIDGE);
return (ENOMEM);
}
priv->numBuckets = MIN_BUCKETS;
priv->hashMask = MIN_BUCKETS - 1;
priv->conf.debugLevel = 1;
priv->conf.loopTimeout = DEFAULT_LOOP_TIMEOUT;
priv->conf.maxStaleness = DEFAULT_MAX_STALENESS;
priv->conf.minStableAge = DEFAULT_MIN_STABLE_AGE;
/*
* This node has all kinds of stuff that could be screwed by SMP.
* Until it gets it's own internal protection, we go through in
* single file. This could hurt a machine bridging beteen two
* GB ethernets so it should be fixed.
* When it's fixed the process SHOULD NOT SLEEP, spinlocks please!
* (and atomic ops )
*/
NG_NODE_FORCE_WRITER(node);
NG_NODE_SET_PRIVATE(node, priv);
priv->node = node;
/* Start timer; timer is always running while node is alive */
callout_reset(&priv->timer, hz, ng_bridge_timeout, priv->node);
/* Done */
return (0);
}
/*
* Method for attaching a new hook
*/
static int
ng_bridge_newhook(node_p node, hook_p hook, const char *name)
{
const priv_p priv = NG_NODE_PRIVATE(node);
/* Check for a link hook */
if (strncmp(name, NG_BRIDGE_HOOK_LINK_PREFIX,
strlen(NG_BRIDGE_HOOK_LINK_PREFIX)) == 0) {
const char *cp;
char *eptr;
u_long linkNum;
cp = name + strlen(NG_BRIDGE_HOOK_LINK_PREFIX);
if (!isdigit(*cp) || (cp[0] == '0' && cp[1] != '\0'))
return (EINVAL);
linkNum = strtoul(cp, &eptr, 10);
if (*eptr != '\0' || linkNum >= NG_BRIDGE_MAX_LINKS)
return (EINVAL);
if (priv->links[linkNum] != NULL)
return (EISCONN);
MALLOC(priv->links[linkNum], struct ng_bridge_link *,
sizeof(*priv->links[linkNum]), M_NETGRAPH_BRIDGE, M_NOWAIT|M_ZERO);
if (priv->links[linkNum] == NULL)
return (ENOMEM);
priv->links[linkNum]->hook = hook;
NG_HOOK_SET_PRIVATE(hook, (void *)linkNum);
priv->numLinks++;
return (0);
}
/* Unknown hook name */
return (EINVAL);
}
/*
* Receive a control message
*/
static int
ng_bridge_rcvmsg(node_p node, item_p item, hook_p lasthook)
{
const priv_p priv = NG_NODE_PRIVATE(node);
struct ng_mesg *resp = NULL;
int error = 0;
struct ng_mesg *msg;
NGI_GET_MSG(item, msg);
switch (msg->header.typecookie) {
case NGM_BRIDGE_COOKIE:
switch (msg->header.cmd) {
case NGM_BRIDGE_GET_CONFIG:
{
struct ng_bridge_config *conf;
NG_MKRESPONSE(resp, msg,
sizeof(struct ng_bridge_config), M_NOWAIT);
if (resp == NULL) {
error = ENOMEM;
break;
}
conf = (struct ng_bridge_config *)resp->data;
*conf = priv->conf; /* no sanity checking needed */
break;
}
case NGM_BRIDGE_SET_CONFIG:
{
struct ng_bridge_config *conf;
int i;
if (msg->header.arglen
!= sizeof(struct ng_bridge_config)) {
error = EINVAL;
break;
}
conf = (struct ng_bridge_config *)msg->data;
priv->conf = *conf;
for (i = 0; i < NG_BRIDGE_MAX_LINKS; i++)
priv->conf.ipfw[i] = !!priv->conf.ipfw[i];
break;
}
case NGM_BRIDGE_RESET:
{
int i;
/* Flush all entries in the hash table */
ng_bridge_remove_hosts(priv, -1);
/* Reset all loop detection counters and stats */
for (i = 0; i < NG_BRIDGE_MAX_LINKS; i++) {
if (priv->links[i] == NULL)
continue;
priv->links[i]->loopCount = 0;
bzero(&priv->links[i]->stats,
sizeof(priv->links[i]->stats));
}
break;
}
case NGM_BRIDGE_GET_STATS:
case NGM_BRIDGE_CLR_STATS:
case NGM_BRIDGE_GETCLR_STATS:
{
struct ng_bridge_link *link;
int linkNum;
/* Get link number */
if (msg->header.arglen != sizeof(u_int32_t)) {
error = EINVAL;
break;
}
linkNum = *((u_int32_t *)msg->data);
if (linkNum < 0 || linkNum >= NG_BRIDGE_MAX_LINKS) {
error = EINVAL;
break;
}
if ((link = priv->links[linkNum]) == NULL) {
error = ENOTCONN;
break;
}
/* Get/clear stats */
if (msg->header.cmd != NGM_BRIDGE_CLR_STATS) {
NG_MKRESPONSE(resp, msg,
sizeof(link->stats), M_NOWAIT);
if (resp == NULL) {
error = ENOMEM;
break;
}
bcopy(&link->stats,
resp->data, sizeof(link->stats));
}
if (msg->header.cmd != NGM_BRIDGE_GET_STATS)
bzero(&link->stats, sizeof(link->stats));
break;
}
case NGM_BRIDGE_GET_TABLE:
{
struct ng_bridge_host_ary *ary;
struct ng_bridge_hent *hent;
int i = 0, bucket;
NG_MKRESPONSE(resp, msg, sizeof(*ary)
+ (priv->numHosts * sizeof(*ary->hosts)), M_NOWAIT);
if (resp == NULL) {
error = ENOMEM;
break;
}
ary = (struct ng_bridge_host_ary *)resp->data;
ary->numHosts = priv->numHosts;
for (bucket = 0; bucket < priv->numBuckets; bucket++) {
SLIST_FOREACH(hent, &priv->tab[bucket], next)
ary->hosts[i++] = hent->host;
}
break;
}
default:
error = EINVAL;
break;
}
break;
default:
error = EINVAL;
break;
}
/* Done */
NG_RESPOND_MSG(error, node, item, resp);
NG_FREE_MSG(msg);
return (error);
}
/*
* Receive data on a hook
*/
static int
ng_bridge_rcvdata(hook_p hook, item_p item)
{
const node_p node = NG_HOOK_NODE(hook);
const priv_p priv = NG_NODE_PRIVATE(node);
struct ng_bridge_host *host;
struct ng_bridge_link *link;
struct ether_header *eh;
int error = 0, linkNum, linksSeen;
int manycast;
struct mbuf *m;
struct ng_bridge_link *firstLink;
NGI_GET_M(item, m);
/* Get link number */
linkNum = (intptr_t)NG_HOOK_PRIVATE(hook);
KASSERT(linkNum >= 0 && linkNum < NG_BRIDGE_MAX_LINKS,
("%s: linkNum=%u", __func__, linkNum));
link = priv->links[linkNum];
KASSERT(link != NULL, ("%s: link%d null", __func__, linkNum));
/* Sanity check packet and pull up header */
if (m->m_pkthdr.len < ETHER_HDR_LEN) {
link->stats.recvRunts++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (EINVAL);
}
if (m->m_len < ETHER_HDR_LEN && !(m = m_pullup(m, ETHER_HDR_LEN))) {
link->stats.memoryFailures++;
NG_FREE_ITEM(item);
return (ENOBUFS);
}
eh = mtod(m, struct ether_header *);
if ((eh->ether_shost[0] & 1) != 0) {
link->stats.recvInvalid++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (EINVAL);
}
/* Is link disabled due to a loopback condition? */
if (link->loopCount != 0) {
link->stats.loopDrops++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (ELOOP); /* XXX is this an appropriate error? */
}
/* Update stats */
link->stats.recvPackets++;
link->stats.recvOctets += m->m_pkthdr.len;
if ((manycast = (eh->ether_dhost[0] & 1)) != 0) {
if (ETHER_EQUAL(eh->ether_dhost, ng_bridge_bcast_addr)) {
link->stats.recvBroadcasts++;
manycast = 2;
} else
link->stats.recvMulticasts++;
}
/* Look up packet's source Ethernet address in hashtable */
if ((host = ng_bridge_get(priv, eh->ether_shost)) != NULL) {
/* Update time since last heard from this host */
host->staleness = 0;
/* Did host jump to a different link? */
if (host->linkNum != linkNum) {
/*
* If the host's old link was recently established
* on the old link and it's already jumped to a new
* link, declare a loopback condition.
*/
if (host->age < priv->conf.minStableAge) {
/* Log the problem */
if (priv->conf.debugLevel >= 2) {
struct ifnet *ifp = m->m_pkthdr.rcvif;
char suffix[32];
if (ifp != NULL)
snprintf(suffix, sizeof(suffix),
" (%s)", ifp->if_xname);
else
*suffix = '\0';
log(LOG_WARNING, "ng_bridge: %s:"
" loopback detected on %s%s\n",
ng_bridge_nodename(node),
NG_HOOK_NAME(hook), suffix);
}
/* Mark link as linka non grata */
link->loopCount = priv->conf.loopTimeout;
link->stats.loopDetects++;
/* Forget all hosts on this link */
ng_bridge_remove_hosts(priv, linkNum);
/* Drop packet */
link->stats.loopDrops++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (ELOOP); /* XXX appropriate? */
}
/* Move host over to new link */
host->linkNum = linkNum;
host->age = 0;
}
} else {
if (!ng_bridge_put(priv, eh->ether_shost, linkNum)) {
link->stats.memoryFailures++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (ENOMEM);
}
}
/* Run packet through ipfw processing, if enabled */
#if 0
if (priv->conf.ipfw[linkNum] && fw_enable && ip_fw_chk_ptr != NULL) {
/* XXX not implemented yet */
}
#endif
/*
* If unicast and destination host known, deliver to host's link,
* unless it is the same link as the packet came in on.
*/
if (!manycast) {
/* Determine packet destination link */
if ((host = ng_bridge_get(priv, eh->ether_dhost)) != NULL) {
struct ng_bridge_link *const destLink
= priv->links[host->linkNum];
/* If destination same as incoming link, do nothing */
KASSERT(destLink != NULL,
("%s: link%d null", __func__, host->linkNum));
if (destLink == link) {
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (0);
}
/* Deliver packet out the destination link */
destLink->stats.xmitPackets++;
destLink->stats.xmitOctets += m->m_pkthdr.len;
NG_FWD_NEW_DATA(error, item, destLink->hook, m);
return (error);
}
/* Destination host is not known */
link->stats.recvUnknown++;
}
/* Distribute unknown, multicast, broadcast pkts to all other links */
firstLink = NULL;
for (linkNum = linksSeen = 0; linksSeen <= priv->numLinks; linkNum++) {
struct ng_bridge_link *destLink;
struct mbuf *m2 = NULL;
/*
* If we have checked all the links then now
* send the original on its reserved link
*/
if (linksSeen == priv->numLinks) {
/* If we never saw a good link, leave. */
if (firstLink == NULL) {
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (0);
}
destLink = firstLink;
} else {
destLink = priv->links[linkNum];
if (destLink != NULL)
linksSeen++;
/* Skip incoming link and disconnected links */
if (destLink == NULL || destLink == link) {
continue;
}
if (firstLink == NULL) {
/*
* This is the first usable link we have found.
* Reserve it for the originals.
* If we never find another we save a copy.
*/
firstLink = destLink;
continue;
}
/*
* It's usable link but not the reserved (first) one.
* Copy mbuf info for sending.
*/
m2 = m_dup(m, M_DONTWAIT); /* XXX m_copypacket() */
if (m2 == NULL) {
link->stats.memoryFailures++;
NG_FREE_ITEM(item);
NG_FREE_M(m);
return (ENOBUFS);
}
}
/* Update stats */
destLink->stats.xmitPackets++;
destLink->stats.xmitOctets += m->m_pkthdr.len;
switch (manycast) {
case 0: /* unicast */
break;
case 1: /* multicast */
destLink->stats.xmitMulticasts++;
break;
case 2: /* broadcast */
destLink->stats.xmitBroadcasts++;
break;
}
/* Send packet */
if (destLink == firstLink) {
/*
* If we've sent all the others, send the original
* on the first link we found.
*/
NG_FWD_NEW_DATA(error, item, destLink->hook, m);
break; /* always done last - not really needed. */
} else {
NG_SEND_DATA_ONLY(error, destLink->hook, m2);
}
}
return (error);
}
/*
* Shutdown node
*/
static int
ng_bridge_shutdown(node_p node)
{
const priv_p priv = NG_NODE_PRIVATE(node);
/*
* Shut down everything except the timer. There's no way to
* avoid another possible timeout event (it may have already
* been dequeued), so we can't free the node yet.
*/
KASSERT(priv->numLinks == 0 && priv->numHosts == 0,
("%s: numLinks=%d numHosts=%d",
__func__, priv->numLinks, priv->numHosts));
FREE(priv->tab, M_NETGRAPH_BRIDGE);
/* NGF_INVALID flag is now set so node will be freed at next timeout */
return (0);
}
/*
* Hook disconnection.
*/
static int
ng_bridge_disconnect(hook_p hook)
{
const priv_p priv = NG_NODE_PRIVATE(NG_HOOK_NODE(hook));
int linkNum;
/* Get link number */
linkNum = (intptr_t)NG_HOOK_PRIVATE(hook);
KASSERT(linkNum >= 0 && linkNum < NG_BRIDGE_MAX_LINKS,
("%s: linkNum=%u", __func__, linkNum));
/* Remove all hosts associated with this link */
ng_bridge_remove_hosts(priv, linkNum);
/* Free associated link information */
KASSERT(priv->links[linkNum] != NULL, ("%s: no link", __func__));
FREE(priv->links[linkNum], M_NETGRAPH_BRIDGE);
priv->links[linkNum] = NULL;
priv->numLinks--;
/* If no more hooks, go away */
if ((NG_NODE_NUMHOOKS(NG_HOOK_NODE(hook)) == 0)
&& (NG_NODE_IS_VALID(NG_HOOK_NODE(hook)))) {
ng_rmnode_self(NG_HOOK_NODE(hook));
}
return (0);
}
/******************************************************************
HASH TABLE FUNCTIONS
******************************************************************/
/*
* Hash algorithm
*/
#define HASH(addr,mask) ( (((const u_int16_t *)(addr))[0] \
^ ((const u_int16_t *)(addr))[1] \
^ ((const u_int16_t *)(addr))[2]) & (mask) )
/*
* Find a host entry in the table.
*/
static struct ng_bridge_host *
ng_bridge_get(priv_p priv, const u_char *addr)
{
const int bucket = HASH(addr, priv->hashMask);
struct ng_bridge_hent *hent;
SLIST_FOREACH(hent, &priv->tab[bucket], next) {
if (ETHER_EQUAL(hent->host.addr, addr))
return (&hent->host);
}
return (NULL);
}
/*
* Add a new host entry to the table. This assumes the host doesn't
* already exist in the table. Returns 1 on success, 0 if there
* was a memory allocation failure.
*/
static int
ng_bridge_put(priv_p priv, const u_char *addr, int linkNum)
{
const int bucket = HASH(addr, priv->hashMask);
struct ng_bridge_hent *hent;
#ifdef INVARIANTS
/* Assert that entry does not already exist in hashtable */
SLIST_FOREACH(hent, &priv->tab[bucket], next) {
KASSERT(!ETHER_EQUAL(hent->host.addr, addr),
("%s: entry %6D exists in table", __func__, addr, ":"));
}
#endif
/* Allocate and initialize new hashtable entry */
MALLOC(hent, struct ng_bridge_hent *,
sizeof(*hent), M_NETGRAPH_BRIDGE, M_NOWAIT);
if (hent == NULL)
return (0);
bcopy(addr, hent->host.addr, ETHER_ADDR_LEN);
hent->host.linkNum = linkNum;
hent->host.staleness = 0;
hent->host.age = 0;
/* Add new element to hash bucket */
SLIST_INSERT_HEAD(&priv->tab[bucket], hent, next);
priv->numHosts++;
/* Resize table if necessary */
ng_bridge_rehash(priv);
return (1);
}
/*
* Resize the hash table. We try to maintain the number of buckets
* such that the load factor is in the range 0.25 to 1.0.
*
* If we can't get the new memory then we silently fail. This is OK
* because things will still work and we'll try again soon anyway.
*/
static void
ng_bridge_rehash(priv_p priv)
{
struct ng_bridge_bucket *newTab;
int oldBucket, newBucket;
int newNumBuckets;
u_int newMask;
/* Is table too full or too empty? */
if (priv->numHosts > priv->numBuckets
&& (priv->numBuckets << 1) <= MAX_BUCKETS)
newNumBuckets = priv->numBuckets << 1;
else if (priv->numHosts < (priv->numBuckets >> 2)
&& (priv->numBuckets >> 2) >= MIN_BUCKETS)
newNumBuckets = priv->numBuckets >> 2;
else
return;
newMask = newNumBuckets - 1;
/* Allocate and initialize new table */
MALLOC(newTab, struct ng_bridge_bucket *,
newNumBuckets * sizeof(*newTab), M_NETGRAPH_BRIDGE, M_NOWAIT | M_ZERO);
if (newTab == NULL)
return;
/* Move all entries from old table to new table */
for (oldBucket = 0; oldBucket < priv->numBuckets; oldBucket++) {
struct ng_bridge_bucket *const oldList = &priv->tab[oldBucket];
while (!SLIST_EMPTY(oldList)) {
struct ng_bridge_hent *const hent
= SLIST_FIRST(oldList);
SLIST_REMOVE_HEAD(oldList, next);
newBucket = HASH(hent->host.addr, newMask);
SLIST_INSERT_HEAD(&newTab[newBucket], hent, next);
}
}
/* Replace old table with new one */
if (priv->conf.debugLevel >= 3) {
log(LOG_INFO, "ng_bridge: %s: table size %d -> %d\n",
ng_bridge_nodename(priv->node),
priv->numBuckets, newNumBuckets);
}
FREE(priv->tab, M_NETGRAPH_BRIDGE);
priv->numBuckets = newNumBuckets;
priv->hashMask = newMask;
priv->tab = newTab;
return;
}
/******************************************************************
MISC FUNCTIONS
******************************************************************/
/*
* Remove all hosts associated with a specific link from the hashtable.
* If linkNum == -1, then remove all hosts in the table.
*/
static void
ng_bridge_remove_hosts(priv_p priv, int linkNum)
{
int bucket;
for (bucket = 0; bucket < priv->numBuckets; bucket++) {
struct ng_bridge_hent **hptr = &SLIST_FIRST(&priv->tab[bucket]);
while (*hptr != NULL) {
struct ng_bridge_hent *const hent = *hptr;
if (linkNum == -1 || hent->host.linkNum == linkNum) {
*hptr = SLIST_NEXT(hent, next);
FREE(hent, M_NETGRAPH_BRIDGE);
priv->numHosts--;
} else
hptr = &SLIST_NEXT(hent, next);
}
}
}
/*
* Handle our once-per-second timeout event. We do two things:
* we decrement link->loopCount for those links being muted due to
* a detected loopback condition, and we remove any hosts from
* the hashtable whom we haven't heard from in a long while.
*
* If the node has the NGF_INVALID flag set, our job is to kill it.
*/
static void
ng_bridge_timeout(void *arg)
{
const node_p node = arg;
const priv_p priv = NG_NODE_PRIVATE(node);
int s, bucket;
int counter = 0;
int linkNum;
/* If node was shut down, this is the final lingering timeout */
s = splnet();
if (NG_NODE_NOT_VALID(node)) {
FREE(priv, M_NETGRAPH_BRIDGE);
NG_NODE_SET_PRIVATE(node, NULL);
NG_NODE_UNREF(node);
splx(s);
return;
}
/* Register a new timeout, keeping the existing node reference */
callout_reset(&priv->timer, hz, ng_bridge_timeout, node);
/* Update host time counters and remove stale entries */
for (bucket = 0; bucket < priv->numBuckets; bucket++) {
struct ng_bridge_hent **hptr = &SLIST_FIRST(&priv->tab[bucket]);
while (*hptr != NULL) {
struct ng_bridge_hent *const hent = *hptr;
/* Make sure host's link really exists */
KASSERT(priv->links[hent->host.linkNum] != NULL,
("%s: host %6D on nonexistent link %d\n",
__func__, hent->host.addr, ":",
hent->host.linkNum));
/* Remove hosts we haven't heard from in a while */
if (++hent->host.staleness >= priv->conf.maxStaleness) {
*hptr = SLIST_NEXT(hent, next);
FREE(hent, M_NETGRAPH_BRIDGE);
priv->numHosts--;
} else {
if (hent->host.age < 0xffff)
hent->host.age++;
hptr = &SLIST_NEXT(hent, next);
counter++;
}
}
}
KASSERT(priv->numHosts == counter,
("%s: hosts: %d != %d", __func__, priv->numHosts, counter));
/* Decrease table size if necessary */
ng_bridge_rehash(priv);
/* Decrease loop counter on muted looped back links */
for (counter = linkNum = 0; linkNum < NG_BRIDGE_MAX_LINKS; linkNum++) {
struct ng_bridge_link *const link = priv->links[linkNum];
if (link != NULL) {
if (link->loopCount != 0) {
link->loopCount--;
if (link->loopCount == 0
&& priv->conf.debugLevel >= 2) {
log(LOG_INFO, "ng_bridge: %s:"
" restoring looped back link%d\n",
ng_bridge_nodename(node), linkNum);
}
}
counter++;
}
}
KASSERT(priv->numLinks == counter,
("%s: links: %d != %d", __func__, priv->numLinks, counter));
/* Done */
splx(s);
}
/*
* Return node's "name", even if it doesn't have one.
*/
static const char *
ng_bridge_nodename(node_p node)
{
static char name[NG_NODESIZ];
if (NG_NODE_NAME(node) != NULL)
snprintf(name, sizeof(name), "%s", NG_NODE_NAME(node));
else
snprintf(name, sizeof(name), "[%x]", ng_node2ID(node));
return name;
}