funny ECMP

Ingo Flaschberger if at xip.at
Mon Aug 23 23:55:24 UTC 2010


Hi,

I have changed the route selection code of ecmp to
balance only between routes of the same weight.
(see attached file)

As Qing Li mentioned months ago, there are problems with static routes and 
interfaces.

Example 1:
ifconfig em1 alias 10.13.13.90/24
 	ping 10.13.13.95 ok
route add 10.13.13.0/24 10.11.11.1 -weight 2
 	ping 10.13.13.95 broken (arpresolve: can't allocate llinfo for
 				10.13.13.95)

Eaxmple 2:
route add 10.13.13.0/24 10.11.11.1 -weight 2
 	ping 10.13.13.95 ok
route add 10.13.13.0/24 10.11.11.2 -weight 3
 	ping 10.13.13.95 ok
ifconfig em1 alias 10.13.13.90/24
 	ping 10.13.13.95 broken (arpresolve: can't allocate llinfo for
 				10.13.13.95)
route delete 10.13.13.0/24 10.11.11.2
 	ping 10.13.13.95 broken
route delete 10.13.13.0/24 10.11.11.1
 	ping -> system crash

deleting the last static multipath routes seems to affect?, destroy?
the interface route.
system crashes in:
/usr/src/sys/net/route.c:370
         rn = rnh->rnh_matchaddr(dst, rnh);
         if (rn && ((rn->rn_flags & RNF_ROOT) == 0)) {
                 newrt = rt = RNTORT(rn);
here ->         RT_LOCK(newrt);
                 RT_ADDREF(newrt);
                 if (needlock)

Kind regards,
 	Ingo Flaschberger
-------------- next part --------------
/*
 * Copyright (C) 2001 WIDE Project.
 * 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 project 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 PROJECT 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 PROJECT 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.
 * THE AUTHORS DO NOT GUARANTEE THAT THIS SOFTWARE DOES NOT INFRINGE
 * ANY OTHERS' INTELLECTUAL PROPERTIES. IN NO EVENT SHALL THE AUTHORS
 * BE LIABLE FOR ANY INFRINGEMENT OF ANY OTHERS' INTELLECTUAL
 * PROPERTIES.
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD: src/sys/net/radix_mpath.c,v 1.10.2.2 2010/04/02 05:02:50 qingli Exp $");

#include "opt_inet.h"
#include "opt_inet6.h"

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/malloc.h>
#include <sys/socket.h>
#include <sys/domain.h>
#include <sys/syslog.h>
#include <net/radix.h>
#include <net/radix_mpath.h>
#include <net/route.h>
#include <net/if.h>
#include <net/if_var.h>

/*
 * give some jitter to hash, to avoid synchronization between routers
 */
static uint32_t hashjitter;

int
rn_mpath_capable(struct radix_node_head *rnh)
{

	return rnh->rnh_multipath;
}

struct radix_node *
rn_mpath_next(struct radix_node *rn)
{
	struct radix_node *next;

	if (!rn->rn_dupedkey)
		return NULL;
	next = rn->rn_dupedkey;
	if (rn->rn_mask == next->rn_mask)
		return next;
	else
		return NULL;
}

uint32_t
rn_mpath_count(struct radix_node *rn)
{
	uint32_t i = 0;
	struct rtentry *rt;
	
	while (rn != NULL) {
		rt = (struct rtentry *)rn;
		i += rt->rt_rmx.rmx_weight;
		rn = rn_mpath_next(rn);
	}
	return (i);
}

struct rtentry *
rt_mpath_matchgate(struct rtentry *rt, struct sockaddr *gate)
{
	struct radix_node *rn;

	if (!rn_mpath_next((struct radix_node *)rt)) {
		printf("rt_mpath_matchgate: no next\n");
		return rt;
	}

	if (!gate) {
		printf("rt_mpath_matchgate: null\n");
		return NULL;
	}

	/* beyond here, we use rn as the master copy */
	rn = (struct radix_node *)rt;
	do {
		rt = (struct rtentry *)rn;
		/*
		 * we are removing an address alias that has 
		 * the same prefix as another address
		 * we need to compare the interface address because
		 * rt_gateway is a special sockadd_dl structure
		 */
		if (rt->rt_gateway->sa_family == AF_LINK) {
			printf("rt_mpath_matchgate: do1\n");
			if (!memcmp(rt->rt_ifa->ifa_addr, gate, gate->sa_len)) {
				printf("rt_mpath_matchgate: do2\n");
				break;
			}
		} else {
			printf("rt_mpath_matchgate: do3\n");
			if (rt->rt_gateway->sa_len == gate->sa_len &&
			    !memcmp(rt->rt_gateway, gate, gate->sa_len)) {
			    	printf("rt_mpath_matchgate: do4\n");
				break;
			}
		}
	} while ((rn = rn_mpath_next(rn)) != NULL);

	printf("rt_mpath_matchgate: return\n");
	return (struct rtentry *)rn;
}

/* 
 * go through the chain and unlink "rt" from the list
 * the caller will free "rt"
 */
int
rt_mpath_deldup(struct rtentry *headrt, struct rtentry *rt)
{
        struct radix_node *t, *tt;

        if (!headrt || !rt)
            return (0);
        t = (struct radix_node *)headrt;
        tt = rn_mpath_next(t);
        while (tt) {
            if (tt == (struct radix_node *)rt) {
                t->rn_dupedkey = tt->rn_dupedkey;
                tt->rn_dupedkey = NULL;
    	        tt->rn_flags &= ~RNF_ACTIVE;
	        tt[1].rn_flags &= ~RNF_ACTIVE;
                return (1);
            }
            t = tt;
            tt = rn_mpath_next((struct radix_node *)t);
        }
        return (0);
}

/*
 * check if we have the same key/mask/gateway on the table already.
 */
int
rt_mpath_conflict(struct radix_node_head *rnh, struct rtentry *rt,
    struct sockaddr *netmask)
{
	struct radix_node *rn, *rn1;
	struct rtentry *rt1;
	char *p, *q, *eq;
	int same, l, skip;

	rn = (struct radix_node *)rt;
	rn1 = rnh->rnh_lookup(rt_key(rt), netmask, rnh);
	if (!rn1 || rn1->rn_flags & RNF_ROOT)
		return 0;

	/*
	 * unlike other functions we have in this file, we have to check
	 * all key/mask/gateway as rnh_lookup can match less specific entry.
	 */
	rt1 = (struct rtentry *)rn1;

	/* compare key. */
	if (rt_key(rt1)->sa_len != rt_key(rt)->sa_len ||
	    bcmp(rt_key(rt1), rt_key(rt), rt_key(rt1)->sa_len))
		goto different;

	/* key was the same.  compare netmask.  hairy... */
	if (rt_mask(rt1) && netmask) {
		skip = rnh->rnh_treetop->rn_offset;
		if (rt_mask(rt1)->sa_len > netmask->sa_len) {
			/*
			 * as rt_mask(rt1) is made optimal by radix.c,
			 * there must be some 1-bits on rt_mask(rt1)
			 * after netmask->sa_len.  therefore, in
			 * this case, the entries are different.
			 */
			if (rt_mask(rt1)->sa_len > skip)
				goto different;
			else {
				/* no bits to compare, i.e. same*/
				goto maskmatched;
			}
		}

		l = rt_mask(rt1)->sa_len;
		if (skip > l) {
			/* no bits to compare, i.e. same */
			goto maskmatched;
		}
		p = (char *)rt_mask(rt1);
		q = (char *)netmask;
		if (bcmp(p + skip, q + skip, l - skip))
			goto different;
		/*
		 * need to go through all the bit, as netmask is not
		 * optimal and can contain trailing 0s
		 */
		eq = (char *)netmask + netmask->sa_len;
		q += l;
		same = 1;
		while (eq > q)
			if (*q++) {
				same = 0;
				break;
			}
		if (!same)
			goto different;
	} else if (!rt_mask(rt1) && !netmask)
		; /* no mask to compare, i.e. same */
	else {
		/* one has mask and the other does not, different */
		goto different;
	}

maskmatched:

	/* key/mask were the same.  compare gateway for all multipaths */
	do {
		rt1 = (struct rtentry *)rn1;

		/* sanity: no use in comparing the same thing */
		if (rn1 == rn)
			continue;
        
		if (rt1->rt_gateway->sa_family == AF_LINK) {
			if (rt1->rt_ifa->ifa_addr->sa_len != rt->rt_ifa->ifa_addr->sa_len ||
			    bcmp(rt1->rt_ifa->ifa_addr, rt->rt_ifa->ifa_addr, 
			    rt1->rt_ifa->ifa_addr->sa_len))
				continue;
		} else {
			if (rt1->rt_gateway->sa_len != rt->rt_gateway->sa_len ||
			    bcmp(rt1->rt_gateway, rt->rt_gateway,
			    rt1->rt_gateway->sa_len))
				continue;
		}

		/* all key/mask/gateway are the same.  conflicting entry. */
		return EEXIST;
	} while ((rn1 = rn_mpath_next(rn1)) != NULL);

different:
	return 0;
}

void
rtalloc_mpath_fib(struct route *ro, uint32_t hash, u_int fibnum)
{
	struct radix_node *rn0, *rn;
	//u_int32_t n;
	struct rtentry *rt;
	int64_t weight;
	int64_t lowest_weight;
	u_int32_t count;

	/*
	 * XXX we don't attempt to lookup cached route again; what should
	 * be done for sendto(3) case?
	 */
	if (ro->ro_rt && ro->ro_rt->rt_ifp && (ro->ro_rt->rt_flags & RTF_UP)
	    && RT_LINK_IS_UP(ro->ro_rt->rt_ifp)) {
	    	printf("cached route\n");
		return;
	}	 
	ro->ro_rt = rtalloc1_fib(&ro->ro_dst, 1, 0, fibnum);

	/* if the route does not exist or it is not multipath, don't care */
	if (ro->ro_rt == NULL) {
		printf("does not exist\n");
		return;
	}
	if (rn_mpath_next((struct radix_node *)ro->ro_rt) == NULL) {
		RT_UNLOCK(ro->ro_rt);
		//printf("no multipath\n");
		return;
	}
	
	// check if this a direct connected interface or host route / no gateway route
	/*
	if ( (ro->ro_rt->rt_flags & RTF_GATEWAY) == 0) {
		return;
	}*/

	/* beyond here, we use rn as the master copy */
	rn0 = rn = (struct radix_node *)ro->ro_rt;
	//n = rn_mpath_count(rn0);
	
	/* find lowest weight route */
	count = 0;
	for ( rt = (struct rtentry *)rn, weight = rt->rt_rmx.rmx_weight; rn != NULL; rn = rn_mpath_next( rn)) {
		/* XXX check if route is up? */
		rt = (struct rtentry *)rn;
		printf("weight: %ld\n", weight);
		if (weight > rt->rt_rmx.rmx_weight) {
			weight = rt->rt_rmx.rmx_weight;
			count = 1;
			printf("weight: new\n");
		} else if (weight == rt->rt_rmx.rmx_weight) {
			count++;
			printf("weight: count %d\n", count);
		}
	}
	lowest_weight = weight;
	
	printf("lowest weight: %ld\n", weight);
	
	// select now one of the lowest weight routes
	hash += hashjitter;
	hash %= count;
	rn = rn0;
	printf("count: %d\n", count);
	printf("hash: %d\n", hash);
	for ( count = 0; rn != NULL; rn = rn_mpath_next( rn)) {
		rt = (struct rtentry *)rn;
		if ( rt->rt_rmx.rmx_weight == lowest_weight) {
			if (count == hash) {
				printf("weight: count == hash\n");
				break;
			}
			count++;
		}
	}
	printf("selcted weight: %ld\n", rt->rt_rmx.rmx_weight);

	
	/* gw selection has failed - there must be only zero weight routes */
	if (!rn) {
		RT_UNLOCK(ro->ro_rt);
		ro->ro_rt = NULL;
		printf("gateway selection failed\n");
		return;
	}
	if (ro->ro_rt != rt) {
		RTFREE_LOCKED(ro->ro_rt);
		ro->ro_rt = (struct rtentry *)rn;
		RT_LOCK(ro->ro_rt);
		RT_ADDREF(ro->ro_rt);

	} 
	RT_UNLOCK(ro->ro_rt);
}

extern int	in6_inithead(void **head, int off);
extern int	in_inithead(void **head, int off);

#ifdef INET
int
rn4_mpath_inithead(void **head, int off)
{
	struct radix_node_head *rnh;

	hashjitter = arc4random();
	if (in_inithead(head, off) == 1) {
		rnh = (struct radix_node_head *)*head;
		rnh->rnh_multipath = 1;
		return 1;
	} else
		return 0;
}
#endif

#ifdef INET6
int
rn6_mpath_inithead(void **head, int off)
{
	struct radix_node_head *rnh;

	hashjitter = arc4random();
	if (in6_inithead(head, off) == 1) {
		rnh = (struct radix_node_head *)*head;
		rnh->rnh_multipath = 1;
		return 1;
	} else
		return 0;
}

#endif


More information about the freebsd-net mailing list