RIP now actually talks to itself.

This commit is contained in:
Pavel Machek 1998-10-20 16:12:43 +00:00
parent 8ca8683c70
commit feb6abe009
3 changed files with 65 additions and 20 deletions

View file

@ -98,6 +98,7 @@ typedef struct rte {
#endif #endif
#ifdef CONFIG_RIP #ifdef CONFIG_RIP
struct { struct {
node garbage; /* List for garbage collection */
byte metric; /* RIP metric */ byte metric; /* RIP metric */
u16 tag; /* External route tag */ u16 tag; /* External route tag */
} rip; } rip;

View file

@ -30,6 +30,9 @@ int infinity = 16;
/* FIXME: should be 30 */ /* FIXME: should be 30 */
#define RIP_TIME 5 #define RIP_TIME 5
/* FIXME: should be 120+180 */
#define RIP_GB_TIME 30
static void static void
rip_reply(struct proto *p) rip_reply(struct proto *p)
{ {
@ -148,7 +151,7 @@ givemore:
packet->block[i].network = c->sendptr->network; packet->block[i].network = c->sendptr->network;
packet->block[i].netmask = ipa_mkmask( c->sendptr->pxlen ); packet->block[i].netmask = ipa_mkmask( c->sendptr->pxlen );
packet->block[i].nexthop = IPA_NONE; /* FIXME: How should I set it? */ packet->block[i].nexthop = IPA_NONE; /* FIXME: How should I set it? */
packet->block[i].metric = htonl( c->sendptr->metric ); packet->block[i].metric = htonl( c->sendptr->metric?:1 );
if (ipa_equal(c->sendptr->whotoldme, s->daddr)) { if (ipa_equal(c->sendptr->whotoldme, s->daddr)) {
debug( "(split horizont)" ); debug( "(split horizont)" );
/* FIXME: should we do it in all cases? */ /* FIXME: should we do it in all cases? */
@ -174,7 +177,7 @@ givemore:
} }
static void static void
rip_sendto( struct proto *p, ip_addr daddr, int dport ) rip_sendto( struct proto *p, ip_addr daddr, int dport, struct iface *iface )
{ {
struct rip_connection *c = mb_alloc( p->pool, sizeof( struct rip_connection )); struct rip_connection *c = mb_alloc( p->pool, sizeof( struct rip_connection ));
static int num = 0; static int num = 0;
@ -192,6 +195,7 @@ rip_sendto( struct proto *p, ip_addr daddr, int dport )
c->send->tx_hook = rip_tx; c->send->tx_hook = rip_tx;
c->send->err_hook = rip_tx_err; c->send->err_hook = rip_tx_err;
c->send->data = c; c->send->data = c;
c->send->iface = iface;
c->send->tbuf = mb_alloc( p->pool, sizeof( struct rip_packet )); c->send->tbuf = mb_alloc( p->pool, sizeof( struct rip_packet ));
if (sk_open(c->send)<0) { if (sk_open(c->send)<0) {
log( L_ERR "Could not open socket for data send to %I:%d\n", daddr, dport ); log( L_ERR "Could not open socket for data send to %I:%d\n", daddr, dport );
@ -223,6 +227,7 @@ advertise_entry( struct proto *p, struct rip_block *b, ip_addr whotoldme )
rta *a, A; rta *a, A;
rte *r; rte *r;
net *n; net *n;
neighbor *neighbor;
debug( "rip: Advertising entry..." ); debug( "rip: Advertising entry..." );
bzero(&A, sizeof(A)); bzero(&A, sizeof(A));
@ -235,13 +240,21 @@ advertise_entry( struct proto *p, struct rip_block *b, ip_addr whotoldme )
A.flags = 0; A.flags = 0;
A.gw = ipa_nonzero(b->nexthop) ? b->nexthop : whotoldme; A.gw = ipa_nonzero(b->nexthop) ? b->nexthop : whotoldme;
A.from = whotoldme; A.from = whotoldme;
A.iface = /* FIXME: need to set somehow */ NULL;
neighbor = neigh_find( p, &A.gw, 0 );
if (!neighbor) {
log( L_ERR "%I asked me to route %I/%I using not-neighbor %I.\n", A.from, b->network, b->netmask, A.gw );
return;
}
A.iface = neighbor->iface;
/* set to: interface of nexthop */ /* set to: interface of nexthop */
a = rta_lookup(&A); a = rta_lookup(&A);
n = net_get( &master_table, 0, b->network, ipa_mklen( b->netmask )); /* FIXME: should verify that it really is netmask */ n = net_get( &master_table, 0, b->network, ipa_mklen( b->netmask )); /* FIXME: should verify that it really is netmask */
r = rte_get_temp(a); r = rte_get_temp(a);
r->u.rip.metric = ntohl(b->metric); r->u.rip.metric = ntohl(b->metric);
r->u.rip.tag = ntohl(b->tag); r->u.rip.tag = ntohl(b->tag);
r->net = n;
r->pflags = 0; /* Here go my flags */ r->pflags = 0; /* Here go my flags */
rte_update( n, p, r ); rte_update( n, p, r );
debug( "done\n" ); debug( "done\n" );
@ -287,7 +300,7 @@ rip_process_packet( struct proto *p, struct rip_packet *packet, int num, ip_addr
switch( packet->heading.command ) { switch( packet->heading.command ) {
case RIPCMD_REQUEST: debug( "Asked to send my routing table\n" ); case RIPCMD_REQUEST: debug( "Asked to send my routing table\n" );
rip_sendto( p, whotoldme, port ); rip_sendto( p, whotoldme, port, NULL ); /* no broadcast */
break; break;
case RIPCMD_RESPONSE: debug( "Part of routing table came\n" ); case RIPCMD_RESPONSE: debug( "Part of routing table came\n" );
if (port != RIP_PORT) { if (port != RIP_PORT) {
@ -368,25 +381,38 @@ rip_timer(timer *t)
CHK_MAGIC; CHK_MAGIC;
debug( "RIP: tick tock\n" ); debug( "RIP: tick tock\n" );
WALK_LIST_DELSAFE( e, et, P->garbage ) {
rte *rte;
rte = SKIP_BACK( struct rte, u.rip.garbage, e );
debug( "Garbage: " ); rte_dump( rte );
#if 0 if (now - rte->lastmod > (RIP_GB_TIME)) {
/* This needs to be completely redone. We need garbage collecting based on main routing table, not on our copy */ debug( "RIP: entry is too old: " ); rte_dump( rte );
WALK_LIST_DELSAFE( e, et, P->rtable ) { rte_discard(rte);
if (!(E->flags & RIP_F_EXTERNAL)) }
if ((now - E->updated) > (180+120)) {
debug( "RIP: entry is too old: " );
rip_dump_entry( E );
kill_entry( p, E );
}
} }
#endif
#if 0
debug( "RIP: Broadcasting routing tables\n" ); debug( "RIP: Broadcasting routing tables\n" );
rip_sendto( p, _MI( 0x0a000001 ), RIP_PORT ); {
#endif struct iface *iface;
/* FIXME: Should broadcast routing tables to all known interfaces every 30 seconds */ /* FIXME: needs to be configured */
WALK_LIST( iface, iface_list ) {
if (!(iface->flags & IF_UP)) continue;
if (iface->flags & (IF_IGNORE | IF_LOOPBACK)) continue;
if (iface->flags & IF_BROADCAST) {
rip_sendto( p, iface->brd, RIP_PORT, iface );
continue;
}
if (iface->flags & IF_UNNUMBERED) {
rip_sendto( p, iface->opposite, RIP_PORT, NULL );
continue;
}
debug( "RIP/send: can not happen\n" );
}
}
debug( "RIP: tick tock done\n" ); debug( "RIP: tick tock done\n" );
} }
@ -407,6 +433,7 @@ rip_start(struct proto *p)
die( "RIP/%s: could not listen\n", p->name ); die( "RIP/%s: could not listen\n", p->name );
init_list( &P->rtable ); init_list( &P->rtable );
init_list( &P->connections ); init_list( &P->connections );
init_list( &P->garbage );
P->timer = tm_new( p->pool ); P->timer = tm_new( p->pool );
P->timer->data = p; P->timer->data = p;
P->timer->randomize = 5; P->timer->randomize = 5;
@ -479,7 +506,7 @@ rip_rt_notify(struct proto *p, struct network *net, struct rte *new, struct rte
} }
} }
static int /* FIXME: should not they pass me struct proto * ? */ static int
rip_rte_better(struct rte *new, struct rte *old) rip_rte_better(struct rte *new, struct rte *old)
{ {
if (old->u.rip.metric < new->u.rip.metric) if (old->u.rip.metric < new->u.rip.metric)
@ -503,18 +530,34 @@ rip_rta_same(rta *a, rta *b)
return 1; return 1;
} }
static void
rip_rte_insert(net *net, rte *rte)
{
struct proto *p = rte->attrs->proto;
add_head( &P->garbage, &rte->u.rip.garbage );
}
static void
rip_rte_remove(net *net, rte *rte)
{
struct proto *p = rte->attrs->proto;
rem_node( &rte->u.rip.garbage );
}
static void static void
rip_preconfig(struct protocol *x) rip_preconfig(struct protocol *x)
{ {
struct proto *p = proto_new(&proto_rip, sizeof(struct rip_data)); struct proto *p = proto_new(&proto_rip, sizeof(struct rip_data));
debug( "RIP: preconfig\n" ); debug( "RIP: preconfig\n" );
p->preference = DEF_PREF_DIRECT; p->preference = DEF_PREF_RIP;
p->start = rip_start; p->start = rip_start;
p->if_notify = rip_if_notify; p->if_notify = rip_if_notify;
p->rt_notify = rip_rt_notify; p->rt_notify = rip_rt_notify;
p->rte_better = rip_rte_better; p->rte_better = rip_rte_better;
p->rta_same = rip_rta_same; p->rta_same = rip_rta_same;
p->rte_insert = rip_rte_insert;
p->rte_remove = rip_rte_remove;
p->dump = rip_dump; p->dump = rip_dump;
} }

View file

@ -60,6 +60,7 @@ struct rip_data {
timer *timer; timer *timer;
list connections; list connections;
list rtable; list rtable;
list garbage;
int magic; int magic;
}; };