diff options
author | fenner <fenner@FreeBSD.org> | 1999-01-20 07:55:57 +0000 |
---|---|---|
committer | fenner <fenner@FreeBSD.org> | 1999-01-20 07:55:57 +0000 |
commit | cdafdeb1c3a36e7c0b59fd4f6f0b90b3b1cc27f8 (patch) | |
tree | ccf540c94528b6508bed4892f971ca6ff14c68a7 /usr.sbin/mrouted/prune.c | |
parent | 1de14dda19dab2b943562326a3d97281534dc709 (diff) | |
download | FreeBSD-src-cdafdeb1c3a36e7c0b59fd4f6f0b90b3b1cc27f8.zip FreeBSD-src-cdafdeb1c3a36e7c0b59fd4f6f0b90b3b1cc27f8.tar.gz |
Merge conflicts from 3.9-beta3+IOS12. The conflicts were huge; cvs's
merge algorithm failed on several files, and the translation of the
mrouted man page to mdoc didn't help...
Diffstat (limited to 'usr.sbin/mrouted/prune.c')
-rw-r--r-- | usr.sbin/mrouted/prune.c | 1274 |
1 files changed, 795 insertions, 479 deletions
diff --git a/usr.sbin/mrouted/prune.c b/usr.sbin/mrouted/prune.c index 23ba638..3586137 100644 --- a/usr.sbin/mrouted/prune.c +++ b/usr.sbin/mrouted/prune.c @@ -5,45 +5,37 @@ * * The mrouted program is COPYRIGHT 1989 by The Board of Trustees of * Leland Stanford Junior University. + * + * + * prune.c,v 3.8.4.59 1998/03/01 02:06:32 fenner Exp */ #ifndef lint static const char rcsid[] = - "$Id$"; + "$Id: prune.c,v 1.15 1998/01/16 07:17:44 charnier Exp $"; #endif /* not lint */ #include "defs.h" extern int cache_lifetime; -extern int max_prune_lifetime; +extern int prune_lifetime; extern struct rtentry *routing_table; extern int phys_vif; +extern int allow_black_holes; + /* - * dither cache lifetime to obtain a value between x and 2*x + * randomize value to obtain a value between .5x and 1.5x + * in order to prevent synchronization */ #ifdef SYSV -#define CACHE_LIFETIME(x) ((x) + (lrand48() % (x))) +#define JITTERED_VALUE(x) ((x)/2 + (lrand48() % (x))) #else -#define CACHE_LIFETIME(x) ((x) + (random() % (x))) +#define JITTERED_VALUE(x) ((x)/2 + (random() % (x))) #endif +#define CACHE_LIFETIME(x) JITTERED_VALUE(x) /* XXX */ -#define CHK_GS(x, y) { \ - switch(x) { \ - case 2: \ - case 4: \ - case 8: \ - case 16: \ - case 32: \ - case 64: \ - case 128: \ - case 256: y = 1; \ - break; \ - default: y = 0; \ - } \ - } - struct gtable *kernel_table; /* ptr to list of kernel grp entries*/ static struct gtable *kernel_no_route; /* list of grp entries w/o routes */ struct gtable *gtp; /* pointer for kernel rt entries */ @@ -52,17 +44,20 @@ unsigned int kroutes; /* current number of cache entries */ /**************************************************************************** Functions that are local to prune.c ****************************************************************************/ +static int scoped_addr __P((vifi_t vifi, u_int32 addr)); static void prun_add_ttls __P((struct gtable *gt)); static int pruning_neighbor __P((vifi_t vifi, u_int32 addr)); static int can_mtrace __P((vifi_t vifi, u_int32 addr)); static struct ptable * find_prune_entry __P((u_int32 vr, struct ptable *pt)); +static void remove_sources __P((struct gtable *gt)); +static void rexmit_prune __P((void *arg)); static void expire_prune __P((vifi_t vifi, struct gtable *gt)); static void send_prune __P((struct gtable *gt)); static void send_graft __P((struct gtable *gt)); static void send_graft_ack __P((u_int32 src, u_int32 dst, - u_int32 origin, u_int32 grp)); + u_int32 origin, u_int32 grp, + vifi_t vifi)); static void update_kernel __P((struct gtable *g)); -static char * scaletime __P((u_long t)); /* * Updates the ttl values for each vif. @@ -84,16 +79,34 @@ prun_add_ttls(gt) /* * checks for scoped multicast addresses + * XXX I want to make the check of allow_black_holes based on ALLOW_BLACK_HOLES + * but macros are not functions. */ #define GET_SCOPE(gt) { \ register vifi_t _i; \ - if ((ntohl((gt)->gt_mcastgrp) & 0xff000000) == 0xef000000) \ + VIFM_CLRALL((gt)->gt_scope); \ + if (allow_black_holes || \ + (ntohl((gt)->gt_mcastgrp) & 0xff000000) == 0xef000000) \ for (_i = 0; _i < numvifs; _i++) \ if (scoped_addr(_i, (gt)->gt_mcastgrp)) \ VIFM_SET(_i, (gt)->gt_scope); \ + } \ + if ((gt)->gt_route == NULL || ((gt)->gt_route->rt_parent != NO_VIF && \ + VIFM_ISSET((gt)->gt_route->rt_parent, (gt)->gt_scope))) \ + VIFM_SETALL((gt)->gt_scope); + +#define APPLY_SCOPE(gt) VIFM_CLR_MASK((gt)->gt_grpmems, (gt)->gt_scope) + +#define GET_MEMBERSHIP(gt, vifi) { \ + if ((gt)->gt_route && \ + VIFM_ISSET((vifi), (gt)->gt_route->rt_children) && \ + (!SUBS_ARE_PRUNED((gt)->gt_route->rt_subordinates, \ + uvifs[vifi].uv_nbrmap, (gt)->gt_prunes) || \ + grplst_mem((vifi), (gt)->gt_mcastgrp))) \ + VIFM_SET((vifi), (gt)->gt_grpmems); \ } -int +static int scoped_addr(vifi, addr) vifi_t vifi; u_int32 addr; @@ -107,6 +120,38 @@ scoped_addr(vifi, addr) return 0; } +/* + * Determine the list of outgoing vifs, based upon + * route subordinates, prunes received, and group + * memberships. + */ +void +determine_forwvifs(gt) + struct gtable *gt; +{ + vifi_t i; + + VIFM_CLRALL(gt->gt_grpmems); + for (i = 0; i < numvifs; i++) { + GET_MEMBERSHIP(gt, i); + } + GET_SCOPE(gt); + APPLY_SCOPE(gt); +} + +/* + * Send a prune or a graft if necessary. + */ +void +send_prune_or_graft(gt) + struct gtable *gt; +{ + if (VIFM_ISEMPTY(gt->gt_grpmems)) + send_prune(gt); + else if (gt->gt_prsent_timer) + send_graft(gt); +} + /* * Determine if mcastgrp has a listener on vifi */ @@ -181,15 +226,8 @@ pruning_neighbor(vifi, addr) if (n == NULL) return 0; - if (n->al_flags & NF_PRUNE) - return 1; - - /* - * Versions from 3.0 to 3.4 relied on the version number to identify - * that they could handle pruning. - */ vers = NBR_VERS(n); - return (vers >= 0x0300 && vers <= 0x0304); + return (vers >= 0x0300 && ((vers & 0xff00) != 0x0a00)); } /* @@ -204,17 +242,10 @@ can_mtrace(vifi, addr) int vers; if (n == NULL) - return 0; + return 1; /* fail "safe" */ - if (n->al_flags & NF_MTRACE) - return 1; - - /* - * Versions 3.3 and 3.4 relied on the version number to identify - * that they could handle traceroute. - */ vers = NBR_VERS(n); - return (vers >= 0x0303 && vers <= 0x0304); + return (vers >= 0x0303 && ((vers & 0xff00) != 0x0a00)); } /* @@ -235,6 +266,76 @@ find_prune_entry(vr, pt) } /* + * Remove all the sources hanging off the group table entry from the kernel + * cache. Remember the packet counts wherever possible, to keep the mtrace + * counters consistent. This prepares for possible prune retransmission, + * either on a multi-access network or when a prune that we sent upstream + * has expired. + */ +static void +remove_sources(gt) + struct gtable *gt; +{ + struct stable *st; + struct sioc_sg_req sg_req; + + sg_req.grp.s_addr = gt->gt_mcastgrp; + + /* + * call k_del_rg() on every one of the gt->gt_srctbl entries + * but first save the packet count so that the mtrace packet + * counters can remain approximately correct. There's a race + * here but it's minor. + */ + for (st = gt->gt_srctbl; st; st = st->st_next) { + if (st->st_ctime == 0) + continue; + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "rexmit_prune deleting (%s %s) (next is %d sec)", + inet_fmt(st->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2), + gt->gt_prune_rexmit); + sg_req.src.s_addr = st->st_origin; + if (ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) < 0) { + sg_req.pktcnt = 0; + } + k_del_rg(st->st_origin, gt); + st->st_ctime = 0; /* flag that it's not in the kernel any more */ + st->st_savpkt += sg_req.pktcnt; + kroutes--; + } + + /* + * Now, add_table_entry will prune when asked to add a cache entry. + */ +} + +/* + * Prepare for possible prune retransmission + */ +static void +rexmit_prune(arg) + void *arg; +{ + struct gtable *gt = *(struct gtable **)arg; + + free(arg); + + gt->gt_rexmit_timer = 0; + + /* Make sure we're still not forwarding traffic */ + if (!VIFM_ISEMPTY(gt->gt_grpmems)) { + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "rexmit_prune (%s %s): gm:%x", + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), + gt->gt_grpmems); + return; + } + + remove_sources(gt); +} + +/* * Send a prune message to the dominant router for * this source. * @@ -248,38 +349,77 @@ send_prune(gt) char *p; int i; int datalen; - u_int32 src; u_int32 dst; u_int32 tmp; + int rexmitting = 0; + struct uvif *v; - /* Don't process any prunes if router is not pruning */ - if (pruning == 0) - return; - - /* Can't process a prune if we don't have an associated route */ - if (gt->gt_route == NULL) + /* + * Can't process a prune if we don't have an associated route + * or if the route points to a local interface. + */ + if (gt->gt_route == NULL || gt->gt_route->rt_parent == NO_VIF || + gt->gt_route->rt_gateway == 0) return; /* Don't send a prune to a non-pruning router */ if (!pruning_neighbor(gt->gt_route->rt_parent, gt->gt_route->rt_gateway)) return; + v = &uvifs[gt->gt_route->rt_parent]; /* * sends a prune message to the router upstream. */ - src = uvifs[gt->gt_route->rt_parent].uv_lcl_addr; +#if 0 + dst = v->uv_flags & VIFF_TUNNEL ? dvmrp_group : gt->gt_route->rt_gateway; /*XXX*/ +#else dst = gt->gt_route->rt_gateway; +#endif p = send_buf + MIN_IP_HEADER_LEN + IGMP_MINLEN; datalen = 0; /* - * determine prune lifetime + * determine prune lifetime, if this isn't a retransmission. + * + * Use interface-specified lifetime if there is one. */ - gt->gt_prsent_timer = gt->gt_timer; - for (pt = gt->gt_pruntbl; pt; pt = pt->pt_next) - if (pt->pt_timer < gt->gt_prsent_timer) - gt->gt_prsent_timer = pt->pt_timer; + if (gt->gt_prsent_timer == 0) { + int l = prune_lifetime; + + if (v->uv_prune_lifetime != 0) + l = v->uv_prune_lifetime; + + gt->gt_prsent_timer = JITTERED_VALUE(l); + for (pt = gt->gt_pruntbl; pt; pt = pt->pt_next) + if (pt->pt_timer < gt->gt_prsent_timer) + gt->gt_prsent_timer = pt->pt_timer; + } else if (gt->gt_prsent_timer < 0) { + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "asked to rexmit? (%s,%s)/%d on vif %d to %s with negative time", + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), + gt->gt_prsent_timer, gt->gt_route->rt_parent, + inet_fmt(gt->gt_route->rt_gateway, s3)); + return; + } else + rexmitting = 1; + + if (rexmitting && !(v->uv_flags & VIFF_REXMIT_PRUNES)) { + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "not rexmitting prune for (%s %s)/%d on vif %d to %s", + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), + gt->gt_prsent_timer, gt->gt_route->rt_parent, + inet_fmt(gt->gt_route->rt_gateway, s3)); + return; + } + if (gt->gt_prsent_timer <= MIN_PRUNE_LIFE) { + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "not bothering to send prune for (%s,%s)/%d on vif %d to %s because it's too short", + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), + gt->gt_prsent_timer, gt->gt_route->rt_parent, + inet_fmt(gt->gt_route->rt_gateway, s3)); + return; + } /* * If we have a graft pending, cancel graft retransmission @@ -295,14 +435,27 @@ send_prune(gt) *p++ = ((char *)&(tmp))[i]; datalen += 12; - send_igmp(src, dst, IGMP_DVMRP, DVMRP_PRUNE, - htonl(MROUTED_LEVEL), datalen); + send_on_vif(v, dst, DVMRP_PRUNE, datalen); - log(LOG_DEBUG, 0, "sent prune for (%s %s)/%d on vif %d to %s", - inet_fmts(gt->gt_route->rt_origin, gt->gt_route->rt_originmask, s1), - inet_fmt(gt->gt_mcastgrp, s2), + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "%s prune for (%s %s)/%d on vif %d to %s", + rexmitting ? "rexmitted" : "sent", + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), gt->gt_prsent_timer, gt->gt_route->rt_parent, inet_fmt(gt->gt_route->rt_gateway, s3)); + + if ((v->uv_flags & VIFF_REXMIT_PRUNES) && + gt->gt_rexmit_timer == 0 && + gt->gt_prsent_timer > gt->gt_prune_rexmit) { + struct gtable **arg = + (struct gtable **)malloc(sizeof (struct gtable **)); + + *arg = gt; + gt->gt_rexmit_timer = timer_setTimer( + JITTERED_VALUE(gt->gt_prune_rexmit), + rexmit_prune, arg); + gt->gt_prune_rexmit *= 2; + } } /* @@ -319,16 +472,28 @@ send_graft(gt) register char *p; register int i; int datalen; - u_int32 src; u_int32 dst; /* Can't send a graft without an associated route */ - if (gt->gt_route == NULL) + if (gt->gt_route == NULL || gt->gt_route->rt_parent == NO_VIF) { + gt->gt_grftsnt = 0; return; - - src = uvifs[gt->gt_route->rt_parent].uv_lcl_addr; + } + + gt->gt_prsent_timer = 0; + gt->gt_prune_rexmit = PRUNE_REXMIT_VAL; + if (gt->gt_rexmit_timer) + timer_clearTimer(gt->gt_rexmit_timer); + + if (gt->gt_grftsnt == 0) + gt->gt_grftsnt = 1; + +#if 0 + dst = uvifs[gt->gt_route->rt_parent].uv_flags & VIFF_TUNNEL ? dvmrp_group : gt->gt_route->rt_gateway; /*XXX*/ +#else dst = gt->gt_route->rt_gateway; - +#endif + p = send_buf + MIN_IP_HEADER_LEN + IGMP_MINLEN; datalen = 0; @@ -338,13 +503,10 @@ send_graft(gt) *p++ = ((char *)&(gt->gt_mcastgrp))[i]; datalen += 8; - if (datalen != 0) { - send_igmp(src, dst, IGMP_DVMRP, DVMRP_GRAFT, - htonl(MROUTED_LEVEL), datalen); - } + send_on_vif(&uvifs[gt->gt_route->rt_parent], dst, DVMRP_GRAFT, datalen); + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "sent graft for (%s %s) to %s on vif %d", - inet_fmts(gt->gt_route->rt_origin, gt->gt_route->rt_originmask, s1), - inet_fmt(gt->gt_mcastgrp, s2), + RT_FMT(gt->gt_route, s1), inet_fmt(gt->gt_mcastgrp, s2), inet_fmt(gt->gt_route->rt_gateway, s3), gt->gt_route->rt_parent); } @@ -352,11 +514,12 @@ send_graft(gt) * Send an ack that a graft was received */ static void -send_graft_ack(src, dst, origin, grp) +send_graft_ack(src, dst, origin, grp, vifi) u_int32 src; u_int32 dst; u_int32 origin; u_int32 grp; + vifi_t vifi; { register char *p; register int i; @@ -371,11 +534,24 @@ send_graft_ack(src, dst, origin, grp) *p++ = ((char *)&(grp))[i]; datalen += 8; - send_igmp(src, dst, IGMP_DVMRP, DVMRP_GRAFT_ACK, - htonl(MROUTED_LEVEL), datalen); + if (vifi == NO_VIF) + send_igmp(src, dst, IGMP_DVMRP, DVMRP_GRAFT_ACK, + htonl(MROUTED_LEVEL), datalen); + else { +#if 0 + if (uvifs[vifi].uv_flags & VIFF_TUNNEL) + dst = dvmrp_group; /* XXX */ +#endif + send_on_vif(&uvifs[vifi], dst, DVMRP_GRAFT_ACK, datalen); + } - log(LOG_DEBUG, 0, "sent graft ack for (%s, %s) to %s", - inet_fmt(origin, s1), inet_fmt(grp, s2), inet_fmt(dst, s3)); + IF_DEBUG(DEBUG_PRUNE) + if (vifi == NO_VIF) + log(LOG_DEBUG, 0, "sent graft ack for (%s, %s) to %s", + inet_fmt(origin, s1), inet_fmt(grp, s2), inet_fmt(dst, s3)); + else + log(LOG_DEBUG, 0, "sent graft ack for (%s, %s) to %s on vif %d", + inet_fmt(origin, s1), inet_fmt(grp, s2), inet_fmt(dst, s3), vifi); } /* @@ -388,7 +564,8 @@ update_kernel(g) struct stable *st; for (st = g->gt_srctbl; st; st = st->st_next) - k_add_rg(st->st_origin, g); + if (st->st_ctime != 0) + k_add_rg(st->st_origin, g); } /**************************************************************************** @@ -404,7 +581,7 @@ update_kernel(g) */ struct gtable * find_grp(grp) - u_long grp; + u_int32 grp; { struct gtable *gt; @@ -424,7 +601,7 @@ find_grp(grp) struct stable * find_grp_src(gt, src) struct gtable *gt; - u_long src; + u_int32 src; { struct stable *st; u_long grp = gt->gt_mcastgrp; @@ -445,9 +622,9 @@ int next_grp_src_mask(gtpp, stpp, grp, src, mask) struct gtable **gtpp; /* ordered by group */ struct stable **stpp; /* ordered by source */ - u_long grp; - u_long src; - u_long mask; + u_int32 grp; + u_int32 src; + u_int32 mask; { struct gtable *gt, *gbest = NULL; struct stable *st, *sbest = NULL; @@ -509,76 +686,6 @@ refresh_sg(sg, gt, st) } /* - * Return pointer to a specific route entry. This must be a separate - * function from find_route() which modifies rtp. - */ -struct rtentry * -snmp_find_route(src, mask) - register u_long src, mask; -{ - register struct rtentry *rt; - - for (rt = routing_table; rt; rt = rt->rt_next) { - if (src == rt->rt_origin && mask == rt->rt_originmask) - return rt; - } - return NULL; -} - -/* - * Find next route entry > specification - */ -int -next_route(rtpp, src, mask) - struct rtentry **rtpp; - u_long src; - u_long mask; -{ - struct rtentry *rt, *rbest = NULL; - - /* Among all entries > spec, find "lowest" one in order */ - for (rt = routing_table; rt; rt=rt->rt_next) { - if ((ntohl(rt->rt_origin) > ntohl(src) - || (ntohl(rt->rt_origin) == ntohl(src) - && ntohl(rt->rt_originmask) > ntohl(mask))) - && (!rbest || (ntohl(rt->rt_origin) < ntohl(rbest->rt_origin)) - || (ntohl(rt->rt_origin) == ntohl(rbest->rt_origin) - && ntohl(rt->rt_originmask) < ntohl(rbest->rt_originmask)))) - rbest = rt; - } - (*rtpp) = rbest; - return (*rtpp)!=0; -} - -/* - * Given a routing table entry, and a vifi, find the next vifi/entry - */ -int -next_route_child(rtpp, src, mask, vifi) - struct rtentry **rtpp; - u_long src; - u_long mask; - vifi_t *vifi; /* vif at which to start looking */ -{ - struct rtentry *rt; - - /* Get (S,M) entry */ - if (!((*rtpp) = snmp_find_route(src,mask))) - if (!next_route(rtpp, src, mask)) - return 0; - - /* Continue until we get one with a valid next vif */ - do { - for (; (*rtpp)->rt_children && *vifi<numvifs; (*vifi)++) - if (VIFM_ISSET(*vifi, (*rtpp)->rt_children)) - return 1; - *vifi = 0; - } while( next_route(rtpp, (*rtpp)->rt_origin, (*rtpp)->rt_originmask) ); - - return 0; -} - -/* * Given a routing table entry, and a vifi, find the next entry * equal to or greater than those */ @@ -586,13 +693,11 @@ int next_child(gtpp, stpp, grp, src, mask, vifi) struct gtable **gtpp; struct stable **stpp; - u_long grp; - u_long src; - u_long mask; + u_int32 grp; + u_int32 src; + u_int32 mask; vifi_t *vifi; /* vif at which to start looking */ { - struct stable *st; - /* Get (G,S,M) entry */ if (mask!=0xFFFFFFFF || !((*gtpp) = find_grp(grp)) @@ -635,7 +740,15 @@ add_table_entry(origin, mcastgrp) struct rtentry *r; struct gtable *gt,**gtnp,*prev_gt; struct stable *st,**stnp; - vifi_t i; + + /* + * Since we have to enable mrouting to get the version number, + * some cache creation requests can sneak through. Ignore them + * since we're not going to do useful stuff until we've performed + * final initialization. + */ + if (!did_final_init) + return; #ifdef DEBUG_MFC md_log(MD_MISS, origin, mcastgrp); @@ -671,35 +784,20 @@ add_table_entry(origin, mcastgrp) gt->gt_mcastgrp = mcastgrp; gt->gt_timer = CACHE_LIFETIME(cache_lifetime); time(>->gt_ctime); - gt->gt_grpmems = 0; - gt->gt_scope = 0; gt->gt_prsent_timer = 0; gt->gt_grftsnt = 0; gt->gt_srctbl = NULL; gt->gt_pruntbl = NULL; gt->gt_route = r; + gt->gt_rexmit_timer = 0; + NBRM_CLRALL(gt->gt_prunes); + gt->gt_prune_rexmit = PRUNE_REXMIT_VAL; #ifdef RSRR gt->gt_rsrr_cache = NULL; #endif - if (r != NULL) { - /* obtain the multicast group membership list */ - for (i = 0; i < numvifs; i++) { - if (VIFM_ISSET(i, r->rt_children) && - !(VIFM_ISSET(i, r->rt_leaves))) - VIFM_SET(i, gt->gt_grpmems); - - if (VIFM_ISSET(i, r->rt_leaves) && grplst_mem(i, mcastgrp)) - VIFM_SET(i, gt->gt_grpmems); - } - GET_SCOPE(gt); - if (VIFM_ISSET(r->rt_parent, gt->gt_scope)) - gt->gt_scope = -1; - gt->gt_grpmems &= ~gt->gt_scope; - } else { - gt->gt_scope = -1; - gt->gt_grpmems = 0; - } + /* Calculate forwarding vifs */ + determine_forwvifs(gt); /* update ttls */ prun_add_ttls(gt); @@ -716,8 +814,7 @@ add_table_entry(origin, mcastgrp) g = gtp ? gtp->gt_gnext : kernel_table; log(LOG_WARNING, 0, "Entry for (%s %s) (rt:%x) exists (rt:%x)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), - inet_fmt(g->gt_mcastgrp, s2), + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), r, g->gt_route); } else { if (gtp) { @@ -751,39 +848,46 @@ add_table_entry(origin, mcastgrp) st->st_origin = origin; st->st_pktcnt = 0; + st->st_savpkt = 0; + time(&st->st_ctime); st->st_next = *stnp; *stnp = st; } else { + if (st->st_ctime == 0) { + /* An old source which we're keeping around for statistics */ + time(&st->st_ctime); + } else { #ifdef DEBUG_MFC - md_log(MD_DUPE, origin, mcastgrp); + md_log(MD_DUPE, origin, mcastgrp); #endif - log(LOG_WARNING, 0, "kernel entry already exists for (%s %s)", - inet_fmt(origin, s1), inet_fmt(mcastgrp, s2)); - /* XXX Doing this should cause no harm, and may ensure - * kernel<>mrouted synchronization */ - k_add_rg(origin, gt); - return; + /* Ignore kernel->mrouted retransmissions */ + if (time(0) - st->st_ctime > 5) + log(LOG_WARNING, 0, "kernel entry already exists for (%s %s)", + inet_fmt(origin, s1), inet_fmt(mcastgrp, s2)); + k_add_rg(origin, gt); + return; + } } kroutes++; k_add_rg(origin, gt); + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "add cache entry (%s %s) gm:%x, parent-vif:%d", inet_fmt(origin, s1), inet_fmt(mcastgrp, s2), gt->gt_grpmems, r ? r->rt_parent : -1); - - /* If there are no leaf vifs - * which have this group, then - * mark this src-grp as a prune candidate. + + /* + * If there are no downstream routers that want traffic for + * this group, send (or retransmit) a prune upstream. */ - if (!gt->gt_prsent_timer && !gt->gt_grpmems && r && r->rt_gateway) + if (VIFM_ISEMPTY(gt->gt_grpmems)) send_prune(gt); } /* - * An mrouter has gone down and come up on an interface - * Forward on that interface immediately + * A router has gone down. Remove prune state pertinent to that router. */ void reset_neighbor_state(vifi, addr) @@ -805,35 +909,29 @@ reset_neighbor_state(vifi, addr) */ if (vifi == r->rt_parent) { if (addr == r->rt_gateway) { + IF_DEBUG(DEBUG_PEER) log(LOG_DEBUG, 0, "reset_neighbor_state parent reset (%s %s)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), - inet_fmt(g->gt_mcastgrp, s2)); + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2)); g->gt_prsent_timer = 0; g->gt_grftsnt = 0; - while ((st = g->gt_srctbl)) { + while ((st = g->gt_srctbl) != NULL) { g->gt_srctbl = st->st_next; - k_del_rg(st->st_origin, g); - kroutes--; + if (st->st_ctime != 0) { + k_del_rg(st->st_origin, g); + kroutes--; + } free(st); } } } else { /* - * Neighbor was not the parent, send grafts to join the groups - */ - if (g->gt_prsent_timer) { - g->gt_grftsnt = 1; - send_graft(g); - g->gt_prsent_timer = 0; - } - - /* * Remove any prunes that this router has sent us. */ ptnp = &g->gt_pruntbl; while ((pt = *ptnp) != NULL) { if (pt->pt_vifi == vifi && pt->pt_router == addr) { + NBRM_CLR(pt->pt_index, g->gt_prunes); *ptnp = pt->pt_next; free(pt); } else @@ -844,15 +942,8 @@ reset_neighbor_state(vifi, addr) * And see if we want to forward again. */ if (!VIFM_ISSET(vifi, g->gt_grpmems)) { - if (VIFM_ISSET(vifi, r->rt_children) && - !(VIFM_ISSET(vifi, r->rt_leaves))) - VIFM_SET(vifi, g->gt_grpmems); - - if (VIFM_ISSET(vifi, r->rt_leaves) && - grplst_mem(vifi, g->gt_mcastgrp)) - VIFM_SET(vifi, g->gt_grpmems); - - g->gt_grpmems &= ~g->gt_scope; + GET_MEMBERSHIP(g, vifi); + APPLY_SCOPE(g); prun_add_ttls(g); /* Update kernel state */ @@ -862,8 +953,17 @@ reset_neighbor_state(vifi, addr) rsrr_cache_send(g,1); #endif /* RSRR */ - log(LOG_DEBUG, 0, "reset member state (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + /* + * If removing this prune causes us to start forwarding + * (e.g. the neighbor rebooted), and we sent a prune upstream, + * send a graft to cancel the prune. + */ + if (!VIFM_ISEMPTY(g->gt_grpmems) && g->gt_prsent_timer) + send_graft(g); + + IF_DEBUG(DEBUG_PEER) + log(LOG_DEBUG, 0, "reset neighbor state (%s %s) gm:%x", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); } } @@ -887,18 +987,20 @@ del_table_entry(r, mcastgrp, del_flag) if (del_flag == DEL_ALL_ROUTES) { g = r->rt_groups; while (g) { + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "del_table_entry deleting (%s %s)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), - inet_fmt(g->gt_mcastgrp, s2)); + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2)); st = g->gt_srctbl; while (st) { - if (k_del_rg(st->st_origin, g) < 0) { - log(LOG_WARNING, errno, - "del_table_entry trying to delete (%s, %s)", - inet_fmt(st->st_origin, s1), - inet_fmt(g->gt_mcastgrp, s2)); + if (st->st_ctime != 0) { + if (k_del_rg(st->st_origin, g) < 0) { + log(LOG_WARNING, errno, + "del_table_entry trying to delete (%s, %s)", + inet_fmt(st->st_origin, s1), + inet_fmt(g->gt_mcastgrp, s2)); + } + kroutes--; } - kroutes--; prev_st = st; st = st->st_next; free(prev_st); @@ -925,6 +1027,9 @@ del_table_entry(r, mcastgrp, del_flag) rsrr_cache_send(g,0); rsrr_cache_clean(g); #endif /* RSRR */ + if (g->gt_rexmit_timer) + timer_clearTimer(g->gt_rexmit_timer); + prev_g = g; g = g->gt_next; free(prev_g); @@ -939,18 +1044,20 @@ del_table_entry(r, mcastgrp, del_flag) prev_g = (struct gtable *)&r->rt_groups; for (g = r->rt_groups; g; g = g->gt_next) { if (g->gt_mcastgrp == mcastgrp) { + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "del_table_entry deleting (%s %s)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), - inet_fmt(g->gt_mcastgrp, s2)); + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2)); st = g->gt_srctbl; while (st) { - if (k_del_rg(st->st_origin, g) < 0) { - log(LOG_WARNING, errno, - "del_table_entry trying to delete (%s, %s)", - inet_fmt(st->st_origin, s1), - inet_fmt(g->gt_mcastgrp, s2)); + if (st->st_ctime != 0) { + if (k_del_rg(st->st_origin, g) < 0) { + log(LOG_WARNING, errno, + "del_table_entry trying to delete (%s, %s)", + inet_fmt(st->st_origin, s1), + inet_fmt(g->gt_mcastgrp, s2)); + } + kroutes--; } - kroutes--; prev_st = st; st = st->st_next; free(prev_st); @@ -978,6 +1085,8 @@ del_table_entry(r, mcastgrp, del_flag) g->gt_next->gt_prev = NULL; prev_g->gt_next = g->gt_next; + if (g->gt_rexmit_timer) + timer_clearTimer(g->gt_rexmit_timer); #ifdef RSRR /* Send route change notification to reservation protocol. */ rsrr_cache_send(g,0); @@ -996,48 +1105,70 @@ del_table_entry(r, mcastgrp, del_flag) * update kernel table entry when a route entry changes */ void -update_table_entry(r) +update_table_entry(r, old_parent_gw) struct rtentry *r; + u_int32 old_parent_gw; { struct gtable *g; - struct ptable *pt, *prev_pt; - vifi_t i; + struct ptable *pt, **ptnp; for (g = r->rt_groups; g; g = g->gt_next) { - pt = g->gt_pruntbl; - while (pt) { - prev_pt = pt->pt_next; - free(pt); - pt = prev_pt; - } - g->gt_pruntbl = NULL; - - g->gt_grpmems = 0; + ptnp = &g->gt_pruntbl; + /* + * Delete prune entries from non-children, or non-subordinates. + */ + while ((pt = *ptnp)) { + if (!VIFM_ISSET(pt->pt_vifi, r->rt_children) || + !NBRM_ISSET(pt->pt_index, r->rt_subordinates)) { + + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "update_table_entry deleting prune for (%s %s) from %s on vif %d -%s%s", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), + inet_fmt(pt->pt_router, s3), pt->pt_vifi, + VIFM_ISSET(pt->pt_vifi, r->rt_children) ? "" : " not a child", + NBRM_ISSET(pt->pt_index, r->rt_subordinates) ? "" : " not a subordinate"); + + if (!NBRM_ISSET(pt->pt_index, g->gt_prunes)) { + log(LOG_WARNING, 0, + "gt_prunes lost track of (%s %s) from %s on vif %d", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), + inet_fmt(pt->pt_router, s3), pt->pt_vifi); + } - /* obtain the multicast group membership list */ - for (i = 0; i < numvifs; i++) { - if (VIFM_ISSET(i, r->rt_children) && - !(VIFM_ISSET(i, r->rt_leaves))) - VIFM_SET(i, g->gt_grpmems); - - if (VIFM_ISSET(i, r->rt_leaves) && grplst_mem(i, g->gt_mcastgrp)) - VIFM_SET(i, g->gt_grpmems); + NBRM_CLR(pt->pt_index, g->gt_prunes); + *ptnp = pt->pt_next; + free(pt); + continue; + } + ptnp = &((*ptnp)->pt_next); } - if (VIFM_ISSET(r->rt_parent, g->gt_scope)) - g->gt_scope = -1; - g->gt_grpmems &= ~g->gt_scope; - log(LOG_DEBUG, 0, "updating cache entries (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), - inet_fmt(g->gt_mcastgrp, s2), + IF_DEBUG(DEBUG_CACHE) + log(LOG_DEBUG, 0, "updating cache entries (%s %s) old gm:%x", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); - if (g->gt_grpmems && g->gt_prsent_timer) { - g->gt_grftsnt = 1; - send_graft(g); + /* + * Forget about a prune or graft that we sent previously if we + * have a new parent router (since the new parent router will + * know nothing about what I sent to the previous parent). The + * old parent will forget any prune state it is keeping for us. + */ + if (old_parent_gw != r->rt_gateway) { g->gt_prsent_timer = 0; + g->gt_grftsnt = 0; } + /* Recalculate membership */ + determine_forwvifs(g); + /* send a prune or graft if needed. */ + send_prune_or_graft(g); + + IF_DEBUG(DEBUG_CACHE) + log(LOG_DEBUG, 0, "updating cache entries (%s %s) new gm:%x", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), + g->gt_grpmems); + /* update ttls and add entry into kernel */ prun_add_ttls(g); update_kernel(g); @@ -1045,12 +1176,6 @@ update_table_entry(r) /* Send route change notification to reservation protocol. */ rsrr_cache_send(g,1); #endif /* RSRR */ - - /* Check if we want to prune this group */ - if (!g->gt_prsent_timer && g->gt_grpmems == 0 && r->rt_gateway) { - g->gt_timer = CACHE_LIFETIME(cache_lifetime); - send_prune(g); - } } } @@ -1065,6 +1190,7 @@ update_lclgrp(vifi, mcastgrp) struct rtentry *r; struct gtable *g; + IF_DEBUG(DEBUG_MEMBER) log(LOG_DEBUG, 0, "group %s joined on vif %d", inet_fmt(mcastgrp, s1), vifi); @@ -1077,13 +1203,14 @@ update_lclgrp(vifi, mcastgrp) VIFM_ISSET(vifi, r->rt_children)) { VIFM_SET(vifi, g->gt_grpmems); - g->gt_grpmems &= ~g->gt_scope; - if (g->gt_grpmems == 0) + APPLY_SCOPE(g); + if (VIFM_ISEMPTY(g->gt_grpmems)) continue; prun_add_ttls(g); + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "update lclgrp (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); update_kernel(g); @@ -1103,9 +1230,9 @@ delete_lclgrp(vifi, mcastgrp) vifi_t vifi; u_int32 mcastgrp; { - struct rtentry *r; struct gtable *g; + IF_DEBUG(DEBUG_MEMBER) log(LOG_DEBUG, 0, "group %s left on vif %d", inet_fmt(mcastgrp, s1), vifi); @@ -1113,28 +1240,14 @@ delete_lclgrp(vifi, mcastgrp) if (ntohl(mcastgrp) < ntohl(g->gt_mcastgrp)) break; - if (g->gt_mcastgrp == mcastgrp) { - int stop_sending = 1; - - r = g->gt_route; - /* - * If this is not a leaf, then we have router neighbors on this - * vif. Only turn off forwarding if they have all pruned. - */ - if (!VIFM_ISSET(vifi, r->rt_leaves)) { - struct listaddr *vr; - - for (vr = uvifs[vifi].uv_neighbors; vr; vr = vr->al_next) - if (find_prune_entry(vr->al_addr, g->gt_pruntbl) == NULL) { - stop_sending = 0; - break; - } - } - - if (stop_sending) { + if (g->gt_mcastgrp == mcastgrp && VIFM_ISSET(vifi, g->gt_grpmems)) { + if (g->gt_route == NULL || + SUBS_ARE_PRUNED(g->gt_route->rt_subordinates, + uvifs[vifi].uv_nbrmap, g->gt_prunes)) { VIFM_CLR(vifi, g->gt_grpmems); + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "delete lclgrp (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(g->gt_route, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); prun_add_ttls(g); @@ -1148,7 +1261,7 @@ delete_lclgrp(vifi, mcastgrp) * If there are no more members of this particular group, * send prune upstream */ - if (!g->gt_prsent_timer && g->gt_grpmems == 0 && r->rt_gateway) + if (VIFM_ISEMPTY(g->gt_grpmems) && g->gt_route->rt_gateway) send_prune(g); } } @@ -1177,15 +1290,9 @@ accept_prune(src, dst, p, datalen) u_int32 prun_tmr; vifi_t vifi; int i; - int stop_sending; struct rtentry *r; struct gtable *g; struct ptable *pt; - struct listaddr *vr; - - /* Don't process any prunes if router is not pruning */ - if (pruning == 0) - return; if ((vifi = find_vif(src, dst)) == NO_VIF) { log(LOG_INFO, 0, @@ -1211,6 +1318,15 @@ accept_prune(src, dst, p, datalen) ((char *)&prun_tmr)[i] = *p++; prun_tmr = ntohl(prun_tmr); + if (prun_tmr <= MIN_PRUNE_LIFE) { + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "ignoring prune from %s on vif %d for (%s %s)/%d because its lifetime is too short", + inet_fmt(src, s1), vifi, + inet_fmt(prun_src, s2), inet_fmt(prun_grp, s3), prun_tmr); + return; + } + + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "%s on vif %d prunes (%s %s)/%d", inet_fmt(src, s1), vifi, inet_fmt(prun_src, s2), inet_fmt(prun_grp, s3), prun_tmr); @@ -1222,10 +1338,19 @@ accept_prune(src, dst, p, datalen) g = gtp ? gtp->gt_gnext : kernel_table; r = g->gt_route; + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "found grp state, (%s %s), metric is %d, children are %x, subords are %08x%08x", + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), r->rt_metric, + r->rt_children, r->rt_subordinates.hi, r->rt_subordinates.lo); if (!VIFM_ISSET(vifi, r->rt_children)) { - log(LOG_WARNING, 0, "prune received from non-child %s for (%s %s)", + IF_DEBUG(DEBUG_PRUNE) + log(LOG_WARNING, 0, "prune received from non-child %s for (%s %s) (dominant on vif %d is %s)", inet_fmt(src, s1), inet_fmt(prun_src, s2), - inet_fmt(prun_grp, s3)); + inet_fmt(prun_grp, s3), vifi, + inet_fmt(r->rt_dominants[vifi], s4)); +#ifdef RINGBUFFER + printringbuf(); +#endif return; } if (VIFM_ISSET(vifi, g->gt_scope)) { @@ -1235,6 +1360,7 @@ accept_prune(src, dst, p, datalen) return; } if ((pt = find_prune_entry(src, g->gt_pruntbl)) != NULL) { + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "%s %d from %s for (%s %s)/%d %s %d %s %x", "duplicate prune received on vif", vifi, inet_fmt(src, s1), inet_fmt(prun_src, s2), @@ -1242,6 +1368,14 @@ accept_prune(src, dst, p, datalen) "old timer:", pt->pt_timer, "cur gm:", g->gt_grpmems); pt->pt_timer = prun_tmr; } else { + struct listaddr *n = neighbor_info(vifi, src); + + if (!n) { + log(LOG_WARNING, 0, "Prune from non-neighbor %s on vif %d!?", + inet_fmt(src, s1), vifi); + return; + } + /* allocate space for the prune structure */ pt = (struct ptable *)(malloc(sizeof(struct ptable))); if (pt == NULL) @@ -1253,28 +1387,37 @@ accept_prune(src, dst, p, datalen) pt->pt_next = g->gt_pruntbl; g->gt_pruntbl = pt; - } - /* Refresh the group's lifetime */ - g->gt_timer = CACHE_LIFETIME(cache_lifetime); - if (g->gt_timer < prun_tmr) - g->gt_timer = prun_tmr; + if (n) { + pt->pt_index = n->al_index; + NBRM_SET(n->al_index, g->gt_prunes); + } + } /* * check if any more packets need to be sent on the * vif which sent this message */ - stop_sending = 1; - for (vr = uvifs[vifi].uv_neighbors; vr; vr = vr->al_next) - if (find_prune_entry(vr->al_addr, g->gt_pruntbl) == NULL) { - stop_sending = 0; - break; - } - - if (stop_sending && !grplst_mem(vifi, prun_grp)) { + if (SUBS_ARE_PRUNED(r->rt_subordinates, + uvifs[vifi].uv_nbrmap, g->gt_prunes) && + !grplst_mem(vifi, prun_grp)) { + nbrbitmap_t tmp; + VIFM_CLR(vifi, g->gt_grpmems); + IF_DEBUG(DEBUG_PRUNE) + log(LOG_DEBUG, 0, "vifnbrs=0x%08x%08x, subord=0x%08x%08x prunes=0x%08x%08x", + uvifs[vifi].uv_nbrmap.hi,uvifs[vifi].uv_nbrmap.lo, + r->rt_subordinates.hi, r->rt_subordinates.lo, + g->gt_prunes.hi, g->gt_prunes.lo); + /* XXX debugging */ + NBRM_COPY(r->rt_subordinates, tmp); + NBRM_MASK(tmp, uvifs[vifi].uv_nbrmap); + if (!NBRM_ISSETALLMASK(g->gt_prunes, tmp)) + log(LOG_WARNING, 0, "subordinate error"); + /* XXX end debugging */ + IF_DEBUG(DEBUG_PRUNE|DEBUG_CACHE) log(LOG_DEBUG, 0, "prune (%s %s), stop sending on vif %d, gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), vifi, g->gt_grpmems); prun_add_ttls(g); @@ -1291,7 +1434,7 @@ accept_prune(src, dst, p, datalen) * interface * Send a prune message then upstream */ - if (!g->gt_prsent_timer && g->gt_grpmems == 0 && r->rt_gateway) { + if (VIFM_ISEMPTY(g->gt_grpmems) && r->rt_gateway) { send_prune(g); } } else { @@ -1300,6 +1443,7 @@ accept_prune(src, dst, p, datalen) * simply ignore the prune, as we are not forwarding this traffic * downstream. */ + IF_DEBUG(DEBUG_PRUNE|DEBUG_CACHE) log(LOG_DEBUG, 0, "%s (%s %s)/%d from %s", "prune message received with no kernel entry for", inet_fmt(prun_src, s1), inet_fmt(prun_grp, s2), @@ -1333,22 +1477,19 @@ chkgrp_graft(vifi, mcastgrp) * If the vif that was joined was a scoped vif, * ignore it ; don't graft back */ - g->gt_grpmems &= ~g->gt_scope; - if (g->gt_grpmems == 0) + APPLY_SCOPE(g); + if (VIFM_ISEMPTY(g->gt_grpmems)) continue; - /* set the flag for graft retransmission */ - g->gt_grftsnt = 1; - /* send graft upwards */ send_graft(g); - /* reset the prune timer and update cache timer*/ - g->gt_prsent_timer = 0; - g->gt_timer = max_prune_lifetime; + /* update cache timer*/ + g->gt_timer = CACHE_LIFETIME(cache_lifetime); + IF_DEBUG(DEBUG_PRUNE|DEBUG_CACHE) log(LOG_DEBUG, 0, "chkgrp graft (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); prun_add_ttls(g); @@ -1388,13 +1529,6 @@ accept_graft(src, dst, p, datalen) struct gtable *g; struct ptable *pt, **ptnp; - if ((vifi = find_vif(src, dst)) == NO_VIF) { - log(LOG_INFO, 0, - "ignoring graft from non-neighbor %s", - inet_fmt(src, s1)); - return; - } - if (datalen < 8) { log(LOG_WARNING, 0, "received non-decipherable graft from %s", @@ -1406,7 +1540,19 @@ accept_graft(src, dst, p, datalen) ((char *)&graft_src)[i] = *p++; for (i = 0; i< 4; i++) ((char *)&graft_grp)[i] = *p++; - + + vifi = find_vif(src, dst); + send_graft_ack(dst, src, graft_src, graft_grp, vifi); + + if (vifi == NO_VIF) { + log(LOG_INFO, 0, + "ignoring graft for (%s %s) from non-neighbor %s", + inet_fmt(graft_src, s2), inet_fmt(graft_grp, s3), + inet_fmt(src, s1)); + return; + } + + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "%s on vif %d grafts (%s %s)", inet_fmt(src, s1), vifi, inet_fmt(graft_src, s2), inet_fmt(graft_grp, s3)); @@ -1428,12 +1574,14 @@ accept_graft(src, dst, p, datalen) ptnp = &g->gt_pruntbl; while ((pt = *ptnp) != NULL) { if ((pt->pt_vifi == vifi) && (pt->pt_router == src)) { + NBRM_CLR(pt->pt_index, g->gt_prunes); *ptnp = pt->pt_next; free(pt); VIFM_SET(vifi, g->gt_grpmems); + IF_DEBUG(DEBUG_PRUNE|DEBUG_CACHE) log(LOG_DEBUG, 0, "accept graft (%s %s) gm:%x", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(g->gt_mcastgrp, s2), g->gt_grpmems); prun_add_ttls(g); @@ -1448,28 +1596,18 @@ accept_graft(src, dst, p, datalen) } } - /* send ack downstream */ - send_graft_ack(dst, src, graft_src, graft_grp); - g->gt_timer = max_prune_lifetime; + g->gt_timer = CACHE_LIFETIME(cache_lifetime); - if (g->gt_prsent_timer) { - /* set the flag for graft retransmission */ - g->gt_grftsnt = 1; - + if (g->gt_prsent_timer) /* send graft upwards */ send_graft(g); - - /* reset the prune sent timer */ - g->gt_prsent_timer = 0; - } } else { /* * We have no state for the source and group in question. - * We can simply acknowledge the graft, since we know - * that we have no prune state, and grafts are requests - * to remove prune state. + * This is fine, since we know that we have no prune state, and + * grafts are requests to remove prune state. */ - send_graft_ack(dst, src, graft_src, graft_grp); + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "%s (%s %s) from %s", "graft received with no kernel entry for", inet_fmt(graft_src, s1), inet_fmt(graft_grp, s2), @@ -1518,6 +1656,7 @@ accept_g_ack(src, dst, p, datalen) for (i = 0; i< 4; i++) ((char *)&grft_grp)[i] = *p++; + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "%s on vif %d acks graft (%s, %s)", inet_fmt(src, s1), vifi, inet_fmt(grft_src, s2), inet_fmt(grft_grp, s3)); @@ -1533,6 +1672,9 @@ accept_g_ack(src, dst, p, datalen) "rcvd graft ack with no kernel entry for", inet_fmt(grft_src, s1), inet_fmt(grft_grp, s2), inet_fmt(src, s3)); +#ifdef RINGBUFFER + printringbuf(); +#endif return; } } @@ -1571,6 +1713,8 @@ free_all_prunes() prev_g = g; g = g->gt_next; + if (prev_g->gt_rexmit_timer) + timer_clearTimer(prev_g->gt_rexmit_timer); free(prev_g); } r->rt_groups = NULL; @@ -1584,6 +1728,8 @@ free_all_prunes() prev_g = g; g = g->gt_next; + if (prev_g->gt_rexmit_timer) + timer_clearTimer(prev_g->gt_rexmit_timer); free(prev_g); } kernel_no_route = NULL; @@ -1607,27 +1753,32 @@ steal_sources(rt) register struct stable *st, **stnp; for (rp = rt->rt_next; rp; rp = rp->rt_next) { + if (rp->rt_groups == NULL) + continue; if ((rt->rt_origin & rp->rt_originmask) == rp->rt_origin) { + IF_DEBUG(DEBUG_ROUTE) log(LOG_DEBUG, 0, "Route for %s stealing sources from %s", - inet_fmts(rt->rt_origin, rt->rt_originmask, s1), - inet_fmts(rp->rt_origin, rp->rt_originmask, s2)); + RT_FMT(rt, s1), RT_FMT(rp, s2)); for (gt = rp->rt_groups; gt; gt = gt->gt_next) { stnp = >->gt_srctbl; while ((st = *stnp) != NULL) { if ((st->st_origin & rt->rt_originmask) == rt->rt_origin) { + IF_DEBUG(DEBUG_ROUTE) log(LOG_DEBUG, 0, "%s stealing (%s %s) from %s", - inet_fmts(rt->rt_origin, rt->rt_originmask, s1), + RT_FMT(rt, s1), inet_fmt(st->st_origin, s3), inet_fmt(gt->gt_mcastgrp, s4), - inet_fmts(rp->rt_origin, rp->rt_originmask, s2)); - if (k_del_rg(st->st_origin, gt) < 0) { - log(LOG_WARNING, errno, "%s (%s, %s)", - "steal_sources trying to delete", - inet_fmt(st->st_origin, s1), - inet_fmt(gt->gt_mcastgrp, s2)); + RT_FMT(rp, s2)); + if (st->st_ctime != 0) { + if (k_del_rg(st->st_origin, gt) < 0) { + log(LOG_WARNING, errno, "%s (%s, %s)", + "steal_sources trying to delete", + inet_fmt(st->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + } + kroutes--; } *stnp = st->st_next; - kroutes--; free(st); } else { stnp = &st->st_next; @@ -1641,22 +1792,27 @@ steal_sources(rt) while ((gt = *gtnp) != NULL) { if (gt->gt_srctbl && ((gt->gt_srctbl->st_origin & rt->rt_originmask) == rt->rt_origin)) { + IF_DEBUG(DEBUG_ROUTE) log(LOG_DEBUG, 0, "%s stealing (%s %s) from %s", - inet_fmts(rt->rt_origin, rt->rt_originmask, s1), + RT_FMT(rt, s1), inet_fmt(gt->gt_srctbl->st_origin, s3), inet_fmt(gt->gt_mcastgrp, s4), "no_route table"); - if (k_del_rg(gt->gt_srctbl->st_origin, gt) < 0) { - log(LOG_WARNING, errno, "%s (%s %s)", - "steal_sources trying to delete", - inet_fmt(gt->gt_srctbl->st_origin, s1), - inet_fmt(gt->gt_mcastgrp, s2)); + if (gt->gt_srctbl->st_ctime != 0) { + if (k_del_rg(gt->gt_srctbl->st_origin, gt) < 0) { + log(LOG_WARNING, errno, "%s (%s %s)", + "steal_sources trying to delete", + inet_fmt(gt->gt_srctbl->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + } + kroutes--; } - kroutes--; free(gt->gt_srctbl); *gtnp = gt->gt_next; if (gt->gt_next) gt->gt_next->gt_prev = gt->gt_prev; + if (gt->gt_rexmit_timer) + timer_clearTimer(gt->gt_rexmit_timer); free(gt); } else { gtnp = >->gt_next; @@ -1678,31 +1834,98 @@ age_table_entry() struct ptable *pt, **ptnp; struct sioc_sg_req sg_req; - log(LOG_DEBUG, 0, "ageing entries"); + IF_DEBUG(DEBUG_PRUNE|DEBUG_CACHE) + log(LOG_DEBUG, 0, "aging forwarding cache entries"); gtnptr = &kernel_table; while ((gt = *gtnptr) != NULL) { + vifi_t i; /* XXX Debugging */ + int fixit = 0; /* XXX Debugging */ + r = gt->gt_route; + /* XXX Debugging... */ + for (i = 0; i < numvifs; i++) { + /* + * If we're not sending on this vif, + * And this group isn't scoped on this vif, + * And I'm the parent for this route on this vif, + * And there are subordinates on this vif, + * And all of the subordinates haven't pruned, + * YELL LOUDLY + * and remember to fix it up later + */ + if (!VIFM_ISSET(i, gt->gt_grpmems) && + !VIFM_ISSET(i, gt->gt_scope) && + VIFM_ISSET(i, r->rt_children) && + NBRM_ISSETMASK(uvifs[i].uv_nbrmap, r->rt_subordinates) && + !SUBS_ARE_PRUNED(r->rt_subordinates, uvifs[i].uv_nbrmap, gt->gt_prunes)) { + log(LOG_WARNING, 0, "(%s %s) is blackholing on vif %d", + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2), i); + fixit = 1; + } + } + if (fixit) { + log(LOG_WARNING, 0, "fixing membership for (%s %s) gm:%x", + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2), gt->gt_grpmems); + determine_forwvifs(gt); + send_prune_or_graft(gt); + log(LOG_WARNING, 0, "fixed membership for (%s %s) gm:%x", + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2), gt->gt_grpmems); +#ifdef RINGBUFFER + printringbuf(); +#endif + } + /*DEBUG2*/ + /* If there are group members, + * and there are recent sources, + * and we have a route, + * and it's not directly connected, + * and we haven't sent a prune, + * if there are any cache entries in the kernel + * [if there aren't we're probably waiting to rexmit], + * YELL LOUDLY + * and send a prune + */ + if (VIFM_ISEMPTY(gt->gt_grpmems) && gt->gt_srctbl && r && r->rt_gateway && gt->gt_prsent_timer == 0) { + for (st = gt->gt_srctbl; st; st = st->st_next) + if (st->st_ctime != 0) + break; + if (st != NULL) { + log(LOG_WARNING, 0, "grpmems for (%s %s) is empty but no prune state!", RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); + send_prune_or_graft(gt); +#ifdef RINGBUFFER + printringbuf(); +#endif + } + } + /* XXX ...Debugging */ + /* advance the timer for the kernel entry */ - gt->gt_timer -= ROUTE_MAX_REPORT_DELAY; + gt->gt_timer -= TIMER_INTERVAL; /* decrement prune timer if need be */ if (gt->gt_prsent_timer > 0) { - gt->gt_prsent_timer -= ROUTE_MAX_REPORT_DELAY; + gt->gt_prsent_timer -= TIMER_INTERVAL; if (gt->gt_prsent_timer <= 0) { + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "upstream prune tmo (%s %s)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); gt->gt_prsent_timer = -1; + /* Reset the prune retransmission timer to its initial value */ + gt->gt_prune_rexmit = PRUNE_REXMIT_VAL; } } - /* retransmit graft if graft sent flag is still set */ + /* retransmit graft with exponential backoff */ if (gt->gt_grftsnt) { register int y; - CHK_GS(gt->gt_grftsnt++, y); - if (y) + + y = ++gt->gt_grftsnt; + while (y && !(y & 1)) + y >>= 1; + if (y == 1) send_graft(gt); } @@ -1713,13 +1936,24 @@ age_table_entry() */ ptnp = >->gt_pruntbl; while ((pt = *ptnp) != NULL) { - if ((pt->pt_timer -= ROUTE_MAX_REPORT_DELAY) <= 0) { + if ((pt->pt_timer -= TIMER_INTERVAL) <= 0) { + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "expire prune (%s %s) from %s on vif %d", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2), inet_fmt(pt->pt_router, s3), pt->pt_vifi); + if (gt->gt_prsent_timer > 0) { + log(LOG_WARNING, 0, "prune (%s %s) from %s on vif %d expires with %d left on prsent timer", + RT_FMT(r, s1), + inet_fmt(gt->gt_mcastgrp, s2), + inet_fmt(pt->pt_router, s3), + pt->pt_vifi, gt->gt_prsent_timer); + /* Send a graft to heal the tree. */ + send_graft(gt); + } + NBRM_CLR(pt->pt_index, gt->gt_prunes); expire_prune(pt->pt_vifi, gt); /* remove the router's prune entry and await new one */ @@ -1737,31 +1971,47 @@ age_table_entry() * Otherwise, the cache entry's timer is refreshed. */ if (gt->gt_timer <= 0) { + IF_DEBUG(DEBUG_CACHE) + log(LOG_DEBUG, 0, "(%s %s) timed out, checking for traffic", + RT_FMT(gt->gt_route, s1), + inet_fmt(gt->gt_mcastgrp, s2)); /* Check for traffic before deleting source entries */ sg_req.grp.s_addr = gt->gt_mcastgrp; stnp = >->gt_srctbl; while ((st = *stnp) != NULL) { - sg_req.src.s_addr = st->st_origin; - if (ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) < 0) { - log(LOG_WARNING, errno, "%s (%s %s)", - "age_table_entry: SIOCGETSGCNT failing for", - inet_fmt(st->st_origin, s1), - inet_fmt(gt->gt_mcastgrp, s2)); - /* Make sure it gets deleted below */ + /* + * Source entries with no ctime are not actually in the + * kernel; they have been removed by rexmit_prune() so + * are safe to remove from the list at this point. + */ + if (st->st_ctime) { + sg_req.src.s_addr = st->st_origin; + if (ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) < 0) { + log(LOG_WARNING, errno, "%s (%s %s)", + "age_table_entry: SIOCGETSGCNT failing for", + inet_fmt(st->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + /* Make sure it gets deleted below */ + sg_req.pktcnt = st->st_pktcnt; + } + } else { sg_req.pktcnt = st->st_pktcnt; } if (sg_req.pktcnt == st->st_pktcnt) { *stnp = st->st_next; + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "age_table_entry deleting (%s %s)", inet_fmt(st->st_origin, s1), inet_fmt(gt->gt_mcastgrp, s2)); - if (k_del_rg(st->st_origin, gt) < 0) { - log(LOG_WARNING, errno, - "age_table_entry trying to delete (%s %s)", - inet_fmt(st->st_origin, s1), - inet_fmt(gt->gt_mcastgrp, s2)); + if (st->st_ctime != 0) { + if (k_del_rg(st->st_origin, gt) < 0) { + log(LOG_WARNING, errno, + "age_table_entry trying to delete (%s %s)", + inet_fmt(st->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + } + kroutes--; } - kroutes--; free(st); } else { st->st_pktcnt = sg_req.pktcnt; @@ -1772,22 +2022,39 @@ age_table_entry() /* * Retain the group entry if we have downstream prunes or if * there is at least one source in the list that still has - * traffic, or if our upstream prune timer is running. + * traffic, or if our upstream prune timer or graft + * retransmission timer is running. */ if (gt->gt_pruntbl != NULL || gt->gt_srctbl != NULL || - gt->gt_prsent_timer > 0) { + gt->gt_prsent_timer > 0 || gt->gt_grftsnt > 0) { + IF_DEBUG(DEBUG_CACHE) + log(LOG_DEBUG, 0, "refresh lifetim of cache entry %s%s%s%s(%s, %s)", + gt->gt_pruntbl ? "(dstrm prunes) " : "", + gt->gt_srctbl ? "(trfc flow) " : "", + gt->gt_prsent_timer > 0 ? "(upstrm prune) " : "", + gt->gt_grftsnt > 0 ? "(grft rexmit) " : "", + RT_FMT(r, s1), + inet_fmt(gt->gt_mcastgrp, s2)); gt->gt_timer = CACHE_LIFETIME(cache_lifetime); - if (gt->gt_prsent_timer == -1) - if (gt->gt_grpmems == 0) - send_prune(gt); - else - gt->gt_prsent_timer = 0; + if (gt->gt_prsent_timer == -1) { + /* + * The upstream prune timed out. Remove any kernel + * state. + */ + gt->gt_prsent_timer = 0; + if (gt->gt_pruntbl) { + log(LOG_WARNING, 0, "upstream prune for (%s %s) expires with downstream prunes active", + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); + } + remove_sources(gt); + } gtnptr = >->gt_gnext; continue; } + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "timeout cache entry (%s, %s)", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); if (gt->gt_prev) @@ -1812,13 +2079,23 @@ age_table_entry() rsrr_cache_send(gt,0); rsrr_cache_clean(gt); #endif /* RSRR */ + if (gt->gt_rexmit_timer) + timer_clearTimer(gt->gt_rexmit_timer); + free((char *)gt); } else { - if (gt->gt_prsent_timer == -1) - if (gt->gt_grpmems == 0) - send_prune(gt); - else - gt->gt_prsent_timer = 0; + if (gt->gt_prsent_timer == -1) { + /* + * The upstream prune timed out. Remove any kernel + * state. + */ + gt->gt_prsent_timer = 0; + if (gt->gt_pruntbl) { + log(LOG_WARNING, 0, "upstream prune for (%s %s) expires with downstream prunes active", + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); + } + remove_sources(gt); + } gtnptr = >->gt_gnext; } } @@ -1830,15 +2107,18 @@ age_table_entry() gtnptr = &kernel_no_route; while ((gt = *gtnptr) != NULL) { /* advance the timer for the kernel entry */ - gt->gt_timer -= ROUTE_MAX_REPORT_DELAY; + gt->gt_timer -= TIMER_INTERVAL; if (gt->gt_timer < 0) { if (gt->gt_srctbl) { - if (k_del_rg(gt->gt_srctbl->st_origin, gt) < 0) { - log(LOG_WARNING, errno, "%s (%s %s)", - "age_table_entry trying to delete no-route", - inet_fmt(gt->gt_srctbl->st_origin, s1), - inet_fmt(gt->gt_mcastgrp, s2)); + if (gt->gt_srctbl->st_ctime != 0) { + if (k_del_rg(gt->gt_srctbl->st_origin, gt) < 0) { + log(LOG_WARNING, errno, "%s (%s %s)", + "age_table_entry trying to delete no-route", + inet_fmt(gt->gt_srctbl->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + } + kroutes--; } free(gt->gt_srctbl); } @@ -1846,6 +2126,9 @@ age_table_entry() if (gt->gt_next) gt->gt_next->gt_prev = gt->gt_prev; + if (gt->gt_rexmit_timer) + timer_clearTimer(gt->gt_rexmit_timer); + free((char *)gt); } else { gtnptr = >->gt_next; @@ -1866,19 +2149,24 @@ expire_prune(vifi, gt) /* * No need to send a graft, any prunes that we sent * will expire before any prunes that we have received. + * However, in the case that we did make a mistake, + * send a graft to compensate. */ - if (gt->gt_prsent_timer > 0) { + if (gt->gt_prsent_timer >= MIN_PRUNE_LIFE) { + IF_DEBUG(DEBUG_PRUNE) log(LOG_DEBUG, 0, "prune expired with %d left on %s", gt->gt_prsent_timer, "prsent_timer"); gt->gt_prsent_timer = 0; + send_graft(gt); } /* modify the kernel entry to forward packets */ if (!VIFM_ISSET(vifi, gt->gt_grpmems)) { struct rtentry *rt = gt->gt_route; VIFM_SET(vifi, gt->gt_grpmems); + IF_DEBUG(DEBUG_CACHE) log(LOG_DEBUG, 0, "forw again (%s %s) gm:%x vif:%d", - inet_fmts(rt->rt_origin, rt->rt_originmask, s1), + RT_FMT(rt, s1), inet_fmt(gt->gt_mcastgrp, s2), gt->gt_grpmems, vifi); prun_add_ttls(gt); @@ -1890,46 +2178,6 @@ expire_prune(vifi, gt) } } - -static char * -scaletime(t) - u_long t; -{ - static char buf1[5]; - static char buf2[5]; - static char *buf=buf1; - char s; - char *p; - - p = buf; - if (buf == buf1) - buf = buf2; - else - buf = buf1; - - if (t < 120) { - s = 's'; - } else if (t < 3600) { - t /= 60; - s = 'm'; - } else if (t < 86400) { - t /= 3600; - s = 'h'; - } else if (t < 864000) { - t /= 86400; - s = 'd'; - } else { - t /= 604800; - s = 'w'; - } - if (t > 999) - return "*** "; - - sprintf(p,"%3d%c", (int)t, s); - - return p; -} - /* * Print the contents of the cache table on file 'fp2'. */ @@ -1947,11 +2195,14 @@ dump_cache(fp2) fprintf(fp2, "Multicast Routing Cache Table (%d entries)\n%s", kroutes, - " Origin Mcast-group CTmr Age Ptmr IVif Forwvifs\n"); + " Origin Mcast-group CTmr Age Ptmr Rx IVif Forwvifs\n"); + fprintf(fp2, + "<(prunesrc:vif[idx]/tmr) prunebitmap\n%s", + ">Source Lifetime SavPkt Pkts Bytes RPFf\n"); for (gt = kernel_no_route; gt; gt = gt->gt_next) { if (gt->gt_srctbl) { - fprintf(fp2, " %-18s %-15s %-4s %-4s - -1\n", + fprintf(fp2, " %-18s %-15s %-8s %-8s - -1 (no route)\n", inet_fmts(gt->gt_srctbl->st_origin, 0xffffffff, s1), inet_fmt(gt->gt_mcastgrp, s2), scaletime(gt->gt_timer), scaletime(thyme - gt->gt_ctime)); @@ -1962,40 +2213,78 @@ dump_cache(fp2) for (gt = kernel_table; gt; gt = gt->gt_gnext) { r = gt->gt_route; fprintf(fp2, " %-18s %-15s", - inet_fmts(r->rt_origin, r->rt_originmask, s1), + RT_FMT(r, s1), inet_fmt(gt->gt_mcastgrp, s2)); - fprintf(fp2, " %-4s", scaletime(gt->gt_timer)); + fprintf(fp2, " %-8s", scaletime(gt->gt_timer)); - fprintf(fp2, " %-4s %-4s ", scaletime(thyme - gt->gt_ctime), + fprintf(fp2, " %-8s %-8s ", scaletime(thyme - gt->gt_ctime), gt->gt_prsent_timer ? scaletime(gt->gt_prsent_timer) : - " -"); + " -"); + + if (gt->gt_prune_rexmit) { + int i = gt->gt_prune_rexmit; + int n = 0; - fprintf(fp2, "%2u%c%c ", r->rt_parent, - gt->gt_prsent_timer ? 'P' : ' ', + while (i > PRUNE_REXMIT_VAL) { + n++; + i /= 2; + } + if (n == 0 && gt->gt_prsent_timer == 0) + fprintf(fp2, " -"); + else + fprintf(fp2, "%2d", n); + } else { + fprintf(fp2, " -"); + } + + fprintf(fp2, " %2u%c%c", r->rt_parent, + gt->gt_prsent_timer ? 'P' : + gt->gt_grftsnt ? 'G' : ' ', VIFM_ISSET(r->rt_parent, gt->gt_scope) ? 'B' : ' '); for (i = 0; i < numvifs; ++i) { if (VIFM_ISSET(i, gt->gt_grpmems)) fprintf(fp2, " %u ", i); else if (VIFM_ISSET(i, r->rt_children) && - !VIFM_ISSET(i, r->rt_leaves)) + NBRM_ISSETMASK(uvifs[i].uv_nbrmap, r->rt_subordinates)) fprintf(fp2, " %u%c", i, - VIFM_ISSET(i, gt->gt_scope) ? 'b' : 'p'); + VIFM_ISSET(i, gt->gt_scope) ? 'b' : + SUBS_ARE_PRUNED(r->rt_subordinates, + uvifs[i].uv_nbrmap, gt->gt_prunes) ? 'p' : '!'); } fprintf(fp2, "\n"); if (gt->gt_pruntbl) { fprintf(fp2, "<"); c = '('; for (pt = gt->gt_pruntbl; pt; pt = pt->pt_next) { - fprintf(fp2, "%c%s:%d/%d", c, inet_fmt(pt->pt_router, s1), - pt->pt_vifi, pt->pt_timer); + fprintf(fp2, "%c%s:%d[%d]/%d", c, inet_fmt(pt->pt_router, s1), + pt->pt_vifi, pt->pt_index, pt->pt_timer); c = ','; } - fprintf(fp2, ")\n"); + fprintf(fp2, ")"); + fprintf(fp2, " 0x%08lx%08lx\n",/*XXX*/ + gt->gt_prunes.hi, gt->gt_prunes.lo); } for (st = gt->gt_srctbl; st; st = st->st_next) { - fprintf(fp2, ">%s\n", inet_fmt(st->st_origin, s1)); + fprintf(fp2, ">%-18s %-8s %6ld", inet_fmt(st->st_origin, s1), + st->st_ctime ? scaletime(thyme - st->st_ctime) : "-", + st->st_savpkt); + if (st->st_ctime) { + struct sioc_sg_req sg_req; + + sg_req.src.s_addr = st->st_origin; + sg_req.grp.s_addr = gt->gt_mcastgrp; + if (ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) < 0) { + log(LOG_WARNING, errno, "SIOCGETSGCNT on (%s %s)", + inet_fmt(st->st_origin, s1), + inet_fmt(gt->gt_mcastgrp, s2)); + } else { + fprintf(fp2, " %8ld %8ld %4ld", sg_req.pktcnt, + sg_req.bytecnt, sg_req.wrong_if); + } + } + fprintf(fp2, "\n"); } } } @@ -2038,14 +2327,17 @@ accept_mtrace(src, dst, group, data, no, datalen) */ if (datalen == QLEN) { type = QUERY; + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Initial traceroute query rcvd from %s to %s", inet_fmt(src, s1), inet_fmt(dst, s2)); } else if ((datalen - QLEN) % RLEN == 0) { type = RESP; + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "In-transit traceroute query rcvd from %s to %s", inet_fmt(src, s1), inet_fmt(dst, s2)); if (IN_MULTICAST(ntohl(dst))) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Dropping multicast response"); return; } @@ -2063,23 +2355,27 @@ accept_mtrace(src, dst, group, data, no, datalen) * if it is a packet with all reports filled, drop it */ if ((rcount = (datalen - QLEN)/RLEN) == no) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "packet with all reports filled in"); return; } + IF_DEBUG(DEBUG_TRACE) { log(LOG_DEBUG, 0, "s: %s g: %s d: %s ", inet_fmt(qry->tr_src, s1), inet_fmt(group, s2), inet_fmt(qry->tr_dst, s3)); log(LOG_DEBUG, 0, "rttl: %d rd: %s", qry->tr_rttl, inet_fmt(qry->tr_raddr, s1)); log(LOG_DEBUG, 0, "rcount:%d, qid:%06x", rcount, qry->tr_qid); + } /* determine the routing table entry for this traceroute */ rt = determine_route(qry->tr_src); + IF_DEBUG(DEBUG_TRACE) if (rt) { log(LOG_DEBUG, 0, "rt parent vif: %d rtr: %s metric: %d", rt->rt_parent, inet_fmt(rt->rt_gateway, s1), rt->rt_metric); log(LOG_DEBUG, 0, "rt origin %s", - inet_fmts(rt->rt_origin, rt->rt_originmask, s1)); + RT_FMT(rt, s1)); } else log(LOG_DEBUG, 0, "...no route"); @@ -2100,11 +2396,13 @@ accept_mtrace(src, dst, group, data, no, datalen) * only get N copies, N <= the number of interfaces on the router, * it is not fatal. */ + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "ignoring duplicate traceroute packet"); return; } if (rt == NULL) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Mcast traceroute: no route entry %s", inet_fmt(qry->tr_src, s1)); if (IN_MULTICAST(ntohl(dst))) @@ -2114,12 +2412,14 @@ accept_mtrace(src, dst, group, data, no, datalen) if (vifi == NO_VIF) { /* The traceroute destination is not on one of my subnet vifs. */ + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Destination %s not an interface", inet_fmt(qry->tr_dst, s1)); if (IN_MULTICAST(ntohl(dst))) return; errcode = TR_WRONG_IF; } else if (rt != NULL && !VIFM_ISSET(vifi, rt->rt_children)) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Destination %s not on forwarding tree for src %s", inet_fmt(qry->tr_dst, s1), inet_fmt(qry->tr_src, s2)); if (IN_MULTICAST(ntohl(dst))) @@ -2134,6 +2434,7 @@ accept_mtrace(src, dst, group, data, no, datalen) * a tunnel or came from a directly attached mrouter. */ if ((vifi = find_vif(src, dst)) == NO_VIF) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Wrong interface for packet"); errcode = TR_WRONG_IF; } @@ -2142,6 +2443,7 @@ accept_mtrace(src, dst, group, data, no, datalen) /* Now that we've decided to send a response, save the qid */ oqid = qry->tr_qid; + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Sending traceroute response"); /* copy the packet to the sending buffer */ @@ -2173,21 +2475,18 @@ accept_mtrace(src, dst, group, data, no, datalen) ((tp.tv_usec << 10) / 15625)); resp->tr_rproto = PROTO_DVMRP; - if (errcode != TR_NO_ERR) { - resp->tr_rflags = errcode; - rt = NULL; /* hack to enforce send straight to requestor */ - goto sendit; - } - resp->tr_outaddr = uvifs[vifi].uv_lcl_addr; - resp->tr_fttl = uvifs[vifi].uv_threshold; - resp->tr_rflags = TR_NO_ERR; + resp->tr_outaddr = (vifi == NO_VIF) ? dst : uvifs[vifi].uv_lcl_addr; + resp->tr_fttl = (vifi == NO_VIF) ? 0 : uvifs[vifi].uv_threshold; + resp->tr_rflags = errcode; /* * obtain # of packets out on interface */ v_req.vifi = vifi; - if (ioctl(udp_socket, SIOCGETVIFCNT, (char *)&v_req) >= 0) + if (vifi != NO_VIF && ioctl(udp_socket, SIOCGETVIFCNT, (char *)&v_req) >= 0) resp->tr_vifout = htonl(v_req.ocount); + else + resp->tr_vifout = 0xffffffff; /* * fill in scoping & pruning information @@ -2201,23 +2500,34 @@ accept_mtrace(src, dst, group, data, no, datalen) gt = NULL; if (gt && gt->gt_mcastgrp == group) { + struct stable *st; + + for (st = gt->gt_srctbl; st; st = st->st_next) + if (qry->tr_src == st->st_origin) + break; + sg_req.src.s_addr = qry->tr_src; sg_req.grp.s_addr = group; - if (ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) >= 0) - resp->tr_pktcnt = htonl(sg_req.pktcnt); + if (st && st->st_ctime != 0 && + ioctl(udp_socket, SIOCGETSGCNT, (char *)&sg_req) >= 0) + resp->tr_pktcnt = htonl(sg_req.pktcnt + st->st_savpkt); + else + resp->tr_pktcnt = htonl(st ? st->st_savpkt : 0xffffffff); if (VIFM_ISSET(vifi, gt->gt_scope)) resp->tr_rflags = TR_SCOPED; else if (gt->gt_prsent_timer) resp->tr_rflags = TR_PRUNED; else if (!VIFM_ISSET(vifi, gt->gt_grpmems)) - if (VIFM_ISSET(vifi, rt->rt_children) && - !VIFM_ISSET(vifi, rt->rt_leaves)) + if (!NBRM_ISEMPTY(uvifs[vifi].uv_nbrmap) && + SUBS_ARE_PRUNED(rt->rt_subordinates, + uvifs[vifi].uv_nbrmap, gt->gt_prunes)) resp->tr_rflags = TR_OPRUNED; else resp->tr_rflags = TR_NO_FWD; } else { - if (scoped_addr(vifi, group)) + if ((vifi != NO_VIF && scoped_addr(vifi, group)) || + (rt && scoped_addr(rt->rt_parent, group))) resp->tr_rflags = TR_SCOPED; else if (rt && !VIFM_ISSET(vifi, rt->rt_children)) resp->tr_rflags = TR_NO_FWD; @@ -2236,12 +2546,15 @@ accept_mtrace(src, dst, group, data, no, datalen) v_req.vifi = rt->rt_parent; if (ioctl(udp_socket, SIOCGETVIFCNT, (char *)&v_req) >= 0) resp->tr_vifin = htonl(v_req.icount); + else + resp->tr_vifin = 0xffffffff; MASK_TO_VAL(rt->rt_originmask, resp->tr_smask); src = uvifs[rt->rt_parent].uv_lcl_addr; resp->tr_inaddr = src; resp->tr_rmtaddr = rt->rt_gateway; if (!VIFM_ISSET(vifi, rt->rt_children)) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Destination %s not on forwarding tree for src %s", inet_fmt(qry->tr_dst, s1), inet_fmt(qry->tr_src, s2)); resp->tr_rflags = TR_WRONG_IF; @@ -2259,6 +2572,7 @@ sendit: * else send to upstream router. If the upstream router can't handle * mtrace, set an error code and send to requestor anyway. */ + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "rcount:%d, no:%d", rcount, no); if ((rcount + 1 == no) || (rt == NULL) || (rt->rt_metric == 1)) { @@ -2280,6 +2594,7 @@ sendit: * If we don't have one, we can't source any multicasts anyway. */ if (phys_vif != -1) { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Sending reply to %s from %s", inet_fmt(dst, s1), inet_fmt(uvifs[phys_vif].uv_lcl_addr, s2)); k_set_ttl(qry->tr_rttl); @@ -2291,6 +2606,7 @@ sendit: log(LOG_INFO, 0, "No enabled phyints -- %s", "dropping traceroute reply"); } else { + IF_DEBUG(DEBUG_TRACE) log(LOG_DEBUG, 0, "Sending %s to %s from %s", resptype == IGMP_MTRACE_RESP ? "reply" : "request on", inet_fmt(dst, s1), inet_fmt(src, s2)); |