something in the eurobsdcon route-a-thon broke my simple home network

router so back out the routing stuff to pre-eurobsdcon where my machine
doesn't crash immediately.

i am happy to test diffs and report success/failures but i am not happy
to have instantaneous crashes when i reboot with a new kernel that was
compiled from pristine sources.

if you are going to be an elitist asshole then you could at least make
sure your code works.

ok and "be crass towards them" deraadt@

1 files changed, 63 insertions, 13 deletions

diff --git a/sys/net/if.c b/sys/net/if.c
index 3aa81a74595..cf75acdaef6 100644
--- a/sys/net/if.c
+++ b/sys/net/if.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: if.c,v 1.139 2005/11/27 16:22:45 henning Exp $	*/
+/*	$OpenBSD: if.c,v 1.140 2005/11/29 02:59:42 jolan Exp $	*/
 /*	$NetBSD: if.c,v 1.35 1996/05/07 05:26:04 thorpej Exp $	*/
 
 /*
@@ -126,6 +126,7 @@
 
 void	if_attachsetup(struct ifnet *);
 void	if_attachdomain1(struct ifnet *);
+int	if_detach_rtdelete(struct radix_node *, void *);
 
 int	ifqmaxlen = IFQ_MAXLEN;
 
@@ -448,6 +449,37 @@ if_attach(struct ifnet *ifp)
 }
 
 /*
+ * Delete a route if it has a specific interface for output.
+ * This function complies to the rn_walktree callback API.
+ *
+ * Note that deleting a RTF_CLONING route can trigger the
+ * deletion of more entries, so we need to cancel the walk
+ * and return EAGAIN.  The caller should restart the walk
+ * as long as EAGAIN is returned.
+ */
+int
+if_detach_rtdelete(struct radix_node *rn, void *vifp)
+{
+	struct ifnet *ifp = vifp;
+	struct rtentry *rt = (struct rtentry *)rn;
+
+	if (rt->rt_ifp == ifp) {
+		int cloning = (rt->rt_flags & RTF_CLONING);
+
+		if (rtrequest(RTM_DELETE, rt_key(rt), rt->rt_gateway,
+		    rt_mask(rt), 0, NULL) == 0 && cloning)
+			return (EAGAIN);
+	}
+
+	/*
+	 * XXX There should be no need to check for rt_ifa belonging to this
+	 * interface, because then rt_ifp is set, right?
+	 */
+
+	return (0);
+}
+
+/*
  * Detach an interface from everything in the kernel.  Also deallocate
  * private resources.
  * XXX So far only the INET protocol family has been looked over
@@ -458,11 +490,10 @@ if_detach(struct ifnet *ifp)
 {
 	struct ifaddr *ifa;
 	struct ifg_list *ifg;
-	int s;
+	int i, s = splimp();
+	struct radix_node_head *rnh;
 	struct domain *dp;
 
-	s = splimp();
-
 	ifp->if_flags &= ~IFF_OACTIVE;
 	ifp->if_start = if_detached_start;
 	ifp->if_ioctl = if_detached_ioctl;
@@ -498,7 +529,17 @@ if_detach(struct ifnet *ifp)
 		altq_detach(&ifp->if_snd);
 #endif
 
-	rt_if_remove(ifp);
+	/*
+	 * Find and remove all routes which is using this interface.
+	 * XXX Factor out into a route.c function?
+	 */
+	for (i = 1; i <= AF_MAX; i++) {
+		rnh = rt_tables[i];
+		if (rnh)
+			while ((*rnh->rnh_walktree)(rnh,
+			    if_detach_rtdelete, ifp) == EAGAIN)
+				;
+	}
 
 #ifdef INET
 	rti_delete(ifp);
@@ -1750,6 +1791,7 @@ if_group_egress_build(void)
 #ifdef INET6
 	struct sockaddr_in6	 sa_in6;
 #endif
+	struct radix_node_head	*rnh;
 	struct radix_node	*rn;
 	struct rtentry		*rt;
 
@@ -1763,34 +1805,42 @@ if_group_egress_build(void)
 			if_delgroup(ifgm->ifgm_ifp, IFG_EGRESS);
 		}
 
+	if ((rnh = rt_tables[AF_INET]) == NULL)
+		return (-1);
+
 	bzero(&sa_in, sizeof(sa_in));
 	sa_in.sin_len = sizeof(sa_in);
 	sa_in.sin_family = AF_INET;
-	if ((rn = rt_lookup(sintosa(&sa_in), sintosa(&sa_in), 0)) != NULL) {
+	if ((rn = rnh->rnh_lookup(&sa_in, &sa_in, rnh))) {
 		do {
 			rt = (struct rtentry *)rn;
 			if (rt->rt_ifp)
 				if_addgroup(rt->rt_ifp, IFG_EGRESS);
 #ifndef SMALL_KERNEL
-			rn = rn_mpath_next(rn);
-#else
-			rn = NULL;
+			if (rn_mpath_capable(rnh))
+				rn = rn_mpath_next(rn);
+			else
 #endif
+				rn = NULL;
 		} while (rn != NULL);
 	}
 
 #ifdef INET6
+	if ((rnh = rt_tables[AF_INET6]) == NULL)
+		return (-1);
+
 	bcopy(&sa6_any, &sa_in6, sizeof(sa_in6));
-	if ((rn = rt_lookup(sin6tosa(&sa_in6), sin6tosa(&sa_in6), 0)) != NULL) {
+	if ((rn = rnh->rnh_lookup(&sa_in6, &sa_in6, rnh))) {
 		do {
 			rt = (struct rtentry *)rn;
 			if (rt->rt_ifp)
 				if_addgroup(rt->rt_ifp, IFG_EGRESS);
 #ifndef SMALL_KERNEL
-			rn = rn_mpath_next(rn);
-#else
-			rn = NULL;
+			if (rn_mpath_capable(rnh))
+				rn = rn_mpath_next(rn);
+			else
 #endif
+				rn = NULL;
 		} while (rn != NULL);
 	}
 #endif