Asynchronous feeding of protocols.

5 files changed, 101 insertions, 39 deletions

diff --git a/TODO b/TODO
index f2380b34..ed32e6b0 100644
--- a/TODO
+++ b/TODO
@@ -5,10 +5,8 @@ Core
 - client: data losses on large dumps?
 
 - table: ocassional core dumps in rt_prune()
-- table: do feeding by parts
 
-- bgp: timing of updates?
-- bgp: dump and get_route_info
+- bgp: get_route_info
 
 Documentation
 ~~~~~~~~~~~~~
@@ -33,6 +31,7 @@ Globals
 Various ideas
 ~~~~~~~~~~~~~
 - static: allow specifying a per-route filter program for setting route attributes?
+- bgp: timing of updates?
 - netlink: realms
 - netlink: import Linux route attributes to our rta's, so that they can be filtered?
 - config: executable config files
diff --git a/nest/proto.c b/nest/proto.c
index e81aadc9..76467aa5 100644
--- a/nest/proto.c
+++ b/nest/proto.c
@@ -410,6 +410,25 @@ proto_fell_down(struct proto *p)
 }
 
 static void
+proto_feed_more(void *P)
+{
+  struct proto *p = P;
+
+  DBG("Feeding protocol %s continued\n", p->name);
+  if (rt_feed_baby(p))
+    {
+      p->core_state = FS_HAPPY;
+      proto_relink(p);
+      DBG("Protocol %s up and running\n", p->name);
+    }
+  else
+    {
+      p->attn->hook = proto_feed_more;
+      ev_schedule(p->attn);		/* Will continue later... */
+    }
+}
+
+static void
 proto_feed(void *P)
 {
   struct proto *p = P;
@@ -417,10 +436,7 @@ proto_feed(void *P)
   DBG("Feeding protocol %s\n", p->name);
   proto_add_announce_hook(p, p->table);
   if_feed_baby(p);
-  rt_feed_baby(p);
-  p->core_state = FS_HAPPY;
-  proto_relink(p);
-  DBG("Protocol %s up and running\n", p->name);
+  proto_feed_more(P);
 }
 
 void
@@ -444,8 +460,12 @@ proto_notify_state(struct proto *p, unsigned ps)
 	}
       else if (cs == FS_FLUSHING)	/* Still flushing... */
 	;
-      else				/* Need to start flushing */
-	goto schedule_flush;
+      else
+	{
+	  if (cs == FS_FEEDING)		/* Need to abort feeding */
+	    rt_feed_baby_abort(p);
+	  goto schedule_flush;		/* Need to start flushing */
+	}
       break;
     case PS_START:
       ASSERT(ops == PS_DOWN);
diff --git a/nest/protocol.h b/nest/protocol.h
index df9064b3..64ea4661 100644
--- a/nest/protocol.h
+++ b/nest/protocol.h
@@ -151,6 +151,9 @@ struct proto {
   struct filter *out_filter;		/* Output filter */
   struct announce_hook *ahooks;		/* Announcement hooks for this protocol */
 
+  struct fib_iterator *feed_iterator;	/* Routing table iterator used during protocol feeding */
+  struct announce_hook *feed_ahook;	/* Announce hook we currently feed */
+
   /* Hic sunt protocol-specific data */
 };
 
diff --git a/nest/route.h b/nest/route.h
index 20709e94..183e80b5 100644
--- a/nest/route.h
+++ b/nest/route.h
@@ -197,7 +197,8 @@ rte *rte_do_cow(rte *);
 static inline rte * rte_cow(rte *r) { return (r->flags & REF_COW) ? rte_do_cow(r) : r; }
 void rt_dump(rtable *);
 void rt_dump_all(void);
-void rt_feed_baby(struct proto *p);
+int rt_feed_baby(struct proto *p);
+void rt_feed_baby_abort(struct proto *p);
 void rt_prune(rtable *tab);
 void rt_prune_all(void);
 struct rtable_config *rt_new_table(struct symbol *s);
diff --git a/nest/rt-table.c b/nest/rt-table.c
index 7133ed49..d81dd857 100644
--- a/nest/rt-table.c
+++ b/nest/rt-table.c
@@ -200,35 +200,6 @@ rte_announce(rtable *tab, net *net, rte *new, rte *old, ea_list *tmpa)
     }
 }
 
-void
-rt_feed_baby(struct proto *p)
-{
-  struct announce_hook *h;
-
-  if (!p->ahooks)
-    return;
-  DBG("Announcing routes to new protocol %s\n", p->name);
-  for(h=p->ahooks; h; h=h->next)
-    {
-      rtable *t = h->table;
-      FIB_WALK(&t->fib, fn)
-	{
-	  net *n = (net *) fn;
-	  rte *e;
-	  for(e=n->routes; e; e=e->next)
-	    {
-	      struct proto *q = e->attrs->proto;
-	      ea_list *tmpa = q->make_tmp_attrs ? q->make_tmp_attrs(e, rte_update_pool) : NULL;
-	      do_rte_announce(h, n, e, NULL, tmpa, ipa_classify(n->n.prefix));
-	      lp_flush(rte_update_pool);
-	      if (p->core_state != FS_FEEDING)
-		return;  /* In the meantime, the protocol fell down. */
-	    }
-	}
-      FIB_WALK_END;
-    }
-}
-
 static inline int
 rte_validate(rte *e)
 {
@@ -683,6 +654,74 @@ rt_commit(struct config *new, struct config *old)
   DBG("\tdone\n");
 }
 
+int
+rt_feed_baby(struct proto *p)
+{
+  struct announce_hook *h;
+  struct fib_iterator *fit;
+  int max_feed = 2;			/* FIXME */
+
+  if (!p->feed_ahook)			/* Need to initialize first */
+    {
+      if (!p->ahooks)
+	return 1;
+      DBG("Announcing routes to new protocol %s\n", p->name);
+      p->feed_ahook = p->ahooks;
+      fit = p->feed_iterator = mb_alloc(p->pool, sizeof(struct fib_iterator));
+      goto next_hook;
+    }
+  fit = p->feed_iterator;
+
+again:
+  h = p->feed_ahook;
+  FIB_ITERATE_START(&h->table->fib, fit, fn)
+    {
+      net *n = (net *) fn;
+      rte *e;
+      for(e=n->routes; e; e=e->next)
+	{
+	  struct proto *q = e->attrs->proto;
+	  ea_list *tmpa;
+
+	  if (p->core_state != FS_FEEDING)
+	    return 1;  /* In the meantime, the protocol fell down. */
+	  rte_update_lock();
+	  tmpa = q->make_tmp_attrs ? q->make_tmp_attrs(e, rte_update_pool) : NULL;
+	  do_rte_announce(h, n, e, NULL, tmpa, ipa_classify(n->n.prefix));
+	  rte_update_unlock();
+	  if (!--max_feed)
+	    {
+	      FIB_ITERATE_PUT(fit, fn);
+	      return 0;
+	    }
+	}
+    }
+  FIB_ITERATE_END(fn);
+  p->feed_ahook = h->next;
+  if (!p->feed_ahook)
+    {
+      mb_free(p->feed_iterator);
+      p->feed_iterator = NULL;
+      return 1;
+    }
+
+next_hook:
+  h = p->feed_ahook;
+  FIB_ITERATE_INIT(fit, &h->table->fib);
+  goto again;
+}
+
+void
+rt_feed_baby_abort(struct proto *p)
+{
+  if (p->feed_ahook)
+    {
+      /* Unlink the iterator and exit */
+      fit_get(&p->feed_ahook->table->fib, p->feed_iterator);
+      p->feed_ahook = NULL;
+    }
+}
+
 /*
  *  CLI commands
  */