From 6892d3292121d02900ac9968e832353ecacca4ad Mon Sep 17 00:00:00 2001
From: Nick Mathewson <nickm@torproject.org>
Date: Thu, 8 Feb 2018 12:14:42 -0500
Subject: [PATCH] Add an address_set to the nodelist.

This set is rebuilt whenever a consensus arrives.  In between
consensuses, it is add-only.
---
 src/or/nodelist.c | 66 +++++++++++++++++++++++++++++++++++++++++++++++
 src/or/nodelist.h |  1 +
 2 files changed, 67 insertions(+)

diff --git a/src/or/nodelist.c b/src/or/nodelist.c
index 0e9a651818..c2080db12c 100644
--- a/src/or/nodelist.c
+++ b/src/or/nodelist.c
@@ -14,6 +14,7 @@
 
 #include "or.h"
 #include "address.h"
+#include "address_set.h"
 #include "config.h"
 #include "control.h"
 #include "dirserv.h"
@@ -52,6 +53,7 @@ static void count_usable_descriptors(int *num_present,
 static void update_router_have_minimum_dir_info(void);
 static double get_frac_paths_needed_for_circs(const or_options_t *options,
                                               const networkstatus_t *ns);
+static void node_add_to_address_set(const node_t *node);
 
 /** A nodelist_t holds a node_t object for every router we're "willing to use
  * for something".  Specifically, it should hold a node_t for every node that
@@ -63,6 +65,8 @@ typedef struct nodelist_t {
   /* Hash table to map from node ID digest to node. */
   HT_HEAD(nodelist_map, node_t) nodes_by_id;
 
+  /* Set of addresses that belong to nodes we believe in. */
+  address_set_t *node_addrs;
 } nodelist_t;
 
 static inline unsigned int
@@ -150,6 +154,50 @@ node_addrs_changed(node_t *node)
   node->country = -1;
 }
 
+/** Add all address information about <b>node</b> to the current address
+ * set (if there is one).
+ */
+static void
+node_add_to_address_set(const node_t *node)
+{
+  if (!the_nodelist || !the_nodelist->node_addrs)
+    return;
+
+  /* These various address sources can be redundant, but it's likely faster
+   * to add them all than to compare them all for equality. */
+
+  if (node->rs) {
+    if (node->rs->addr)
+      address_set_add_ipv4h(the_nodelist->node_addrs, node->rs->addr);
+    if (!tor_addr_is_null(&node->rs->ipv6_addr))
+      address_set_add(the_nodelist->node_addrs, &node->rs->ipv6_addr);
+  }
+  if (node->ri) {
+    if (node->ri->addr)
+      address_set_add_ipv4h(the_nodelist->node_addrs, node->ri->addr);
+    if (!tor_addr_is_null(&node->ri->ipv6_addr))
+      address_set_add(the_nodelist->node_addrs, &node->ri->ipv6_addr);
+  }
+  if (node->md) {
+    if (!tor_addr_is_null(&node->md->ipv6_addr))
+      address_set_add(the_nodelist->node_addrs, &node->md->ipv6_addr);
+  }
+}
+
+/** Return true if <b>addr</b> is the address of some node in the nodelist.
+ * If not, probably return false. */
+int
+nodelist_probably_contains_address(const tor_addr_t *addr)
+{
+  if (BUG(!addr))
+    return 0;
+
+  if (!the_nodelist || !the_nodelist->node_addrs)
+    return 0;
+
+  return address_set_probably_contains(the_nodelist->node_addrs, addr);
+}
+
 /** Add <b>ri</b> to an appropriate node in the nodelist.  If we replace an
  * old routerinfo, and <b>ri_old_out</b> is not NULL, set *<b>ri_old_out</b>
  * to the previous routerinfo.
@@ -188,6 +236,8 @@ nodelist_set_routerinfo(routerinfo_t *ri, routerinfo_t **ri_old_out)
     dirserv_set_node_flags_from_authoritative_status(node, status);
   }
 
+  node_add_to_address_set(node);
+
   return node;
 }
 
@@ -219,6 +269,9 @@ nodelist_add_microdesc(microdesc_t *md)
     node->md = md;
     md->held_by_nodes++;
   }
+
+  node_add_to_address_set(node);
+
   return node;
 }
 
@@ -240,6 +293,11 @@ nodelist_set_consensus(networkstatus_t *ns)
   SMARTLIST_FOREACH(the_nodelist->nodes, node_t *, node,
                     node->rs = NULL);
 
+  /* Conservatively estimate that every node will have 2 addresses. */
+  const int estimated_addresses = smartlist_len(ns->routerstatus_list) * 2;
+  address_set_free(the_nodelist->node_addrs);
+  the_nodelist->node_addrs = address_set_new(estimated_addresses);
+
   SMARTLIST_FOREACH_BEGIN(ns->routerstatus_list, routerstatus_t *, rs) {
     node_t *node = node_get_or_create(rs->identity_digest);
     node->rs = rs;
@@ -278,6 +336,11 @@ nodelist_set_consensus(networkstatus_t *ns)
 
   nodelist_purge();
 
+  /* Now add all the nodes we have to the address set. */
+  SMARTLIST_FOREACH_BEGIN(the_nodelist->nodes, node_t *, node) {
+    node_add_to_address_set(node);
+  } SMARTLIST_FOREACH_END(node);
+
   if (! authdir) {
     SMARTLIST_FOREACH_BEGIN(the_nodelist->nodes, node_t *, node) {
       /* We have no routerstatus for this router. Clear flags so we can skip
@@ -430,6 +493,9 @@ nodelist_free_all(void)
 
   smartlist_free(the_nodelist->nodes);
 
+  address_set_free(the_nodelist->node_addrs);
+  the_nodelist->node_addrs = NULL;
+
   tor_free(the_nodelist);
 }
 
diff --git a/src/or/nodelist.h b/src/or/nodelist.h
index 71a91e107f..355057f398 100644
--- a/src/or/nodelist.h
+++ b/src/or/nodelist.h
@@ -22,6 +22,7 @@ const node_t *node_get_by_hex_id(const char *identity_digest);
 node_t *nodelist_set_routerinfo(routerinfo_t *ri, routerinfo_t **ri_old_out);
 node_t *nodelist_add_microdesc(microdesc_t *md);
 void nodelist_set_consensus(networkstatus_t *ns);
+int nodelist_probably_contains_address(const tor_addr_t *addr);
 
 void nodelist_remove_microdesc(const char *identity_digest, microdesc_t *md);
 void nodelist_remove_routerinfo(routerinfo_t *ri);