Merge branch 'split_stats'

46 files changed, 1150 insertions, 901 deletions

diff --git a/src/app/config/config.c b/src/app/config/config.c
index 91e2758650..d44b17ce65 100644
--- a/src/app/config/config.c
+++ b/src/app/config/config.c
@@ -96,13 +96,16 @@
 #include "feature/nodelist/networkstatus.h"
 #include "feature/nodelist/nickname.h"
 #include "feature/nodelist/nodelist.h"
+#include "feature/nodelist/routerlist.h"
 #include "feature/nodelist/routerset.h"
 #include "feature/relay/dns.h"
 #include "feature/relay/ext_orport.h"
 #include "feature/relay/routermode.h"
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "lib/compress/compress.h"
 #include "lib/crypt_ops/crypto_init.h"
@@ -8346,6 +8349,11 @@ config_load_geoip_file_(sa_family_t family,
                         const char *fname,
                         const char *default_fname)
 {
+  const or_options_t *options = get_options();
+  const char *msg = "";
+  int severity = options_need_geoip_info(options, &msg) ? LOG_WARN : LOG_INFO;
+  int r;
+
 #ifdef _WIN32
   char *free_fname = NULL; /* Used to hold any temporary-allocated value */
   /* XXXX Don't use this "<default>" junk; make our filename options
@@ -8355,12 +8363,16 @@ config_load_geoip_file_(sa_family_t family,
     tor_asprintf(&free_fname, "%s\\%s", conf_root, default_fname);
     fname = free_fname;
   }
-  geoip_load_file(family, fname);
+  r = geoip_load_file(family, fname, severity);
   tor_free(free_fname);
 #else /* !(defined(_WIN32)) */
   (void)default_fname;
-  geoip_load_file(family, fname);
+  r = geoip_load_file(family, fname, severity);
 #endif /* defined(_WIN32) */
+
+  if (r < 0 && severity == LOG_WARN) {
+    log_warn(LD_GENERAL, "%s", msg);
+  }
 }
 
 /** Load geoip files for IPv4 and IPv6 if <a>options</a> and
@@ -8374,13 +8386,19 @@ config_maybe_load_geoip_files_(const or_options_t *options,
   if (options->GeoIPFile &&
       ((!old_options || !opt_streq(old_options->GeoIPFile,
                                    options->GeoIPFile))
-       || !geoip_is_loaded(AF_INET)))
+       || !geoip_is_loaded(AF_INET))) {
     config_load_geoip_file_(AF_INET, options->GeoIPFile, "geoip");
+    /* Okay, now we need to maybe change our mind about what is in
+     * which country. We do this for IPv4 only since that's what we
+     * store in node->country. */
+    refresh_all_country_info();
+  }
   if (options->GeoIPv6File &&
       ((!old_options || !opt_streq(old_options->GeoIPv6File,
                                    options->GeoIPv6File))
-       || !geoip_is_loaded(AF_INET6)))
+       || !geoip_is_loaded(AF_INET6))) {
     config_load_geoip_file_(AF_INET6, options->GeoIPv6File, "geoip6");
+  }
 }
 
 /** Initialize cookie authentication (used so far by the ControlPort
diff --git a/src/app/main/main.c b/src/app/main/main.c
index 90bc8c510f..8a24df2105 100644
--- a/src/app/main/main.c
+++ b/src/app/main/main.c
@@ -27,6 +27,7 @@
 #include "core/or/command.h"
 #include "core/or/connection_edge.h"
 #include "core/or/connection_or.h"
+#include "core/or/dos.h"
 #include "core/or/policies.h"
 #include "core/or/protover.h"
 #include "core/or/relay.h"
@@ -60,7 +61,9 @@
 #include "feature/rend/rendcache.h"
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "lib/compress/compress.h"
 #include "lib/container/buffers.h"
@@ -754,6 +757,7 @@ tor_free_all(int postfork)
     evdns_shutdown(1);
   }
   geoip_free_all();
+  geoip_stats_free_all();
   dirvote_free_all();
   routerlist_free_all();
   networkstatus_free_all();
diff --git a/src/core/include.am b/src/core/include.am
index 64a593ca51..e51819fa20 100644
--- a/src/core/include.am
+++ b/src/core/include.am
@@ -60,6 +60,7 @@ LIBTOR_APP_A_SOURCES = 				\
 	src/feature/client/transports.c		\
 	src/feature/control/control.c		\
 	src/feature/control/fmt_serverstatus.c  \
+	src/feature/control/getinfo_geoip.c	\
 	src/feature/dirauth/keypin.c		\
 	src/feature/dircache/conscache.c	\
 	src/feature/dircache/consdiffmgr.c	\
@@ -116,8 +117,9 @@ LIBTOR_APP_A_SOURCES = 				\
 	src/feature/rend/rendcommon.c		\
 	src/feature/rend/rendmid.c		\
 	src/feature/rend/rendservice.c		\
-	src/feature/stats/geoip.c		\
-	src/feature/stats/rephist.c
+	src/feature/stats/geoip_stats.c		\
+	src/feature/stats/rephist.c		\
+	src/feature/stats/predict_ports.c
 
 # These should eventually move into module_dirauth_sources, but for now
 # the separation is only in the code location.
@@ -253,6 +255,7 @@ noinst_HEADERS +=					\
 	src/feature/control/control.h			\
 	src/feature/control/control_connection_st.h	\
 	src/feature/control/fmt_serverstatus.h		\
+	src/feature/control/getinfo_geoip.h		\
 	src/feature/dirauth/authmode.h			\
 	src/feature/dirauth/bwauth.h			\
 	src/feature/dirauth/dircollate.h		\
@@ -346,8 +349,9 @@ noinst_HEADERS +=					\
 	src/feature/rend/rendcommon.h			\
 	src/feature/rend/rendmid.h			\
 	src/feature/rend/rendservice.h			\
-	src/feature/stats/geoip.h			\
-	src/feature/stats/rephist.h
+	src/feature/stats/geoip_stats.h			\
+	src/feature/stats/rephist.h			\
+	src/feature/stats/predict_ports.h
 
 noinst_HEADERS +=			\
 	src/app/config/auth_dirs.inc	\
diff --git a/src/core/mainloop/connection.c b/src/core/mainloop/connection.c
index 18863fc04f..d8326a5be8 100644
--- a/src/core/mainloop/connection.c
+++ b/src/core/mainloop/connection.c
@@ -102,7 +102,7 @@
 #include "feature/relay/routermode.h"
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "feature/stats/rephist.h"
 #include "lib/crypt_ops/crypto_util.h"
 
diff --git a/src/core/mainloop/mainloop.c b/src/core/mainloop/mainloop.c
index 6e7033ec4b..6c305c2b47 100644
--- a/src/core/mainloop/mainloop.c
+++ b/src/core/mainloop/mainloop.c
@@ -92,7 +92,8 @@
 #include "feature/relay/selftest.h"
 #include "feature/rend/rendcache.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/geoip.h"
+#include "feature/stats/geoip_stats.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "lib/container/buffers.h"
 #include "lib/crypt_ops/crypto_rand.h"
diff --git a/src/core/or/channel.c b/src/core/or/channel.c
index e9011520ae..fe3785d52e 100644
--- a/src/core/or/channel.c
+++ b/src/core/or/channel.c
@@ -58,28 +58,29 @@
 #define CHANNEL_PRIVATE_
 
 #include "core/or/or.h"
+#include "app/config/config.h"
+#include "core/mainloop/mainloop.h"
 #include "core/or/channel.h"
-#include "core/or/channeltls.h"
 #include "core/or/channelpadding.h"
+#include "core/or/channeltls.h"
 #include "core/or/circuitbuild.h"
 #include "core/or/circuitlist.h"
+#include "core/or/circuitmux.h"
 #include "core/or/circuitstats.h"
-#include "app/config/config.h"
 #include "core/or/connection_or.h" /* For var_cell_free() */
-#include "core/or/circuitmux.h"
-#include "feature/client/entrynodes.h"
-#include "feature/stats/geoip.h"
-#include "core/mainloop/mainloop.h"
-#include "feature/nodelist/nodelist.h"
+#include "core/or/dos.h"
 #include "core/or/relay.h"
-#include "feature/stats/rephist.h"
-#include "feature/relay/router.h"
-#include "feature/nodelist/routerlist.h"
 #include "core/or/scheduler.h"
-#include "lib/time/compat_time.h"
+#include "feature/client/entrynodes.h"
 #include "feature/nodelist/networkstatus.h"
+#include "feature/nodelist/nodelist.h"
+#include "feature/nodelist/routerlist.h"
+#include "feature/relay/router.h"
 #include "feature/rend/rendservice.h"
+#include "feature/stats/geoip_stats.h"
+#include "feature/stats/rephist.h"
 #include "lib/evloop/timers.h"
+#include "lib/time/compat_time.h"
 
 #include "core/or/cell_queue_st.h"
 
diff --git a/src/core/or/circuitbuild.c b/src/core/or/circuitbuild.c
index 26f0fc4d4f..822129d80e 100644
--- a/src/core/or/circuitbuild.c
+++ b/src/core/or/circuitbuild.c
@@ -67,7 +67,7 @@
 #include "feature/relay/routermode.h"
 #include "feature/relay/selftest.h"
 #include "feature/rend/rendcommon.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/predict_ports.h"
 #include "lib/crypt_ops/crypto_rand.h"
 
 #include "core/or/cell_st.h"
diff --git a/src/core/or/circuitlist.c b/src/core/or/circuitlist.c
index c07c8e97a0..5ff142c15c 100644
--- a/src/core/or/circuitlist.c
+++ b/src/core/or/circuitlist.c
@@ -85,6 +85,7 @@
 #include "core/crypto/relay_crypto.h"
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "feature/nodelist/routerlist.h"
 #include "feature/nodelist/routerset.h"
diff --git a/src/core/or/circuituse.c b/src/core/or/circuituse.c
index d035d66ea2..088358a4d6 100644
--- a/src/core/or/circuituse.c
+++ b/src/core/or/circuituse.c
@@ -57,7 +57,7 @@
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/predict_ports.h"
 #include "lib/math/fp.h"
 #include "lib/time/tvdiff.h"
 
diff --git a/src/core/or/connection_edge.c b/src/core/or/connection_edge.c
index d8b45fe350..627a2f1a1c 100644
--- a/src/core/or/connection_edge.c
+++ b/src/core/or/connection_edge.c
@@ -95,6 +95,7 @@
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
 #include "feature/rend/rendservice.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "lib/container/buffers.h"
 #include "lib/crypt_ops/crypto_util.h"
diff --git a/src/core/or/connection_or.c b/src/core/or/connection_or.c
index 4a724a24d3..65f4e28c92 100644
--- a/src/core/or/connection_or.c
+++ b/src/core/or/connection_or.c
@@ -43,7 +43,7 @@
 #include "lib/crypt_ops/crypto_util.h"
 #include "feature/dirauth/reachability.h"
 #include "feature/client/entrynodes.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "core/mainloop/mainloop.h"
 #include "trunnel/link_handshake.h"
 #include "feature/nodelist/microdesc.h"
diff --git a/src/core/or/dos.c b/src/core/or/dos.c
index c59e2f6361..4303b8fd59 100644
--- a/src/core/or/dos.c
+++ b/src/core/or/dos.c
@@ -9,17 +9,17 @@
 #define DOS_PRIVATE
 
 #include "core/or/or.h"
-#include "core/or/channel.h"
 #include "app/config/config.h"
 #include "core/mainloop/connection.h"
-#include "core/or/connection_or.h"
-#include "lib/crypt_ops/crypto_rand.h"
-#include "feature/stats/geoip.h"
 #include "core/mainloop/mainloop.h"
+#include "core/or/channel.h"
+#include "core/or/connection_or.h"
+#include "core/or/relay.h"
 #include "feature/nodelist/networkstatus.h"
 #include "feature/nodelist/nodelist.h"
-#include "core/or/relay.h"
 #include "feature/relay/routermode.h"
+#include "feature/stats/geoip_stats.h"
+#include "lib/crypt_ops/crypto_rand.h"
 
 #include "core/or/dos.h"
 
diff --git a/src/core/or/or.h b/src/core/or/or.h
index 7557c1321e..acf092c8dc 100644
--- a/src/core/or/or.h
+++ b/src/core/or/or.h
@@ -41,6 +41,7 @@
 #include "lib/fs/mmap.h"
 #include "lib/fs/path.h"
 #include "lib/fs/userdb.h"
+#include "lib/geoip/country.h"
 #include "lib/intmath/addsub.h"
 #include "lib/intmath/bits.h"
 #include "lib/intmath/cmp.h"
@@ -798,9 +799,6 @@ typedef struct download_status_t download_status_t;
 
 typedef struct signed_descriptor_t signed_descriptor_t;
 
-/** A signed integer representing a country code. */
-typedef int16_t country_t;
-
 /** Flags used to summarize the declared protocol versions of a relay,
  * so we don't need to parse them again and again. */
 typedef struct protover_summary_flags_t {
diff --git a/src/core/or/policies.c b/src/core/or/policies.c
index fc359627b9..c3fded1fbd 100644
--- a/src/core/or/policies.c
+++ b/src/core/or/policies.c
@@ -27,7 +27,7 @@
 #include "feature/relay/router.h"
 #include "feature/relay/routermode.h"
 #include "feature/nodelist/routerparse.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "ht.h"
 #include "lib/encoding/confline.h"
 
diff --git a/src/core/or/relay.c b/src/core/or/relay.c
index 88bcdf0703..9e45a35246 100644
--- a/src/core/or/relay.c
+++ b/src/core/or/relay.c
@@ -65,7 +65,7 @@
 #include "lib/crypt_ops/crypto_util.h"
 #include "feature/dircommon/directory.h"
 #include "feature/relay/dns.h"
-#include "feature/stats/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "feature/hs/hs_cache.h"
 #include "core/mainloop/mainloop.h"
 #include "feature/nodelist/networkstatus.h"
diff --git a/src/core/or/status.c b/src/core/or/status.c
index d8a9da0a55..cdb6df1230 100644
--- a/src/core/or/status.c
+++ b/src/core/or/status.c
@@ -30,13 +30,13 @@
 #include "feature/hs/hs_stats.h"
 #include "feature/hs/hs_service.h"
 #include "core/or/dos.h"
+#include "feature/stats/geoip_stats.h"
 
 #include "app/config/or_state_st.h"
 #include "feature/nodelist/routerinfo_st.h"
 #include "lib/tls/tortls.h"
 
 static void log_accounting(const time_t now, const or_options_t *options);
-#include "feature/stats/geoip.h"
 
 /** Return the total number of circuits. */
 STATIC int
diff --git a/src/feature/control/control.c b/src/feature/control/control.c
index d8d08f2c86..d02ecbac80 100644
--- a/src/feature/control/control.c
+++ b/src/feature/control/control.c
@@ -60,6 +60,7 @@
 #include "feature/client/entrynodes.h"
 #include "feature/control/control.h"
 #include "feature/control/fmt_serverstatus.h"
+#include "feature/control/getinfo_geoip.h"
 #include "feature/dircache/dirserv.h"
 #include "feature/dirclient/dirclient.h"
 #include "feature/dirclient/dlstatus.h"
@@ -83,8 +84,8 @@
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/geoip.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/geoip_stats.h"
+#include "feature/stats/predict_ports.h"
 #include "lib/container/buffers.h"
 #include "lib/crypt_ops/crypto_rand.h"
 #include "lib/crypt_ops/crypto_util.h"
diff --git a/src/feature/control/getinfo_geoip.c b/src/feature/control/getinfo_geoip.c
new file mode 100644
index 0000000000..d188725fa3
--- /dev/null
+++ b/src/feature/control/getinfo_geoip.c
@@ -0,0 +1,45 @@
+
+#include "core/or/or.h"
+#include "core/mainloop/connection.h"
+#include "feature/control/control.h"
+#include "feature/control/getinfo_geoip.h"
+#include "lib/geoip/geoip.h"
+
+/** Helper used to implement GETINFO ip-to-country/... controller command. */
+int
+getinfo_helper_geoip(control_connection_t *control_conn,
+                     const char *question, char **answer,
+                     const char **errmsg)
+{
+  (void)control_conn;
+  if (!strcmpstart(question, "ip-to-country/")) {
+    int c;
+    sa_family_t family;
+    tor_addr_t addr;
+    question += strlen("ip-to-country/");
+
+    if (!strcmp(question, "ipv4-available") ||
+        !strcmp(question, "ipv6-available")) {
+      family = !strcmp(question, "ipv4-available") ? AF_INET : AF_INET6;
+      const int available = geoip_is_loaded(family);
+      tor_asprintf(answer, "%d", !! available);
+      return 0;
+    }
+
+    family = tor_addr_parse(&addr, question);
+    if (family != AF_INET && family != AF_INET6) {
+      *errmsg = "Invalid address family";
+      return -1;
+    }
+    if (!geoip_is_loaded(family)) {
+      *errmsg = "GeoIP data not loaded";
+      return -1;
+    }
+    if (family == AF_INET)
+      c = geoip_get_country_by_ipv4(tor_addr_to_ipv4h(&addr));
+    else                      /* AF_INET6 */
+      c = geoip_get_country_by_ipv6(tor_addr_to_in6(&addr));
+    *answer = tor_strdup(geoip_get_country_name(c));
+  }
+  return 0;
+}
diff --git a/src/feature/control/getinfo_geoip.h b/src/feature/control/getinfo_geoip.h
new file mode 100644
index 0000000000..ff77cefecd
--- /dev/null
+++ b/src/feature/control/getinfo_geoip.h
@@ -0,0 +1,14 @@
+/* Copyright (c) 2001 Matej Pfajfar.
+ * Copyright (c) 2001-2004, Roger Dingledine.
+ * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
+ * Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+#ifndef TOR_GETINFO_GEOIP_H
+#define TOR_GETINFO_GEOIP_H
+
+int getinfo_helper_geoip(control_connection_t *control_conn,
+                     const char *question, char **answer,
+                     const char **errmsg);
+
+#endif
diff --git a/src/feature/dircache/dircache.c b/src/feature/dircache/dircache.c
index 76bde0b787..872a88018f 100644
--- a/src/feature/dircache/dircache.c
+++ b/src/feature/dircache/dircache.c
@@ -25,7 +25,7 @@
 #include "feature/nodelist/routerlist.h"
 #include "feature/relay/routermode.h"
 #include "feature/rend/rendcache.h"
-#include "feature/stats/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "feature/stats/rephist.h"
 #include "lib/compress/compress.h"
 
diff --git a/src/feature/dircache/dirserv.c b/src/feature/dircache/dirserv.c
index 4fa3cd7f93..57178cd506 100644
--- a/src/feature/dircache/dirserv.c
+++ b/src/feature/dircache/dirserv.c
@@ -16,7 +16,7 @@
 #include "feature/nodelist/routerlist.h"
 #include "feature/relay/router.h"
 #include "feature/relay/routermode.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/predict_ports.h"
 
 #include "feature/dircache/cached_dir_st.h"
 #include "feature/dircommon/dir_connection_st.h"
diff --git a/src/feature/dirclient/dirclient.c b/src/feature/dirclient/dirclient.c
index 334012602c..f6a712e429 100644
--- a/src/feature/dirclient/dirclient.c
+++ b/src/feature/dirclient/dirclient.c
@@ -43,7 +43,7 @@
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/predict_ports.h"
 
 #include "lib/compress/compress.h"
 #include "lib/crypt_ops/crypto_format.h"
diff --git a/src/feature/dircommon/directory.c b/src/feature/dircommon/directory.c
index 8116d72c62..dff68fcf0e 100644
--- a/src/feature/dircommon/directory.c
+++ b/src/feature/dircommon/directory.c
@@ -12,7 +12,7 @@
 #include "feature/dirclient/dirclient.h"
 #include "feature/dircommon/directory.h"
 #include "feature/dircommon/fp_pair.h"
-#include "feature/stats/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "lib/compress/compress.h"
 
 #include "feature/dircommon/dir_connection_st.h"
diff --git a/src/feature/nodelist/nodelist.c b/src/feature/nodelist/nodelist.c
index e3b77d562c..90ca579b7d 100644
--- a/src/feature/nodelist/nodelist.c
+++ b/src/feature/nodelist/nodelist.c
@@ -65,7 +65,7 @@
 #include "feature/nodelist/routerset.h"
 #include "feature/nodelist/torcert.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "lib/net/address.h"
 
 #include <string.h>
diff --git a/src/feature/nodelist/routerset.c b/src/feature/nodelist/routerset.c
index 08124835ae..cb07a436d1 100644
--- a/src/feature/nodelist/routerset.c
+++ b/src/feature/nodelist/routerset.c
@@ -34,7 +34,7 @@ n * Copyright (c) 2001-2004, Roger Dingledine.
 #include "feature/nodelist/nodelist.h"
 #include "feature/nodelist/routerparse.h"
 #include "feature/nodelist/routerset.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 
 #include "core/or/addr_policy_st.h"
 #include "core/or/extend_info_st.h"
diff --git a/src/feature/relay/router.c b/src/feature/relay/router.c
index 8629ad9561..fdaadb205b 100644
--- a/src/feature/relay/router.c
+++ b/src/feature/relay/router.c
@@ -36,7 +36,8 @@
 #include "feature/relay/routerkeys.h"
 #include "feature/relay/routermode.h"
 #include "feature/relay/selftest.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "feature/stats/rephist.h"
 #include "lib/crypt_ops/crypto_ed25519.h"
 #include "lib/crypt_ops/crypto_format.h"
diff --git a/src/feature/rend/rendservice.c b/src/feature/rend/rendservice.c
index 8257919ac4..057ca2d0c7 100644
--- a/src/feature/rend/rendservice.c
+++ b/src/feature/rend/rendservice.c
@@ -36,7 +36,7 @@
 #include "feature/rend/rendclient.h"
 #include "feature/rend/rendcommon.h"
 #include "feature/rend/rendservice.h"
-#include "feature/stats/rephist.h"
+#include "feature/stats/predict_ports.h"
 #include "lib/crypt_ops/crypto_dh.h"
 #include "lib/crypt_ops/crypto_rand.h"
 #include "lib/crypt_ops/crypto_util.h"
diff --git a/src/feature/stats/geoip.c b/src/feature/stats/geoip_stats.c
index 019aa7a283..3e647bd46c 100644
--- a/src/feature/stats/geoip.c
+++ b/src/feature/stats/geoip_stats.c
@@ -27,55 +27,26 @@
  * for each country.
  */
 
-#define GEOIP_PRIVATE
 #include "core/or/or.h"
+
 #include "ht.h"
 #include "lib/container/buffers.h"
 #include "app/config/config.h"
 #include "feature/control/control.h"
 #include "feature/client/dnsserv.h"
 #include "core/or/dos.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "feature/nodelist/routerlist.h"
 
 #include "lib/container/order.h"
 #include "lib/time/tvdiff.h"
 
-static void init_geoip_countries(void);
-
-/** An entry from the GeoIP IPv4 file: maps an IPv4 range to a country. */
-typedef struct geoip_ipv4_entry_t {
-  uint32_t ip_low; /**< The lowest IP in the range, in host order */
-  uint32_t ip_high; /**< The highest IP in the range, in host order */
-  intptr_t country; /**< An index into geoip_countries */
-} geoip_ipv4_entry_t;
-
-/** An entry from the GeoIP IPv6 file: maps an IPv6 range to a country. */
-typedef struct geoip_ipv6_entry_t {
-  struct in6_addr ip_low; /**< The lowest IP in the range, in host order */
-  struct in6_addr ip_high; /**< The highest IP in the range, in host order */
-  intptr_t country; /**< An index into geoip_countries */
-} geoip_ipv6_entry_t;
-
-/** A per-country record for GeoIP request history. */
-typedef struct geoip_country_t {
-  char countrycode[3];
-  uint32_t n_v3_ns_requests;
-} geoip_country_t;
-
-/** A list of geoip_country_t */
-static smartlist_t *geoip_countries = NULL;
-/** A map from lowercased country codes to their position in geoip_countries.
- * The index is encoded in the pointer, and 1 is added so that NULL can mean
- * not found. */
-static strmap_t *country_idxplus1_by_lc_code = NULL;
-/** Lists of all known geoip_ipv4_entry_t and geoip_ipv6_entry_t, sorted
- * by their respective ip_low. */
-static smartlist_t *geoip_ipv4_entries = NULL, *geoip_ipv6_entries = NULL;
-
-/** SHA1 digest of the GeoIP files to include in extra-info descriptors. */
-static char geoip_digest[DIGEST_LEN];
-static char geoip6_digest[DIGEST_LEN];
+/** Number of entries in n_v3_ns_requests */
+static size_t n_v3_ns_requests_len = 0;
+/** Array, indexed by country index, of number of v3 networkstatus requests
+ * received from that country */
+static uint32_t *n_v3_ns_requests;
 
 /* Total size in bytes of the geoip client history cache. Used by the OOM
  * handler. */
@@ -109,194 +80,30 @@ geoip_decrement_client_history_cache_size(size_t bytes)
   geoip_client_history_cache_size -= bytes;
 }
 
-/** Return the index of the <b>country</b>'s entry in the GeoIP
- * country list if it is a valid 2-letter country code, otherwise
- * return -1. */
-MOCK_IMPL(country_t,
-geoip_get_country,(const char *country))
-{
-  void *idxplus1_;
-  intptr_t idx;
-
-  idxplus1_ = strmap_get_lc(country_idxplus1_by_lc_code, country);
-  if (!idxplus1_)
-    return -1;
-
-  idx = ((uintptr_t)idxplus1_)-1;
-  return (country_t)idx;
-}
-
-/** Add an entry to a GeoIP table, mapping all IP addresses between <b>low</b>
- * and <b>high</b>, inclusive, to the 2-letter country code <b>country</b>. */
+/** Add 1 to the count of v3 ns requests received from <b>country</b>. */
 static void
-geoip_add_entry(const tor_addr_t *low, const tor_addr_t *high,
-                const char *country)
+increment_v3_ns_request(country_t country)
 {
-  intptr_t idx;
-  void *idxplus1_;
-
-  IF_BUG_ONCE(tor_addr_family(low) != tor_addr_family(high))
-    return;
-  IF_BUG_ONCE(tor_addr_compare(high, low, CMP_EXACT) < 0)
+  if (country < 0)
     return;
 
-  idxplus1_ = strmap_get_lc(country_idxplus1_by_lc_code, country);
-
-  if (!idxplus1_) {
-    geoip_country_t *c = tor_malloc_zero(sizeof(geoip_country_t));
-    strlcpy(c->countrycode, country, sizeof(c->countrycode));
-    tor_strlower(c->countrycode);
-    smartlist_add(geoip_countries, c);
-    idx = smartlist_len(geoip_countries) - 1;
-    strmap_set_lc(country_idxplus1_by_lc_code, country, (void*)(idx+1));
-  } else {
-    idx = ((uintptr_t)idxplus1_)-1;
-  }
-  {
-    geoip_country_t *c = smartlist_get(geoip_countries, (int)idx);
-    tor_assert(!strcasecmp(c->countrycode, country));
-  }
-
-  if (tor_addr_family(low) == AF_INET) {
-    geoip_ipv4_entry_t *ent = tor_malloc_zero(sizeof(geoip_ipv4_entry_t));
-    ent->ip_low = tor_addr_to_ipv4h(low);
-    ent->ip_high = tor_addr_to_ipv4h(high);
-    ent->country = idx;
-    smartlist_add(geoip_ipv4_entries, ent);
-  } else if (tor_addr_family(low) == AF_INET6) {
-    geoip_ipv6_entry_t *ent = tor_malloc_zero(sizeof(geoip_ipv6_entry_t));
-    ent->ip_low = *tor_addr_to_in6_assert(low);
-    ent->ip_high = *tor_addr_to_in6_assert(high);
-    ent->country = idx;
-    smartlist_add(geoip_ipv6_entries, ent);
-  }
-}
-
-/** Add an entry to the GeoIP table indicated by <b>family</b>,
- * parsing it from <b>line</b>. The format is as for geoip_load_file(). */
-STATIC int
-geoip_parse_entry(const char *line, sa_family_t family)
-{
-  tor_addr_t low_addr, high_addr;
-  char c[3];
-  char *country = NULL;
-
-  if (!geoip_countries)
-    init_geoip_countries();
-  if (family == AF_INET) {
-    if (!geoip_ipv4_entries)
-      geoip_ipv4_entries = smartlist_new();
-  } else if (family == AF_INET6) {
-    if (!geoip_ipv6_entries)
-      geoip_ipv6_entries = smartlist_new();
-  } else {
-    log_warn(LD_GENERAL, "Unsupported family: %d", family);
-    return -1;
+  if ((size_t)country >= n_v3_ns_requests_len) {
+    /* We need to reallocate the array. */
+    size_t new_len;
+    if (n_v3_ns_requests_len == 0)
+      new_len = 256;
+    else
+      new_len = n_v3_ns_requests_len * 2;
+    if (new_len <= (size_t)country)
+      new_len = ((size_t)country)+1;
+    n_v3_ns_requests = tor_reallocarray(n_v3_ns_requests, new_len,
+                                        sizeof(uint32_t));
+    memset(n_v3_ns_requests + n_v3_ns_requests_len, 0,
+           sizeof(uint32_t)*(new_len - n_v3_ns_requests_len));
+    n_v3_ns_requests_len = new_len;
   }
 
-  while (TOR_ISSPACE(*line))
-    ++line;
-  if (*line == '#')
-    return 0;
-
-  char buf[512];
-  if (family == AF_INET) {
-    unsigned int low, high;
-    if (tor_sscanf(line,"%u,%u,%2s", &low, &high, c) == 3 ||
-        tor_sscanf(line,"\"%u\",\"%u\",\"%2s\",", &low, &high, c) == 3) {
-      tor_addr_from_ipv4h(&low_addr, low);
-      tor_addr_from_ipv4h(&high_addr, high);
-    } else
-      goto fail;
-    country = c;
-  } else {                      /* AF_INET6 */
-    char *low_str, *high_str;
-    struct in6_addr low, high;
-    char *strtok_state;
-    strlcpy(buf, line, sizeof(buf));
-    low_str = tor_strtok_r(buf, ",", &strtok_state);
-    if (!low_str)
-      goto fail;
-    high_str = tor_strtok_r(NULL, ",", &strtok_state);
-    if (!high_str)
-      goto fail;
-    country = tor_strtok_r(NULL, "\n", &strtok_state);
-    if (!country)
-      goto fail;
-    if (strlen(country) != 2)
-      goto fail;
-    if (tor_inet_pton(AF_INET6, low_str, &low) <= 0)
-      goto fail;
-    tor_addr_from_in6(&low_addr, &low);
-    if (tor_inet_pton(AF_INET6, high_str, &high) <= 0)
-      goto fail;
-    tor_addr_from_in6(&high_addr, &high);
-  }
-  geoip_add_entry(&low_addr, &high_addr, country);
-  return 0;
-
-  fail:
-  log_warn(LD_GENERAL, "Unable to parse line from GEOIP %s file: %s",
-           family == AF_INET ? "IPv4" : "IPv6", escaped(line));
-  return -1;
-}
-
-/** Sorting helper: return -1, 1, or 0 based on comparison of two
- * geoip_ipv4_entry_t */
-static int
-geoip_ipv4_compare_entries_(const void **_a, const void **_b)
-{
-  const geoip_ipv4_entry_t *a = *_a, *b = *_b;
-  if (a->ip_low < b->ip_low)
-    return -1;
-  else if (a->ip_low > b->ip_low)
-    return 1;
-  else
-    return 0;
-}
-
-/** bsearch helper: return -1, 1, or 0 based on comparison of an IP (a pointer
- * to a uint32_t in host order) to a geoip_ipv4_entry_t */
-static int
-geoip_ipv4_compare_key_to_entry_(const void *_key, const void **_member)
-{
-  /* No alignment issue here, since _key really is a pointer to uint32_t */
-  const uint32_t addr = *(uint32_t *)_key;
-  const geoip_ipv4_entry_t *entry = *_member;
-  if (addr < entry->ip_low)
-    return -1;
-  else if (addr > entry->ip_high)
-    return 1;
-  else
-    return 0;
-}
-
-/** Sorting helper: return -1, 1, or 0 based on comparison of two
- * geoip_ipv6_entry_t */
-static int
-geoip_ipv6_compare_entries_(const void **_a, const void **_b)
-{
-  const geoip_ipv6_entry_t *a = *_a, *b = *_b;
-  return fast_memcmp(a->ip_low.s6_addr, b->ip_low.s6_addr,
-                     sizeof(struct in6_addr));
-}
-
-/** bsearch helper: return -1, 1, or 0 based on comparison of an IPv6
- * (a pointer to a in6_addr) to a geoip_ipv6_entry_t */
-static int
-geoip_ipv6_compare_key_to_entry_(const void *_key, const void **_member)
-{
-  const struct in6_addr *addr = (struct in6_addr *)_key;
-  const geoip_ipv6_entry_t *entry = *_member;
-
-  if (fast_memcmp(addr->s6_addr, entry->ip_low.s6_addr,
-             sizeof(struct in6_addr)) < 0)
-    return -1;
-  else if (fast_memcmp(addr->s6_addr, entry->ip_high.s6_addr,
-                  sizeof(struct in6_addr)) > 0)
-    return 1;
-  else
-    return 0;
+  n_v3_ns_requests[country] += 1;
 }
 
 /** Return 1 if we should collect geoip stats on bridge users, and
@@ -307,208 +114,6 @@ should_record_bridge_info(const or_options_t *options)
   return options->BridgeRelay && options->BridgeRecordUsageByCountry;
 }
 
-/** Set up a new list of geoip countries with no countries (yet) set in it,
- * except for the unknown country.
- */
-static void
-init_geoip_countries(void)
-{
-  geoip_country_t *geoip_unresolved;
-  geoip_countries = smartlist_new();
-  /* Add a geoip_country_t for requests that could not be resolved to a
-   * country as first element (index 0) to geoip_countries. */
-  geoip_unresolved = tor_malloc_zero(sizeof(geoip_country_t));
-  strlcpy(geoip_unresolved->countrycode, "??",
-          sizeof(geoip_unresolved->countrycode));
-  smartlist_add(geoip_countries, geoip_unresolved);
-  country_idxplus1_by_lc_code = strmap_new();
-  strmap_set_lc(country_idxplus1_by_lc_code, "??", (void*)(1));
-}
-
-/** Clear appropriate GeoIP database, based on <b>family</b>, and
- * reload it from the file <b>filename</b>. Return 0 on success, -1 on
- * failure.
- *
- * Recognized line formats for IPv4 are:
- *   INTIPLOW,INTIPHIGH,CC
- * and
- *   "INTIPLOW","INTIPHIGH","CC","CC3","COUNTRY NAME"
- * where INTIPLOW and INTIPHIGH are IPv4 addresses encoded as 4-byte unsigned
- * integers, and CC is a country code.
- *
- * Recognized line format for IPv6 is:
- *   IPV6LOW,IPV6HIGH,CC
- * where IPV6LOW and IPV6HIGH are IPv6 addresses and CC is a country code.
- *
- * It also recognizes, and skips over, blank lines and lines that start
- * with '#' (comments).
- */
-int
-geoip_load_file(sa_family_t family, const char *filename)
-{
-  FILE *f;
-  const char *msg = "";
-  const or_options_t *options = get_options();
-  int severity = options_need_geoip_info(options, &msg) ? LOG_WARN : LOG_INFO;
-  crypto_digest_t *geoip_digest_env = NULL;
-
-  tor_assert(family == AF_INET || family == AF_INET6);
-
-  if (!(f = tor_fopen_cloexec(filename, "r"))) {
-    log_fn(severity, LD_GENERAL, "Failed to open GEOIP file %s.  %s",
-           filename, msg);
-    return -1;
-  }
-  if (!geoip_countries)
-    init_geoip_countries();
-
-  if (family == AF_INET) {
-    if (geoip_ipv4_entries) {
-      SMARTLIST_FOREACH(geoip_ipv4_entries, geoip_ipv4_entry_t *, e,
-                        tor_free(e));
-      smartlist_free(geoip_ipv4_entries);
-    }
-    geoip_ipv4_entries = smartlist_new();
-  } else { /* AF_INET6 */
-    if (geoip_ipv6_entries) {
-      SMARTLIST_FOREACH(geoip_ipv6_entries, geoip_ipv6_entry_t *, e,
-                        tor_free(e));
-      smartlist_free(geoip_ipv6_entries);
-    }
-    geoip_ipv6_entries = smartlist_new();
-  }
-  geoip_digest_env = crypto_digest_new();
-
-  log_notice(LD_GENERAL, "Parsing GEOIP %s file %s.",
-             (family == AF_INET) ? "IPv4" : "IPv6", filename);
-  while (!feof(f)) {
-    char buf[512];
-    if (fgets(buf, (int)sizeof(buf), f) == NULL)
-      break;
-    crypto_digest_add_bytes(geoip_digest_env, buf, strlen(buf));
-    /* FFFF track full country name. */
-    geoip_parse_entry(buf, family);
-  }
-  /*XXXX abort and return -1 if no entries/illformed?*/
-  fclose(f);
-
-  /* Sort list and remember file digests so that we can include it in
-   * our extra-info descriptors. */
-  if (family == AF_INET) {
-    smartlist_sort(geoip_ipv4_entries, geoip_ipv4_compare_entries_);
-    /* Okay, now we need to maybe change our mind about what is in
-     * which country. We do this for IPv4 only since that's what we
-     * store in node->country. */
-    refresh_all_country_info();
-    crypto_digest_get_digest(geoip_digest_env, geoip_digest, DIGEST_LEN);
-  } else {
-    /* AF_INET6 */
-    smartlist_sort(geoip_ipv6_entries, geoip_ipv6_compare_entries_);
-    crypto_digest_get_digest(geoip_digest_env, geoip6_digest, DIGEST_LEN);
-  }
-  crypto_digest_free(geoip_digest_env);
-
-  return 0;
-}
-
-/** Given an IP address in host order, return a number representing the
- * country to which that address belongs, -1 for "No geoip information
- * available", or 0 for the 'unknown country'.  The return value will always
- * be less than geoip_get_n_countries().  To decode it, call
- * geoip_get_country_name().
- */
-STATIC int
-geoip_get_country_by_ipv4(uint32_t ipaddr)
-{
-  geoip_ipv4_entry_t *ent;
-  if (!geoip_ipv4_entries)
-    return -1;
-  ent = smartlist_bsearch(geoip_ipv4_entries, &ipaddr,
-                          geoip_ipv4_compare_key_to_entry_);
-  return ent ? (int)ent->country : 0;
-}
-
-/** Given an IPv6 address, return a number representing the country to
- * which that address belongs, -1 for "No geoip information available", or
- * 0 for the 'unknown country'.  The return value will always be less than
- * geoip_get_n_countries().  To decode it, call geoip_get_country_name().
- */
-STATIC int
-geoip_get_country_by_ipv6(const struct in6_addr *addr)
-{
-  geoip_ipv6_entry_t *ent;
-
-  if (!geoip_ipv6_entries)
-    return -1;
-  ent = smartlist_bsearch(geoip_ipv6_entries, addr,
-                          geoip_ipv6_compare_key_to_entry_);
-  return ent ? (int)ent->country : 0;
-}
-
-/** Given an IP address, return a number representing the country to which
- * that address belongs, -1 for "No geoip information available", or 0 for
- * the 'unknown country'.  The return value will always be less than
- * geoip_get_n_countries().  To decode it, call geoip_get_country_name().
- */
-MOCK_IMPL(int,
-geoip_get_country_by_addr,(const tor_addr_t *addr))
-{
-  if (tor_addr_family(addr) == AF_INET) {
-    return geoip_get_country_by_ipv4(tor_addr_to_ipv4h(addr));
-  } else if (tor_addr_family(addr) == AF_INET6) {
-    return geoip_get_country_by_ipv6(tor_addr_to_in6(addr));
-  } else {
-    return -1;
-  }
-}
-
-/** Return the number of countries recognized by the GeoIP country list. */
-MOCK_IMPL(int,
-geoip_get_n_countries,(void))
-{
-  if (!geoip_countries)
-    init_geoip_countries();
-  return (int) smartlist_len(geoip_countries);
-}
-
-/** Return the two-letter country code associated with the number <b>num</b>,
- * or "??" for an unknown value. */
-const char *
-geoip_get_country_name(country_t num)
-{
-  if (geoip_countries && num >= 0 && num < smartlist_len(geoip_countries)) {
-    geoip_country_t *c = smartlist_get(geoip_countries, num);
-    return c->countrycode;
-  } else
-    return "??";
-}
-
-/** Return true iff we have loaded a GeoIP database.*/
-MOCK_IMPL(int,
-geoip_is_loaded,(sa_family_t family))
-{
-  tor_assert(family == AF_INET || family == AF_INET6);
-  if (geoip_countries == NULL)
-    return 0;
-  if (family == AF_INET)
-    return geoip_ipv4_entries != NULL;
-  else                          /* AF_INET6 */
-    return geoip_ipv6_entries != NULL;
-}
-
-/** Return the hex-encoded SHA1 digest of the loaded GeoIP file. The
- * result does not need to be deallocated, but will be overwritten by the
- * next call of hex_str(). */
-const char *
-geoip_db_digest(sa_family_t family)
-{
-  tor_assert(family == AF_INET || family == AF_INET6);
-  if (family == AF_INET)
-    return hex_str(geoip_digest, DIGEST_LEN);
-  else                          /* AF_INET6 */
-    return hex_str(geoip6_digest, DIGEST_LEN);
-}
-
 /** Largest allowable value for last_seen_in_minutes.  (It's a 30-bit field,
  * so it can hold up to (1u<<30)-1, or 0x3fffffffu.
  */
@@ -660,10 +265,7 @@ geoip_note_client_seen(geoip_client_action_t action,
     int country_idx = geoip_get_country_by_addr(addr);
     if (country_idx < 0)
       country_idx = 0; /** unresolved requests are stored at index 0. */
-    if (country_idx >= 0 && country_idx < smartlist_len(geoip_countries)) {
-      geoip_country_t *country = smartlist_get(geoip_countries, country_idx);
-      ++country->n_v3_ns_requests;
-    }
+    increment_v3_ns_request(country_idx);
   }
 }
 
@@ -1275,14 +877,14 @@ geoip_get_request_history(void)
   char *result;
   unsigned granularity = IP_GRANULARITY;
 
-  if (!geoip_countries)
-    return NULL;
-
   entries = smartlist_new();
-  SMARTLIST_FOREACH_BEGIN(geoip_countries, geoip_country_t *, c) {
+  SMARTLIST_FOREACH_BEGIN(geoip_get_countries(), const geoip_country_t *, c) {
       uint32_t tot = 0;
       c_hist_t *ent;
-      tot = c->n_v3_ns_requests;
+      if ((size_t)c_sl_idx < n_v3_ns_requests_len)
+        tot = n_v3_ns_requests[c_sl_idx];
+      else
+        tot = 0;
       if (!tot)
         continue;
       ent = tor_malloc_zero(sizeof(c_hist_t));
@@ -1319,9 +921,8 @@ geoip_dirreq_stats_init(time_t now)
 void
 geoip_reset_dirreq_stats(time_t now)
 {
-  SMARTLIST_FOREACH(geoip_countries, geoip_country_t *, c, {
-      c->n_v3_ns_requests = 0;
-  });
+  memset(n_v3_ns_requests, 0,
+         n_v3_ns_requests_len * sizeof(uint32_t));
   {
     clientmap_entry_t **ent, **next, *this;
     for (ent = HT_START(clientmap, &client_history); ent != NULL;
@@ -1793,74 +1394,9 @@ geoip_entry_stats_write(time_t now)
   return start_of_entry_stats_interval + WRITE_STATS_INTERVAL;
 }
 
-/** Helper used to implement GETINFO ip-to-country/... controller command. */
-int
-getinfo_helper_geoip(control_connection_t *control_conn,
-                     const char *question, char **answer,
-                     const char **errmsg)
-{
-  (void)control_conn;
-  if (!strcmpstart(question, "ip-to-country/")) {
-    int c;
-    sa_family_t family;
-    tor_addr_t addr;
-    question += strlen("ip-to-country/");
-
-    if (!strcmp(question, "ipv4-available") ||
-        !strcmp(question, "ipv6-available")) {
-      family = !strcmp(question, "ipv4-available") ? AF_INET : AF_INET6;
-      const int available = geoip_is_loaded(family);
-      tor_asprintf(answer, "%d", !! available);
-      return 0;
-    }
-
-    family = tor_addr_parse(&addr, question);
-    if (family != AF_INET && family != AF_INET6) {
-      *errmsg = "Invalid address family";
-      return -1;
-    }
-    if (!geoip_is_loaded(family)) {
-      *errmsg = "GeoIP data not loaded";
-      return -1;
-    }
-    if (family == AF_INET)
-      c = geoip_get_country_by_ipv4(tor_addr_to_ipv4h(&addr));
-    else                      /* AF_INET6 */
-      c = geoip_get_country_by_ipv6(tor_addr_to_in6(&addr));
-    *answer = tor_strdup(geoip_get_country_name(c));
-  }
-  return 0;
-}
-
-/** Release all storage held by the GeoIP databases and country list. */
-STATIC void
-clear_geoip_db(void)
-{
-  if (geoip_countries) {
-    SMARTLIST_FOREACH(geoip_countries, geoip_country_t *, c, tor_free(c));
-    smartlist_free(geoip_countries);
-  }
-
-  strmap_free(country_idxplus1_by_lc_code, NULL);
-  if (geoip_ipv4_entries) {
-    SMARTLIST_FOREACH(geoip_ipv4_entries, geoip_ipv4_entry_t *, ent,
-                      tor_free(ent));
-    smartlist_free(geoip_ipv4_entries);
-  }
-  if (geoip_ipv6_entries) {
-    SMARTLIST_FOREACH(geoip_ipv6_entries, geoip_ipv6_entry_t *, ent,
-                      tor_free(ent));
-    smartlist_free(geoip_ipv6_entries);
-  }
-  geoip_countries = NULL;
-  country_idxplus1_by_lc_code = NULL;
-  geoip_ipv4_entries = NULL;
-  geoip_ipv6_entries = NULL;
-}
-
 /** Release all storage held in this file. */
 void
-geoip_free_all(void)
+geoip_stats_free_all(void)
 {
   {
     clientmap_entry_t **ent, **next, *this;
@@ -1881,9 +1417,6 @@ geoip_free_all(void)
     HT_CLEAR(dirreqmap, &dirreq_map);
   }
 
-  clear_geoip_db();
   tor_free(bridge_stats_extrainfo);
-
-  memset(geoip_digest, 0, sizeof(geoip_digest));
-  memset(geoip6_digest, 0, sizeof(geoip6_digest));
+  tor_free(n_v3_ns_requests);
 }
diff --git a/src/feature/stats/geoip.h b/src/feature/stats/geoip_stats.h
index a2175c01e5..97011d62ce 100644
--- a/src/feature/stats/geoip.h
+++ b/src/feature/stats/geoip_stats.h
@@ -5,14 +5,13 @@
 /* See LICENSE for licensing information */
 
 /**
- * \file geoip.h
- * \brief Header file for geoip.c.
+ * \file geoip_stats.h
+ * \brief Header file for geoip_stats.c.
  **/
 
-#ifndef TOR_GEOIP_H
-#define TOR_GEOIP_H
+#ifndef TOR_GEOIP_STATS_H
+#define TOR_GEOIP_STATS_H
 
-#include "lib/testsupport/testsupport.h"
 #include "core/or/dos.h"
 
 /** Indicates an action that we might be noting geoip statistics on.
@@ -73,13 +72,6 @@ typedef enum {
   DIRREQ_CHANNEL_BUFFER_FLUSHED = 4
 } dirreq_state_t;
 
-#ifdef GEOIP_PRIVATE
-STATIC int geoip_parse_entry(const char *line, sa_family_t family);
-STATIC int geoip_get_country_by_ipv4(uint32_t ipaddr);
-STATIC int geoip_get_country_by_ipv6(const struct in6_addr *addr);
-STATIC void clear_geoip_db(void);
-#endif /* defined(GEOIP_PRIVATE) */
-
 /** Entry in a map from IP address to the last time we've seen an incoming
  * connection from that IP address. Used by bridges only to track which
  * countries have them blocked, or the DoS mitigation subsystem if enabled. */
@@ -103,13 +95,6 @@ typedef struct clientmap_entry_t {
 } clientmap_entry_t;
 
 int should_record_bridge_info(const or_options_t *options);
-int geoip_load_file(sa_family_t family, const char *filename);
-MOCK_DECL(int, geoip_get_country_by_addr, (const tor_addr_t *addr));
-MOCK_DECL(int, geoip_get_n_countries, (void));
-const char *geoip_get_country_name(country_t num);
-MOCK_DECL(int, geoip_is_loaded, (sa_family_t family));
-const char *geoip_db_digest(sa_family_t family);
-MOCK_DECL(country_t, geoip_get_country, (const char *countrycode));
 
 void geoip_note_client_seen(geoip_client_action_t action,
                             const tor_addr_t *addr, const char *transport_name,
@@ -126,10 +111,7 @@ char *geoip_get_transport_history(void);
 int geoip_get_client_history(geoip_client_action_t action,
                              char **country_str, char **ipver_str);
 char *geoip_get_request_history(void);
-int getinfo_helper_geoip(control_connection_t *control_conn,
-                         const char *question, char **answer,
-                         const char **errmsg);
-void geoip_free_all(void);
+void geoip_stats_free_all(void);
 
 void geoip_start_dirreq(uint64_t dirreq_id, size_t response_size,
                         dirreq_type_t type);
@@ -154,4 +136,4 @@ const char *geoip_get_bridge_stats_extrainfo(time_t);
 char *geoip_get_bridge_stats_controller(time_t);
 char *format_client_stats_heartbeat(time_t now);
 
-#endif /* !defined(TOR_GEOIP_H) */
+#endif /* !defined(TOR_GEOIP_STATS_H) */
diff --git a/src/feature/stats/predict_ports.c b/src/feature/stats/predict_ports.c
new file mode 100644
index 0000000000..8abf34f50f
--- /dev/null
+++ b/src/feature/stats/predict_ports.c
@@ -0,0 +1,311 @@
+/* Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
+ * Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+/**
+ * \file predict_ports.c
+ * \brief Remember what ports we've needed so we can have circuits ready.
+ *
+ * Predicted ports are used by clients to remember how long it's been
+ * since they opened an exit connection to each given target
+ * port. Clients use this information in order to try to keep circuits
+ * open to exit nodes that can connect to the ports that they care
+ * about.  (The predicted ports mechanism also handles predicted circuit
+ * usage that _isn't_ port-specific, such as resolves, internal circuits,
+ * and so on.)
+ **/
+
+#include "core/or/or.h"
+
+#include "app/config/config.h"
+#include "core/or/channelpadding.h"
+#include "core/or/circuituse.h"
+#include "feature/relay/routermode.h"
+#include "feature/relay/selftest.h"
+#include "feature/stats/predict_ports.h"
+#include "lib/container/bitarray.h"
+#include "lib/time/tvdiff.h"
+
+static size_t predicted_ports_total_alloc = 0;
+
+static void predicted_ports_alloc(void);
+
+/** A single predicted port: used to remember which ports we've made
+ * connections to, so that we can try to keep making circuits that can handle
+ * those ports. */
+typedef struct predicted_port_t {
+  /** The port we connected to */
+  uint16_t port;
+  /** The time at which we last used it */
+  time_t time;
+} predicted_port_t;
+
+/** A list of port numbers that have been used recently. */
+static smartlist_t *predicted_ports_list=NULL;
+/** How long do we keep predicting circuits? */
+static time_t prediction_timeout=0;
+/** When was the last time we added a prediction entry (HS or port) */
+static time_t last_prediction_add_time=0;
+
+/**
+ * How much time left until we stop predicting circuits?
+ */
+int
+predicted_ports_prediction_time_remaining(time_t now)
+{
+  time_t seconds_waited;
+  time_t seconds_left;
+
+  /* Protect against overflow of return value. This can happen if the clock
+   * jumps backwards in time. Update the last prediction time (aka last
+   * active time) to prevent it. This update is preferable to using monotonic
+   * time because it prevents clock jumps into the past from simply causing
+   * very long idle timeouts while the monotonic time stands still. */
+  seconds_waited = time_diff(last_prediction_add_time, now);
+  if (seconds_waited == TIME_MAX) {
+    last_prediction_add_time = now;
+    seconds_waited = 0;
+  }
+
+  /* Protect against underflow of the return value. This can happen for very
+   * large periods of inactivity/system sleep. */
+  if (seconds_waited > prediction_timeout)
+    return 0;
+
+  seconds_left = time_diff(seconds_waited, prediction_timeout);
+  if (BUG(seconds_left == TIME_MAX))
+    return INT_MAX;
+
+  return (int)(seconds_left);
+}
+
+/** We just got an application request for a connection with
+ * port <b>port</b>. Remember it for the future, so we can keep
+ * some circuits open that will exit to this port.
+ */
+static void
+add_predicted_port(time_t now, uint16_t port)
+{
+  predicted_port_t *pp = tor_malloc(sizeof(predicted_port_t));
+
+  //  If the list is empty, re-randomize predicted ports lifetime
+  if (!any_predicted_circuits(now)) {
+    prediction_timeout =
+     (time_t)channelpadding_get_circuits_available_timeout();
+  }
+
+  last_prediction_add_time = now;
+
+  log_info(LD_CIRC,
+          "New port prediction added. Will continue predictive circ building "
+          "for %d more seconds.",
+          predicted_ports_prediction_time_remaining(now));
+
+  pp->port = port;
+  pp->time = now;
+  predicted_ports_total_alloc += sizeof(*pp);
+  smartlist_add(predicted_ports_list, pp);
+}
+
+/** Remember that <b>port</b> has been asked for as of time <b>now</b>.
+ * This is used for predicting what sorts of streams we'll make in the
+ * future and making exit circuits to anticipate that.
+ */
+void
+rep_hist_note_used_port(time_t now, uint16_t port)
+{
+  tor_assert(predicted_ports_list);
+
+  if (!port) /* record nothing */
+    return;
+
+  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
+    if (pp->port == port) {
+      pp->time = now;
+
+      last_prediction_add_time = now;
+      log_info(LD_CIRC,
+               "New port prediction added. Will continue predictive circ "
+               "building for %d more seconds.",
+               predicted_ports_prediction_time_remaining(now));
+      return;
+    }
+  } SMARTLIST_FOREACH_END(pp);
+  /* it's not there yet; we need to add it */
+  add_predicted_port(now, port);
+}
+
+/** Return a newly allocated pointer to a list of uint16_t * for ports that
+ * are likely to be asked for in the near future.
+ */
+smartlist_t *
+rep_hist_get_predicted_ports(time_t now)
+{
+  int predicted_circs_relevance_time;
+  smartlist_t *out = smartlist_new();
+  tor_assert(predicted_ports_list);
+
+  predicted_circs_relevance_time = (int)prediction_timeout;
+
+  /* clean out obsolete entries */
+  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
+    if (pp->time + predicted_circs_relevance_time < now) {
+      log_debug(LD_CIRC, "Expiring predicted port %d", pp->port);
+
+      predicted_ports_total_alloc -= sizeof(predicted_port_t);
+      tor_free(pp);
+      SMARTLIST_DEL_CURRENT(predicted_ports_list, pp);
+    } else {
+      smartlist_add(out, tor_memdup(&pp->port, sizeof(uint16_t)));
+    }
+  } SMARTLIST_FOREACH_END(pp);
+  return out;
+}
+
+/**
+ * Take a list of uint16_t *, and remove every port in the list from the
+ * current list of predicted ports.
+ */
+void
+rep_hist_remove_predicted_ports(const smartlist_t *rmv_ports)
+{
+  /* Let's do this on O(N), not O(N^2). */
+  bitarray_t *remove_ports = bitarray_init_zero(UINT16_MAX);
+  SMARTLIST_FOREACH(rmv_ports, const uint16_t *, p,
+                    bitarray_set(remove_ports, *p));
+  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
+    if (bitarray_is_set(remove_ports, pp->port)) {
+      tor_free(pp);
+      predicted_ports_total_alloc -= sizeof(*pp);
+      SMARTLIST_DEL_CURRENT(predicted_ports_list, pp);
+    }
+  } SMARTLIST_FOREACH_END(pp);
+  bitarray_free(remove_ports);
+}
+
+/** The user asked us to do a resolve. Rather than keeping track of
+ * timings and such of resolves, we fake it for now by treating
+ * it the same way as a connection to port 80. This way we will continue
+ * to have circuits lying around if the user only uses Tor for resolves.
+ */
+void
+rep_hist_note_used_resolve(time_t now)
+{
+  rep_hist_note_used_port(now, 80);
+}
+
+/** The last time at which we needed an internal circ. */
+static time_t predicted_internal_time = 0;
+/** The last time we needed an internal circ with good uptime. */
+static time_t predicted_internal_uptime_time = 0;
+/** The last time we needed an internal circ with good capacity. */
+static time_t predicted_internal_capacity_time = 0;
+
+/** Remember that we used an internal circ at time <b>now</b>. */
+void
+rep_hist_note_used_internal(time_t now, int need_uptime, int need_capacity)
+{
+  // If the list is empty, re-randomize predicted ports lifetime
+  if (!any_predicted_circuits(now)) {
+    prediction_timeout = channelpadding_get_circuits_available_timeout();
+  }
+
+  last_prediction_add_time = now;
+
+  log_info(LD_CIRC,
+          "New port prediction added. Will continue predictive circ building "
+          "for %d more seconds.",
+          predicted_ports_prediction_time_remaining(now));
+
+  predicted_internal_time = now;
+  if (need_uptime)
+    predicted_internal_uptime_time = now;
+  if (need_capacity)
+    predicted_internal_capacity_time = now;
+}
+
+/** Return 1 if we've used an internal circ recently; else return 0. */
+int
+rep_hist_get_predicted_internal(time_t now, int *need_uptime,
+                                int *need_capacity)
+{
+  int predicted_circs_relevance_time;
+
+  predicted_circs_relevance_time = (int)prediction_timeout;
+
+  if (!predicted_internal_time) { /* initialize it */
+    predicted_internal_time = now;
+    predicted_internal_uptime_time = now;
+    predicted_internal_capacity_time = now;
+  }
+  if (predicted_internal_time + predicted_circs_relevance_time < now)
+    return 0; /* too long ago */
+  if (predicted_internal_uptime_time + predicted_circs_relevance_time >= now)
+    *need_uptime = 1;
+  // Always predict that we need capacity.
+  *need_capacity = 1;
+  return 1;
+}
+
+/** Any ports used lately? These are pre-seeded if we just started
+ * up or if we're running a hidden service. */
+int
+any_predicted_circuits(time_t now)
+{
+  int predicted_circs_relevance_time;
+  predicted_circs_relevance_time = (int)prediction_timeout;
+
+  return smartlist_len(predicted_ports_list) ||
+         predicted_internal_time + predicted_circs_relevance_time >= now;
+}
+
+/** Return 1 if we have no need for circuits currently, else return 0. */
+int
+rep_hist_circbuilding_dormant(time_t now)
+{
+  const or_options_t *options = get_options();
+
+  if (any_predicted_circuits(now))
+    return 0;
+
+  /* see if we'll still need to build testing circuits */
+  if (server_mode(options) &&
+      (!check_whether_orport_reachable(options) ||
+       !circuit_enough_testing_circs()))
+    return 0;
+  if (!check_whether_dirport_reachable(options))
+    return 0;
+
+  return 1;
+}
+
+/**
+ * Allocate whatever memory and structs are needed for predicting
+ * which ports will be used. Also seed it with port 80, so we'll build
+ * circuits on start-up.
+ */
+static void
+predicted_ports_alloc(void)
+{
+  predicted_ports_list = smartlist_new();
+}
+
+void
+predicted_ports_init(void)
+{
+  predicted_ports_alloc();
+  add_predicted_port(time(NULL), 443); // Add a port to get us started
+}
+
+/** Free whatever memory is needed for predicting which ports will
+ * be used.
+ */
+void
+predicted_ports_free_all(void)
+{
+  predicted_ports_total_alloc -=
+    smartlist_len(predicted_ports_list)*sizeof(predicted_port_t);
+  SMARTLIST_FOREACH(predicted_ports_list, predicted_port_t *,
+                    pp, tor_free(pp));
+  smartlist_free(predicted_ports_list);
+}
diff --git a/src/feature/stats/predict_ports.h b/src/feature/stats/predict_ports.h
new file mode 100644
index 0000000000..ecb2e9caf2
--- /dev/null
+++ b/src/feature/stats/predict_ports.h
@@ -0,0 +1,30 @@
+/* Copyright (c) 2001 Matej Pfajfar.
+ * Copyright (c) 2001-2004, Roger Dingledine.
+ * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
+ * Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+/**
+ * \file predict_portst.h
+ * \brief Header file for predict_ports.c.
+ **/
+
+#ifndef TOR_PREDICT_PORTS_H
+#define TOR_PREDICT_PORTS_H
+
+void predicted_ports_init(void);
+void rep_hist_note_used_port(time_t now, uint16_t port);
+smartlist_t *rep_hist_get_predicted_ports(time_t now);
+void rep_hist_remove_predicted_ports(const smartlist_t *rmv_ports);
+void rep_hist_note_used_resolve(time_t now);
+void rep_hist_note_used_internal(time_t now, int need_uptime,
+                                 int need_capacity);
+int rep_hist_get_predicted_internal(time_t now, int *need_uptime,
+                                    int *need_capacity);
+
+int any_predicted_circuits(time_t now);
+int rep_hist_circbuilding_dormant(time_t now);
+int predicted_ports_prediction_time_remaining(time_t now);
+void predicted_ports_free_all(void);
+
+#endif
diff --git a/src/feature/stats/rephist.c b/src/feature/stats/rephist.c
index fa4bd0ba4e..37c4759f37 100644
--- a/src/feature/stats/rephist.c
+++ b/src/feature/stats/rephist.c
@@ -76,38 +76,29 @@
 
 #define REPHIST_PRIVATE
 #include "core/or/or.h"
-#include "core/or/circuitlist.h"
-#include "core/or/circuituse.h"
 #include "app/config/config.h"
-#include "lib/crypt_ops/crypto_rand.h"
+#include "app/config/statefile.h"
+#include "core/or/circuitlist.h"
+#include "core/or/connection_or.h"
+#include "feature/dirauth/authmode.h"
 #include "feature/nodelist/networkstatus.h"
 #include "feature/nodelist/nodelist.h"
-#include "feature/stats/rephist.h"
 #include "feature/relay/routermode.h"
-#include "feature/relay/selftest.h"
-#include "feature/nodelist/routerlist.h"
-#include "ht.h"
-#include "core/or/channelpadding.h"
-#include "core/or/connection_or.h"
-#include "app/config/statefile.h"
-#include "feature/dirauth/authmode.h"
+#include "feature/stats/predict_ports.h"
+#include "feature/stats/rephist.h"
+#include "lib/container/order.h"
+#include "lib/crypt_ops/crypto_rand.h"
+#include "lib/math/laplace.h"
 
 #include "feature/nodelist/networkstatus_st.h"
 #include "core/or/or_circuit_st.h"
 #include "app/config/or_state_st.h"
 
-#include "lib/container/bloomfilt.h"
-#include "lib/container/order.h"
-#include "lib/math/fp.h"
-#include "lib/math/laplace.h"
-#include "lib/time/tvdiff.h"
-
 #ifdef HAVE_FCNTL_H
 #include <fcntl.h>
 #endif
 
 static void bw_arrays_init(void);
-static void predicted_ports_alloc(void);
 
 /** Total number of bytes currently allocated in fields used by rephist.c. */
 uint64_t rephist_total_alloc=0;
@@ -242,7 +233,6 @@ rep_hist_init(void)
 {
   history_map = digestmap_new();
   bw_arrays_init();
-  predicted_ports_alloc();
 }
 
 /** We have just decided that this router with identity digest <b>id</b> is
@@ -1537,287 +1527,6 @@ rep_hist_load_state(or_state_t *state, char **err)
   return 0;
 }
 
-/*********************************************************************/
-
-/** A single predicted port: used to remember which ports we've made
- * connections to, so that we can try to keep making circuits that can handle
- * those ports. */
-typedef struct predicted_port_t {
-  /** The port we connected to */
-  uint16_t port;
-  /** The time at which we last used it */
-  time_t time;
-} predicted_port_t;
-
-/** A list of port numbers that have been used recently. */
-static smartlist_t *predicted_ports_list=NULL;
-/** How long do we keep predicting circuits? */
-static time_t prediction_timeout=0;
-/** When was the last time we added a prediction entry (HS or port) */
-static time_t last_prediction_add_time=0;
-
-/**
- * How much time left until we stop predicting circuits?
- */
-int
-predicted_ports_prediction_time_remaining(time_t now)
-{
-  time_t seconds_waited;
-  time_t seconds_left;
-
-  /* Protect against overflow of return value. This can happen if the clock
-   * jumps backwards in time. Update the last prediction time (aka last
-   * active time) to prevent it. This update is preferable to using monotonic
-   * time because it prevents clock jumps into the past from simply causing
-   * very long idle timeouts while the monotonic time stands still. */
-  seconds_waited = time_diff(last_prediction_add_time, now);
-  if (seconds_waited == TIME_MAX) {
-    last_prediction_add_time = now;
-    seconds_waited = 0;
-  }
-
-  /* Protect against underflow of the return value. This can happen for very
-   * large periods of inactivity/system sleep. */
-  if (seconds_waited > prediction_timeout)
-    return 0;
-
-  seconds_left = time_diff(seconds_waited, prediction_timeout);
-  if (BUG(seconds_left == TIME_MAX))
-    return INT_MAX;
-
-  return (int)(seconds_left);
-}
-
-/** We just got an application request for a connection with
- * port <b>port</b>. Remember it for the future, so we can keep
- * some circuits open that will exit to this port.
- */
-static void
-add_predicted_port(time_t now, uint16_t port)
-{
-  predicted_port_t *pp = tor_malloc(sizeof(predicted_port_t));
-
-  //  If the list is empty, re-randomize predicted ports lifetime
-  if (!any_predicted_circuits(now)) {
-    prediction_timeout =
-     (time_t)channelpadding_get_circuits_available_timeout();
-  }
-
-  last_prediction_add_time = now;
-
-  log_info(LD_CIRC,
-          "New port prediction added. Will continue predictive circ building "
-          "for %d more seconds.",
-          predicted_ports_prediction_time_remaining(now));
-
-  pp->port = port;
-  pp->time = now;
-  rephist_total_alloc += sizeof(*pp);
-  smartlist_add(predicted_ports_list, pp);
-}
-
-/**
- * Allocate whatever memory and structs are needed for predicting
- * which ports will be used. Also seed it with port 80, so we'll build
- * circuits on start-up.
- */
-static void
-predicted_ports_alloc(void)
-{
-  predicted_ports_list = smartlist_new();
-}
-
-void
-predicted_ports_init(void)
-{
-  add_predicted_port(time(NULL), 443); // Add a port to get us started
-}
-
-/** Free whatever memory is needed for predicting which ports will
- * be used.
- */
-static void
-predicted_ports_free_all(void)
-{
-  rephist_total_alloc -=
-    smartlist_len(predicted_ports_list)*sizeof(predicted_port_t);
-  SMARTLIST_FOREACH(predicted_ports_list, predicted_port_t *,
-                    pp, tor_free(pp));
-  smartlist_free(predicted_ports_list);
-}
-
-/** Remember that <b>port</b> has been asked for as of time <b>now</b>.
- * This is used for predicting what sorts of streams we'll make in the
- * future and making exit circuits to anticipate that.
- */
-void
-rep_hist_note_used_port(time_t now, uint16_t port)
-{
-  tor_assert(predicted_ports_list);
-
-  if (!port) /* record nothing */
-    return;
-
-  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
-    if (pp->port == port) {
-      pp->time = now;
-
-      last_prediction_add_time = now;
-      log_info(LD_CIRC,
-               "New port prediction added. Will continue predictive circ "
-               "building for %d more seconds.",
-               predicted_ports_prediction_time_remaining(now));
-      return;
-    }
-  } SMARTLIST_FOREACH_END(pp);
-  /* it's not there yet; we need to add it */
-  add_predicted_port(now, port);
-}
-
-/** Return a newly allocated pointer to a list of uint16_t * for ports that
- * are likely to be asked for in the near future.
- */
-smartlist_t *
-rep_hist_get_predicted_ports(time_t now)
-{
-  int predicted_circs_relevance_time;
-  smartlist_t *out = smartlist_new();
-  tor_assert(predicted_ports_list);
-
-  predicted_circs_relevance_time = (int)prediction_timeout;
-
-  /* clean out obsolete entries */
-  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
-    if (pp->time + predicted_circs_relevance_time < now) {
-      log_debug(LD_CIRC, "Expiring predicted port %d", pp->port);
-
-      rephist_total_alloc -= sizeof(predicted_port_t);
-      tor_free(pp);
-      SMARTLIST_DEL_CURRENT(predicted_ports_list, pp);
-    } else {
-      smartlist_add(out, tor_memdup(&pp->port, sizeof(uint16_t)));
-    }
-  } SMARTLIST_FOREACH_END(pp);
-  return out;
-}
-
-/**
- * Take a list of uint16_t *, and remove every port in the list from the
- * current list of predicted ports.
- */
-void
-rep_hist_remove_predicted_ports(const smartlist_t *rmv_ports)
-{
-  /* Let's do this on O(N), not O(N^2). */
-  bitarray_t *remove_ports = bitarray_init_zero(UINT16_MAX);
-  SMARTLIST_FOREACH(rmv_ports, const uint16_t *, p,
-                    bitarray_set(remove_ports, *p));
-  SMARTLIST_FOREACH_BEGIN(predicted_ports_list, predicted_port_t *, pp) {
-    if (bitarray_is_set(remove_ports, pp->port)) {
-      tor_free(pp);
-      rephist_total_alloc -= sizeof(*pp);
-      SMARTLIST_DEL_CURRENT(predicted_ports_list, pp);
-    }
-  } SMARTLIST_FOREACH_END(pp);
-  bitarray_free(remove_ports);
-}
-
-/** The user asked us to do a resolve. Rather than keeping track of
- * timings and such of resolves, we fake it for now by treating
- * it the same way as a connection to port 80. This way we will continue
- * to have circuits lying around if the user only uses Tor for resolves.
- */
-void
-rep_hist_note_used_resolve(time_t now)
-{
-  rep_hist_note_used_port(now, 80);
-}
-
-/** The last time at which we needed an internal circ. */
-static time_t predicted_internal_time = 0;
-/** The last time we needed an internal circ with good uptime. */
-static time_t predicted_internal_uptime_time = 0;
-/** The last time we needed an internal circ with good capacity. */
-static time_t predicted_internal_capacity_time = 0;
-
-/** Remember that we used an internal circ at time <b>now</b>. */
-void
-rep_hist_note_used_internal(time_t now, int need_uptime, int need_capacity)
-{
-  // If the list is empty, re-randomize predicted ports lifetime
-  if (!any_predicted_circuits(now)) {
-    prediction_timeout = channelpadding_get_circuits_available_timeout();
-  }
-
-  last_prediction_add_time = now;
-
-  log_info(LD_CIRC,
-          "New port prediction added. Will continue predictive circ building "
-          "for %d more seconds.",
-          predicted_ports_prediction_time_remaining(now));
-
-  predicted_internal_time = now;
-  if (need_uptime)
-    predicted_internal_uptime_time = now;
-  if (need_capacity)
-    predicted_internal_capacity_time = now;
-}
-
-/** Return 1 if we've used an internal circ recently; else return 0. */
-int
-rep_hist_get_predicted_internal(time_t now, int *need_uptime,
-                                int *need_capacity)
-{
-  int predicted_circs_relevance_time;
-
-  predicted_circs_relevance_time = (int)prediction_timeout;
-
-  if (!predicted_internal_time) { /* initialize it */
-    predicted_internal_time = now;
-    predicted_internal_uptime_time = now;
-    predicted_internal_capacity_time = now;
-  }
-  if (predicted_internal_time + predicted_circs_relevance_time < now)
-    return 0; /* too long ago */
-  if (predicted_internal_uptime_time + predicted_circs_relevance_time >= now)
-    *need_uptime = 1;
-  // Always predict that we need capacity.
-  *need_capacity = 1;
-  return 1;
-}
-
-/** Any ports used lately? These are pre-seeded if we just started
- * up or if we're running a hidden service. */
-int
-any_predicted_circuits(time_t now)
-{
-  int predicted_circs_relevance_time;
-  predicted_circs_relevance_time = (int)prediction_timeout;
-
-  return smartlist_len(predicted_ports_list) ||
-         predicted_internal_time + predicted_circs_relevance_time >= now;
-}
-
-/** Return 1 if we have no need for circuits currently, else return 0. */
-int
-rep_hist_circbuilding_dormant(time_t now)
-{
-  const or_options_t *options = get_options();
-
-  if (any_predicted_circuits(now))
-    return 0;
-
-  /* see if we'll still need to build testing circuits */
-  if (server_mode(options) &&
-      (!check_whether_orport_reachable(options) ||
-       !circuit_enough_testing_circs()))
-    return 0;
-  if (!check_whether_dirport_reachable(options))
-    return 0;
-
-  return 1;
-}
-
 /*** Exit port statistics ***/
 
 /* Some constants */
diff --git a/src/feature/stats/rephist.h b/src/feature/stats/rephist.h
index 842184f815..0584b4684a 100644
--- a/src/feature/stats/rephist.h
+++ b/src/feature/stats/rephist.h
@@ -41,20 +41,6 @@ double rep_hist_get_weighted_fractional_uptime(const char *id, time_t when);
 long rep_hist_get_weighted_time_known(const char *id, time_t when);
 int rep_hist_have_measured_enough_stability(void);
 
-void predicted_ports_init(void);
-void rep_hist_note_used_port(time_t now, uint16_t port);
-smartlist_t *rep_hist_get_predicted_ports(time_t now);
-void rep_hist_remove_predicted_ports(const smartlist_t *rmv_ports);
-void rep_hist_note_used_resolve(time_t now);
-void rep_hist_note_used_internal(time_t now, int need_uptime,
-                                 int need_capacity);
-int rep_hist_get_predicted_internal(time_t now, int *need_uptime,
-                                    int *need_capacity);
-
-int any_predicted_circuits(time_t now);
-int rep_hist_circbuilding_dormant(time_t now);
-int predicted_ports_prediction_time_remaining(time_t now);
-
 void rep_hist_exit_stats_init(time_t now);
 void rep_hist_reset_exit_stats(time_t now);
 void rep_hist_exit_stats_term(void);
diff --git a/src/include.am b/src/include.am
index b5cca7e30a..d2f83da814 100644
--- a/src/include.am
+++ b/src/include.am
@@ -11,6 +11,7 @@ include src/lib/encoding/include.am
 include src/lib/evloop/include.am
 include src/lib/fdio/include.am
 include src/lib/fs/include.am
+include src/lib/geoip/include.am
 include src/lib/include.libdonna.am
 include src/lib/intmath/include.am
 include src/lib/lock/include.am
diff --git a/src/lib/geoip/.may_include b/src/lib/geoip/.may_include
new file mode 100644
index 0000000000..b1ee2dcfe9
--- /dev/null
+++ b/src/lib/geoip/.may_include
@@ -0,0 +1,13 @@
+orconfig.h
+lib/cc/*.h
+lib/container/*.h
+lib/crypt_ops/*.h
+lib/ctime/*.h
+lib/encoding/*.h
+lib/fs/*.h
+lib/geoip/*.h
+lib/log/*.h
+lib/malloc/*.h
+lib/net/*.h
+lib/string/*.h
+lib/testsupport/*.h
diff --git a/src/lib/geoip/country.h b/src/lib/geoip/country.h
new file mode 100644
index 0000000000..e4ad0752b3
--- /dev/null
+++ b/src/lib/geoip/country.h
@@ -0,0 +1,14 @@
+/* Copyright (c) 2001 Matej Pfajfar.
+ * Copyright (c) 2001-2004, Roger Dingledine.
+ * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
+ * Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+#ifndef TOR_COUNTRY_H
+#define TOR_COUNTRY_H
+
+#include "lib/cc/torint.h"
+/** A signed integer representing a country code. */
+typedef int16_t country_t;
+
+#endif
diff --git a/src/lib/geoip/geoip.c b/src/lib/geoip/geoip.c
new file mode 100644
index 0000000000..b1c0973d03
--- /dev/null
+++ b/src/lib/geoip/geoip.c
@@ -0,0 +1,510 @@
+/* Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+/**
+ * \file geoip.c
+ * \brief Functions related to maintaining an IP-to-country database;
+ * to summarizing client connections by country to entry guards, bridges,
+ * and directory servers; and for statistics on answering network status
+ * requests.
+ *
+ * There are two main kinds of functions in this module: geoip functions,
+ * which map groups of IPv4 and IPv6 addresses to country codes, and
+ * statistical functions, which collect statistics about different kinds of
+ * per-country usage.
+ *
+ * The geoip lookup tables are implemented as sorted lists of disjoint address
+ * ranges, each mapping to a singleton geoip_country_t.  These country objects
+ * are also indexed by their names in a hashtable.
+ *
+ * The tables are populated from disk at startup by the geoip_load_file()
+ * function.  For more information on the file format they read, see that
+ * function.  See the scripts and the README file in src/config for more
+ * information about how those files are generated.
+ *
+ * Tor uses GeoIP information in order to implement user requests (such as
+ * ExcludeNodes {cc}), and to keep track of how much usage relays are getting
+ * for each country.
+ */
+
+#define GEOIP_PRIVATE
+#include "lib/geoip/geoip.h"
+#include "lib/container/map.h"
+#include "lib/container/order.h"
+#include "lib/container/smartlist.h"
+#include "lib/crypt_ops/crypto_digest.h"
+#include "lib/ctime/di_ops.h"
+#include "lib/encoding/binascii.h"
+#include "lib/fs/files.h"
+#include "lib/log/escape.h"
+#include "lib/malloc/malloc.h"
+#include "lib/net/address.h" //????
+#include "lib/net/inaddr.h"
+#include "lib/string/compat_ctype.h"
+#include "lib/string/compat_string.h"
+#include "lib/string/scanf.h"
+#include "lib/string/util_string.h"
+
+#include <stdio.h>
+#include <string.h>
+
+static void init_geoip_countries(void);
+
+/** An entry from the GeoIP IPv4 file: maps an IPv4 range to a country. */
+typedef struct geoip_ipv4_entry_t {
+  uint32_t ip_low; /**< The lowest IP in the range, in host order */
+  uint32_t ip_high; /**< The highest IP in the range, in host order */
+  intptr_t country; /**< An index into geoip_countries */
+} geoip_ipv4_entry_t;
+
+/** An entry from the GeoIP IPv6 file: maps an IPv6 range to a country. */
+typedef struct geoip_ipv6_entry_t {
+  struct in6_addr ip_low; /**< The lowest IP in the range, in host order */
+  struct in6_addr ip_high; /**< The highest IP in the range, in host order */
+  intptr_t country; /**< An index into geoip_countries */
+} geoip_ipv6_entry_t;
+
+/** A list of geoip_country_t */
+static smartlist_t *geoip_countries = NULL;
+/** A map from lowercased country codes to their position in geoip_countries.
+ * The index is encoded in the pointer, and 1 is added so that NULL can mean
+ * not found. */
+static strmap_t *country_idxplus1_by_lc_code = NULL;
+/** Lists of all known geoip_ipv4_entry_t and geoip_ipv6_entry_t, sorted
+ * by their respective ip_low. */
+static smartlist_t *geoip_ipv4_entries = NULL, *geoip_ipv6_entries = NULL;
+
+/** SHA1 digest of the GeoIP files to include in extra-info descriptors. */
+static char geoip_digest[DIGEST_LEN];
+static char geoip6_digest[DIGEST_LEN];
+
+/** Return a list of geoip_country_t for all known countries. */
+const smartlist_t *
+geoip_get_countries(void)
+{
+  if (geoip_countries == NULL) {
+    init_geoip_countries();
+  }
+  return geoip_countries;
+}
+
+/** Return the index of the <b>country</b>'s entry in the GeoIP
+ * country list if it is a valid 2-letter country code, otherwise
+ * return -1. */
+MOCK_IMPL(country_t,
+geoip_get_country,(const char *country))
+{
+  void *idxplus1_;
+  intptr_t idx;
+
+  idxplus1_ = strmap_get_lc(country_idxplus1_by_lc_code, country);
+  if (!idxplus1_)
+    return -1;
+
+  idx = ((uintptr_t)idxplus1_)-1;
+  return (country_t)idx;
+}
+
+/** Add an entry to a GeoIP table, mapping all IP addresses between <b>low</b>
+ * and <b>high</b>, inclusive, to the 2-letter country code <b>country</b>. */
+static void
+geoip_add_entry(const tor_addr_t *low, const tor_addr_t *high,
+                const char *country)
+{
+  intptr_t idx;
+  void *idxplus1_;
+
+  IF_BUG_ONCE(tor_addr_family(low) != tor_addr_family(high))
+    return;
+  IF_BUG_ONCE(tor_addr_compare(high, low, CMP_EXACT) < 0)
+    return;
+
+  idxplus1_ = strmap_get_lc(country_idxplus1_by_lc_code, country);
+
+  if (!idxplus1_) {
+    geoip_country_t *c = tor_malloc_zero(sizeof(geoip_country_t));
+    strlcpy(c->countrycode, country, sizeof(c->countrycode));
+    tor_strlower(c->countrycode);
+    smartlist_add(geoip_countries, c);
+    idx = smartlist_len(geoip_countries) - 1;
+    strmap_set_lc(country_idxplus1_by_lc_code, country, (void*)(idx+1));
+  } else {
+    idx = ((uintptr_t)idxplus1_)-1;
+  }
+  {
+    geoip_country_t *c = smartlist_get(geoip_countries, (int)idx);
+    tor_assert(!strcasecmp(c->countrycode, country));
+  }
+
+  if (tor_addr_family(low) == AF_INET) {
+    geoip_ipv4_entry_t *ent = tor_malloc_zero(sizeof(geoip_ipv4_entry_t));
+    ent->ip_low = tor_addr_to_ipv4h(low);
+    ent->ip_high = tor_addr_to_ipv4h(high);
+    ent->country = idx;
+    smartlist_add(geoip_ipv4_entries, ent);
+  } else if (tor_addr_family(low) == AF_INET6) {
+    geoip_ipv6_entry_t *ent = tor_malloc_zero(sizeof(geoip_ipv6_entry_t));
+    ent->ip_low = *tor_addr_to_in6_assert(low);
+    ent->ip_high = *tor_addr_to_in6_assert(high);
+    ent->country = idx;
+    smartlist_add(geoip_ipv6_entries, ent);
+  }
+}
+
+/** Add an entry to the GeoIP table indicated by <b>family</b>,
+ * parsing it from <b>line</b>. The format is as for geoip_load_file(). */
+STATIC int
+geoip_parse_entry(const char *line, sa_family_t family)
+{
+  tor_addr_t low_addr, high_addr;
+  char c[3];
+  char *country = NULL;
+
+  if (!geoip_countries)
+    init_geoip_countries();
+  if (family == AF_INET) {
+    if (!geoip_ipv4_entries)
+      geoip_ipv4_entries = smartlist_new();
+  } else if (family == AF_INET6) {
+    if (!geoip_ipv6_entries)
+      geoip_ipv6_entries = smartlist_new();
+  } else {
+    log_warn(LD_GENERAL, "Unsupported family: %d", family);
+    return -1;
+  }
+
+  while (TOR_ISSPACE(*line))
+    ++line;
+  if (*line == '#')
+    return 0;
+
+  char buf[512];
+  if (family == AF_INET) {
+    unsigned int low, high;
+    if (tor_sscanf(line,"%u,%u,%2s", &low, &high, c) == 3 ||
+        tor_sscanf(line,"\"%u\",\"%u\",\"%2s\",", &low, &high, c) == 3) {
+      tor_addr_from_ipv4h(&low_addr, low);
+      tor_addr_from_ipv4h(&high_addr, high);
+    } else
+      goto fail;
+    country = c;
+  } else {                      /* AF_INET6 */
+    char *low_str, *high_str;
+    struct in6_addr low, high;
+    char *strtok_state;
+    strlcpy(buf, line, sizeof(buf));
+    low_str = tor_strtok_r(buf, ",", &strtok_state);
+    if (!low_str)
+      goto fail;
+    high_str = tor_strtok_r(NULL, ",", &strtok_state);
+    if (!high_str)
+      goto fail;
+    country = tor_strtok_r(NULL, "\n", &strtok_state);
+    if (!country)
+      goto fail;
+    if (strlen(country) != 2)
+      goto fail;
+    if (tor_inet_pton(AF_INET6, low_str, &low) <= 0)
+      goto fail;
+    tor_addr_from_in6(&low_addr, &low);
+    if (tor_inet_pton(AF_INET6, high_str, &high) <= 0)
+      goto fail;
+    tor_addr_from_in6(&high_addr, &high);
+  }
+  geoip_add_entry(&low_addr, &high_addr, country);
+  return 0;
+
+  fail:
+  log_warn(LD_GENERAL, "Unable to parse line from GEOIP %s file: %s",
+           family == AF_INET ? "IPv4" : "IPv6", escaped(line));
+  return -1;
+}
+
+/** Sorting helper: return -1, 1, or 0 based on comparison of two
+ * geoip_ipv4_entry_t */
+static int
+geoip_ipv4_compare_entries_(const void **_a, const void **_b)
+{
+  const geoip_ipv4_entry_t *a = *_a, *b = *_b;
+  if (a->ip_low < b->ip_low)
+    return -1;
+  else if (a->ip_low > b->ip_low)
+    return 1;
+  else
+    return 0;
+}
+
+/** bsearch helper: return -1, 1, or 0 based on comparison of an IP (a pointer
+ * to a uint32_t in host order) to a geoip_ipv4_entry_t */
+static int
+geoip_ipv4_compare_key_to_entry_(const void *_key, const void **_member)
+{
+  /* No alignment issue here, since _key really is a pointer to uint32_t */
+  const uint32_t addr = *(uint32_t *)_key;
+  const geoip_ipv4_entry_t *entry = *_member;
+  if (addr < entry->ip_low)
+    return -1;
+  else if (addr > entry->ip_high)
+    return 1;
+  else
+    return 0;
+}
+
+/** Sorting helper: return -1, 1, or 0 based on comparison of two
+ * geoip_ipv6_entry_t */
+static int
+geoip_ipv6_compare_entries_(const void **_a, const void **_b)
+{
+  const geoip_ipv6_entry_t *a = *_a, *b = *_b;
+  return fast_memcmp(a->ip_low.s6_addr, b->ip_low.s6_addr,
+                     sizeof(struct in6_addr));
+}
+
+/** bsearch helper: return -1, 1, or 0 based on comparison of an IPv6
+ * (a pointer to a in6_addr) to a geoip_ipv6_entry_t */
+static int
+geoip_ipv6_compare_key_to_entry_(const void *_key, const void **_member)
+{
+  const struct in6_addr *addr = (struct in6_addr *)_key;
+  const geoip_ipv6_entry_t *entry = *_member;
+
+  if (fast_memcmp(addr->s6_addr, entry->ip_low.s6_addr,
+             sizeof(struct in6_addr)) < 0)
+    return -1;
+  else if (fast_memcmp(addr->s6_addr, entry->ip_high.s6_addr,
+                  sizeof(struct in6_addr)) > 0)
+    return 1;
+  else
+    return 0;
+}
+
+/** Set up a new list of geoip countries with no countries (yet) set in it,
+ * except for the unknown country.
+ */
+static void
+init_geoip_countries(void)
+{
+  geoip_country_t *geoip_unresolved;
+  geoip_countries = smartlist_new();
+  /* Add a geoip_country_t for requests that could not be resolved to a
+   * country as first element (index 0) to geoip_countries. */
+  geoip_unresolved = tor_malloc_zero(sizeof(geoip_country_t));
+  strlcpy(geoip_unresolved->countrycode, "??",
+          sizeof(geoip_unresolved->countrycode));
+  smartlist_add(geoip_countries, geoip_unresolved);
+  country_idxplus1_by_lc_code = strmap_new();
+  strmap_set_lc(country_idxplus1_by_lc_code, "??", (void*)(1));
+}
+
+/** Clear appropriate GeoIP database, based on <b>family</b>, and
+ * reload it from the file <b>filename</b>. Return 0 on success, -1 on
+ * failure.
+ *
+ * Recognized line formats for IPv4 are:
+ *   INTIPLOW,INTIPHIGH,CC
+ * and
+ *   "INTIPLOW","INTIPHIGH","CC","CC3","COUNTRY NAME"
+ * where INTIPLOW and INTIPHIGH are IPv4 addresses encoded as 4-byte unsigned
+ * integers, and CC is a country code.
+ *
+ * Recognized line format for IPv6 is:
+ *   IPV6LOW,IPV6HIGH,CC
+ * where IPV6LOW and IPV6HIGH are IPv6 addresses and CC is a country code.
+ *
+ * It also recognizes, and skips over, blank lines and lines that start
+ * with '#' (comments).
+ */
+int
+geoip_load_file(sa_family_t family, const char *filename, int severity)
+{
+  FILE *f;
+  crypto_digest_t *geoip_digest_env = NULL;
+
+  tor_assert(family == AF_INET || family == AF_INET6);
+
+  if (!(f = tor_fopen_cloexec(filename, "r"))) {
+    log_fn(severity, LD_GENERAL, "Failed to open GEOIP file %s.",
+           filename);
+    return -1;
+  }
+  if (!geoip_countries)
+    init_geoip_countries();
+
+  if (family == AF_INET) {
+    if (geoip_ipv4_entries) {
+      SMARTLIST_FOREACH(geoip_ipv4_entries, geoip_ipv4_entry_t *, e,
+                        tor_free(e));
+      smartlist_free(geoip_ipv4_entries);
+    }
+    geoip_ipv4_entries = smartlist_new();
+  } else { /* AF_INET6 */
+    if (geoip_ipv6_entries) {
+      SMARTLIST_FOREACH(geoip_ipv6_entries, geoip_ipv6_entry_t *, e,
+                        tor_free(e));
+      smartlist_free(geoip_ipv6_entries);
+    }
+    geoip_ipv6_entries = smartlist_new();
+  }
+  geoip_digest_env = crypto_digest_new();
+
+  log_notice(LD_GENERAL, "Parsing GEOIP %s file %s.",
+             (family == AF_INET) ? "IPv4" : "IPv6", filename);
+  while (!feof(f)) {
+    char buf[512];
+    if (fgets(buf, (int)sizeof(buf), f) == NULL)
+      break;
+    crypto_digest_add_bytes(geoip_digest_env, buf, strlen(buf));
+    /* FFFF track full country name. */
+    geoip_parse_entry(buf, family);
+  }
+  /*XXXX abort and return -1 if no entries/illformed?*/
+  fclose(f);
+
+  /* Sort list and remember file digests so that we can include it in
+   * our extra-info descriptors. */
+  if (family == AF_INET) {
+    smartlist_sort(geoip_ipv4_entries, geoip_ipv4_compare_entries_);
+    crypto_digest_get_digest(geoip_digest_env, geoip_digest, DIGEST_LEN);
+  } else {
+    /* AF_INET6 */
+    smartlist_sort(geoip_ipv6_entries, geoip_ipv6_compare_entries_);
+    crypto_digest_get_digest(geoip_digest_env, geoip6_digest, DIGEST_LEN);
+  }
+  crypto_digest_free(geoip_digest_env);
+
+  return 0;
+}
+
+/** Given an IP address in host order, return a number representing the
+ * country to which that address belongs, -1 for "No geoip information
+ * available", or 0 for the 'unknown country'.  The return value will always
+ * be less than geoip_get_n_countries().  To decode it, call
+ * geoip_get_country_name().
+ */
+int
+geoip_get_country_by_ipv4(uint32_t ipaddr)
+{
+  geoip_ipv4_entry_t *ent;
+  if (!geoip_ipv4_entries)
+    return -1;
+  ent = smartlist_bsearch(geoip_ipv4_entries, &ipaddr,
+                          geoip_ipv4_compare_key_to_entry_);
+  return ent ? (int)ent->country : 0;
+}
+
+/** Given an IPv6 address, return a number representing the country to
+ * which that address belongs, -1 for "No geoip information available", or
+ * 0 for the 'unknown country'.  The return value will always be less than
+ * geoip_get_n_countries().  To decode it, call geoip_get_country_name().
+ */
+int
+geoip_get_country_by_ipv6(const struct in6_addr *addr)
+{
+  geoip_ipv6_entry_t *ent;
+
+  if (!geoip_ipv6_entries)
+    return -1;
+  ent = smartlist_bsearch(geoip_ipv6_entries, addr,
+                          geoip_ipv6_compare_key_to_entry_);
+  return ent ? (int)ent->country : 0;
+}
+
+/** Given an IP address, return a number representing the country to which
+ * that address belongs, -1 for "No geoip information available", or 0 for
+ * the 'unknown country'.  The return value will always be less than
+ * geoip_get_n_countries().  To decode it, call geoip_get_country_name().
+ */
+MOCK_IMPL(int,
+geoip_get_country_by_addr,(const tor_addr_t *addr))
+{
+  if (tor_addr_family(addr) == AF_INET) {
+    return geoip_get_country_by_ipv4(tor_addr_to_ipv4h(addr));
+  } else if (tor_addr_family(addr) == AF_INET6) {
+    return geoip_get_country_by_ipv6(tor_addr_to_in6(addr));
+  } else {
+    return -1;
+  }
+}
+
+/** Return the number of countries recognized by the GeoIP country list. */
+MOCK_IMPL(int,
+geoip_get_n_countries,(void))
+{
+  if (!geoip_countries)
+    init_geoip_countries();
+  return (int) smartlist_len(geoip_countries);
+}
+
+/** Return the two-letter country code associated with the number <b>num</b>,
+ * or "??" for an unknown value. */
+const char *
+geoip_get_country_name(country_t num)
+{
+  if (geoip_countries && num >= 0 && num < smartlist_len(geoip_countries)) {
+    geoip_country_t *c = smartlist_get(geoip_countries, num);
+    return c->countrycode;
+  } else
+    return "??";
+}
+
+/** Return true iff we have loaded a GeoIP database.*/
+MOCK_IMPL(int,
+geoip_is_loaded,(sa_family_t family))
+{
+  tor_assert(family == AF_INET || family == AF_INET6);
+  if (geoip_countries == NULL)
+    return 0;
+  if (family == AF_INET)
+    return geoip_ipv4_entries != NULL;
+  else                          /* AF_INET6 */
+    return geoip_ipv6_entries != NULL;
+}
+
+/** Return the hex-encoded SHA1 digest of the loaded GeoIP file. The
+ * result does not need to be deallocated, but will be overwritten by the
+ * next call of hex_str(). */
+const char *
+geoip_db_digest(sa_family_t family)
+{
+  tor_assert(family == AF_INET || family == AF_INET6);
+  if (family == AF_INET)
+    return hex_str(geoip_digest, DIGEST_LEN);
+  else                          /* AF_INET6 */
+    return hex_str(geoip6_digest, DIGEST_LEN);
+}
+
+/** Release all storage held by the GeoIP databases and country list. */
+STATIC void
+clear_geoip_db(void)
+{
+  if (geoip_countries) {
+    SMARTLIST_FOREACH(geoip_countries, geoip_country_t *, c, tor_free(c));
+    smartlist_free(geoip_countries);
+  }
+
+  strmap_free(country_idxplus1_by_lc_code, NULL);
+  if (geoip_ipv4_entries) {
+    SMARTLIST_FOREACH(geoip_ipv4_entries, geoip_ipv4_entry_t *, ent,
+                      tor_free(ent));
+    smartlist_free(geoip_ipv4_entries);
+  }
+  if (geoip_ipv6_entries) {
+    SMARTLIST_FOREACH(geoip_ipv6_entries, geoip_ipv6_entry_t *, ent,
+                      tor_free(ent));
+    smartlist_free(geoip_ipv6_entries);
+  }
+  geoip_countries = NULL;
+  country_idxplus1_by_lc_code = NULL;
+  geoip_ipv4_entries = NULL;
+  geoip_ipv6_entries = NULL;
+}
+
+/** Release all storage held in this file. */
+void
+geoip_free_all(void)
+{
+  clear_geoip_db();
+
+  memset(geoip_digest, 0, sizeof(geoip_digest));
+  memset(geoip6_digest, 0, sizeof(geoip6_digest));
+}
diff --git a/src/lib/geoip/geoip.h b/src/lib/geoip/geoip.h
new file mode 100644
index 0000000000..6ef27d66d0
--- /dev/null
+++ b/src/lib/geoip/geoip.h
@@ -0,0 +1,50 @@
+/* Copyright (c) 2001 Matej Pfajfar.
+ * Copyright (c) 2001-2004, Roger Dingledine.
+ * Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
+ * Copyright (c) 2007-2018, The Tor Project, Inc. */
+/* See LICENSE for licensing information */
+
+/**
+ * \file geoip.h
+ * \brief Header file for geoip.c.
+ **/
+
+#ifndef TOR_GEOIP_H
+#define TOR_GEOIP_H
+
+#include "orconfig.h"
+#include "lib/net/nettypes.h"
+#include "lib/testsupport/testsupport.h"
+#include "lib/net/inaddr_st.h"
+#include "lib/geoip/country.h"
+
+#ifdef GEOIP_PRIVATE
+STATIC int geoip_parse_entry(const char *line, sa_family_t family);
+STATIC void clear_geoip_db(void);
+#endif /* defined(GEOIP_PRIVATE) */
+
+struct in6_addr;
+struct tor_addr_t;
+
+int geoip_get_country_by_ipv4(uint32_t ipaddr);
+int geoip_get_country_by_ipv6(const struct in6_addr *addr);
+
+/** A per-country GeoIP record. */
+typedef struct geoip_country_t {
+  char countrycode[3];
+} geoip_country_t;
+
+struct smartlist_t;
+const struct smartlist_t *geoip_get_countries(void);
+
+int geoip_load_file(sa_family_t family, const char *filename, int severity);
+MOCK_DECL(int, geoip_get_country_by_addr, (const struct tor_addr_t *addr));
+MOCK_DECL(int, geoip_get_n_countries, (void));
+const char *geoip_get_country_name(country_t num);
+MOCK_DECL(int, geoip_is_loaded, (sa_family_t family));
+const char *geoip_db_digest(sa_family_t family);
+MOCK_DECL(country_t, geoip_get_country, (const char *countrycode));
+
+void geoip_free_all(void);
+
+#endif /* !defined(TOR_GEOIP_H) */
diff --git a/src/lib/geoip/include.am b/src/lib/geoip/include.am
new file mode 100644
index 0000000000..9710d75ac7
--- /dev/null
+++ b/src/lib/geoip/include.am
@@ -0,0 +1,17 @@
+noinst_LIBRARIES += src/lib/libtor-geoip.a
+
+if UNITTESTS_ENABLED
+noinst_LIBRARIES += src/lib/libtor-geoip-testing.a
+endif
+
+src_lib_libtor_geoip_a_SOURCES =			\
+	src/lib/geoip/geoip.c
+
+src_lib_libtor_geoip_testing_a_SOURCES = \
+	$(src_lib_libtor_geoip_a_SOURCES)
+src_lib_libtor_geoip_testing_a_CPPFLAGS = $(AM_CPPFLAGS) $(TEST_CPPFLAGS)
+src_lib_libtor_geoip_testing_a_CFLAGS = $(AM_CFLAGS) $(TEST_CFLAGS)
+
+noinst_HEADERS +=					\
+	src/lib/geoip/geoip.h   			\
+	src/lib/geoip/country.h
diff --git a/src/test/test_config.c b/src/test/test_config.c
index 77c0d248c2..dae4d83766 100644
--- a/src/test/test_config.c
+++ b/src/test/test_config.c
@@ -29,7 +29,7 @@
 #include "feature/client/entrynodes.h"
 #include "feature/client/transports.h"
 #include "feature/relay/ext_orport.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "feature/hibernate/hibernate.h"
 #include "core/mainloop/mainloop.h"
 #include "feature/nodelist/networkstatus.h"
diff --git a/src/test/test_dir_handle_get.c b/src/test/test_dir_handle_get.c
index f05b0fc79f..623d9e9b9d 100644
--- a/src/test/test_dir_handle_get.c
+++ b/src/test/test_dir_handle_get.c
@@ -32,7 +32,8 @@
 #include "feature/nodelist/routerparse.h"
 #include "feature/nodelist/networkstatus.h"
 #include "core/proto/proto_http.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "feature/dircache/dirserv.h"
 #include "feature/dirauth/dirvote.h"
 #include "test/log_test_helpers.h"
diff --git a/src/test/test_dos.c b/src/test/test_dos.c
index b411e7b38a..40a4c6ba29 100644
--- a/src/test/test_dos.c
+++ b/src/test/test_dos.c
@@ -9,7 +9,7 @@
 #include "core/or/dos.h"
 #include "core/or/circuitlist.h"
 #include "lib/crypt_ops/crypto_rand.h"
-#include "feature/stats/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "core/or/channel.h"
 #include "feature/nodelist/microdesc.h"
 #include "feature/nodelist/networkstatus.h"
@@ -500,4 +500,3 @@ struct testcase_t dos_tests[] = {
     NULL, NULL },
   END_OF_TESTCASES
 };
-
diff --git a/src/test/test_geoip.c b/src/test/test_geoip.c
index 9df8ea7988..6f9c39063b 100644
--- a/src/test/test_geoip.c
+++ b/src/test/test_geoip.c
@@ -10,7 +10,8 @@
 #define GEOIP_PRIVATE
 #include "core/or/or.h"
 #include "app/config/config.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
+#include "feature/stats/geoip_stats.h"
 #include "test/test.h"
 
   /* Record odd numbered fake-IPs using ipv6, even numbered fake-IPs
@@ -403,7 +404,8 @@ test_geoip_load_file(void *arg)
 
   /* A nonexistant filename should fail. */
   tt_int_op(-1, OP_EQ,
-            geoip_load_file(AF_INET, "/you/did/not/put/a/file/here/I/hope"));
+            geoip_load_file(AF_INET, "/you/did/not/put/a/file/here/I/hope",
+                            LOG_INFO));
 
   /* We start out with only "Ningunpartia" in the database. */
   tt_int_op(1, OP_EQ, geoip_get_n_countries());
@@ -417,7 +419,7 @@ test_geoip_load_file(void *arg)
   const char *fname = get_fname("geoip");
   tt_int_op(0, OP_EQ, write_str_to_file(fname, GEOIP_CONTENT, 1));
 
-  int rv = geoip_load_file(AF_INET, fname);
+  int rv = geoip_load_file(AF_INET, fname, LOG_WARN);
   if (rv != 0) {
     TT_GRIPE(("Unable to load geoip from %s", escaped(fname)));
   }
@@ -467,7 +469,8 @@ test_geoip6_load_file(void *arg)
 
   /* A nonexistant filename should fail. */
   tt_int_op(-1, OP_EQ,
-            geoip_load_file(AF_INET6, "/you/did/not/put/a/file/here/I/hope"));
+            geoip_load_file(AF_INET6, "/you/did/not/put/a/file/here/I/hope",
+                            LOG_INFO));
 
   /* Any lookup attempt should say "-1" because we have no info */
   tor_inet_pton(AF_INET6, "2001:4860:4860::8888", &iaddr6);
@@ -493,7 +496,7 @@ test_geoip6_load_file(void *arg)
     "2001:4878:205::,2001:4878:214:ffff:ffff:ffff:ffff:ffff,US\n";
   tt_int_op(0, OP_EQ, write_str_to_file(fname6, CONTENT, 1));
 
-  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET6, fname6));
+  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET6, fname6, LOG_WARN));
 
   /* Check that we loaded some countries; this will fail if there are ever
    * fewer than 5 countries in our test data above. */
@@ -545,11 +548,11 @@ test_geoip_load_2nd_file(void *arg)
   tt_int_op(0, OP_EQ, write_str_to_file(fname_empty, "\n", 1));
 
   /* Load 1st geoip file */
-  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET, fname_geoip));
+  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET, fname_geoip, LOG_WARN));
 
   /* Load 2nd geoip (empty) file */
   /* It has to be the same IP address family */
-  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET, fname_empty));
+  tt_int_op(0, OP_EQ, geoip_load_file(AF_INET, fname_empty, LOG_WARN));
 
   /* Check that there is no geoip information for 8.8.8.8, */
   /* since loading the empty 2nd file should have delete it. */
diff --git a/src/test/test_options.c b/src/test/test_options.c
index a4de0e992a..f14e620eeb 100644
--- a/src/test/test_options.c
+++ b/src/test/test_options.c
@@ -8,7 +8,7 @@
 #include "app/config/confparse.h"
 #include "app/config/config.h"
 #include "test/test.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 
 #define ROUTERSET_PRIVATE
 #include "feature/nodelist/routerset.h"
diff --git a/src/test/test_routerset.c b/src/test/test_routerset.c
index 2017ef0050..fede419f0b 100644
--- a/src/test/test_routerset.c
+++ b/src/test/test_routerset.c
@@ -4,7 +4,7 @@
 #define ROUTERSET_PRIVATE
 
 #include "core/or/or.h"
-#include "feature/stats/geoip.h"
+#include "lib/geoip/geoip.h"
 #include "feature/nodelist/routerset.h"
 #include "feature/nodelist/routerparse.h"
 #include "core/or/policies.h"
diff --git a/src/test/testing_common.c b/src/test/testing_common.c
index 2d00ecb651..c52683afca 100644
--- a/src/test/testing_common.c
+++ b/src/test/testing_common.c
@@ -16,6 +16,7 @@
 #include "lib/crypt_ops/crypto_dh.h"
 #include "lib/crypt_ops/crypto_ed25519.h"
 #include "lib/crypt_ops/crypto_rand.h"
+#include "feature/stats/predict_ports.h"
 #include "feature/stats/rephist.h"
 #include "lib/err/backtrace.h"
 #include "test/test.h"