Merge branch 'develop' into anoa/blacklist_ip_ranges

* develop: (45 commits)
  URL preview blacklisting fixes (#5155)
  Revert 085ae346ac
  Add a DUMMY stage to captcha-only registration flow
  Make Prometheus snippet less confusing on the metrics collection doc (#4288)
  Set syslog identifiers in systemd units (#5023)
  Run Black on the tests again (#5170)
  Add AllowEncodedSlashes to apache (#5068)
  remove instructions for jessie installation (#5164)
  Run `black` on per_destination_queue
  Limit the number of EDUs in transactions to 100 as expected by receiver (#5138)
  Fix bogus imports in tests (#5154)
  add options to require an access_token to GET /profile and /publicRooms on CS API (#5083)
  Do checks on aliases for incoming m.room.aliases events (#5128)
  Remove the requirement to authenticate for /admin/server_version. (#5122)
  Fix spelling in server notices admin API docs (#5142)
  Fix sample config
  0.99.3.2
  include disco in deb build target list
  changelog
  Debian: we now need libpq-dev.
  ...

changelog.d/5043.feature

@@ -0,0 +1 @@
+Add ability to blacklist IP ranges for the federation client.

docs/sample_config.yaml

@@ -115,6 +115,24 @@ pid_file: DATADIR/homeserver.pid
 #  - nyc.example.com
 #  - syd.example.com
 
+# Prevent federation requests from being sent to the following
+# blacklist IP address CIDR ranges. If this option is not specified, or
+# specified with an empty list, no ip range blacklist will be enforced.
+#
+# (0.0.0.0 and :: are always blacklisted, whether or not they are explicitly
+# listed here, since they correspond to unroutable addresses.)
+#
+federation_ip_range_blacklist:
+  - '127.0.0.0/8'
+  - '10.0.0.0/8'
+  - '172.16.0.0/12'
+  - '192.168.0.0/16'
+  - '100.64.0.0/10'
+  - '169.254.0.0/16'
+  - '::1/128'
+  - 'fe80::/64'
+  - 'fc00::/7'
+
 # List of ports that Synapse should listen on, their purpose and their
 # configuration.
 #

synapse/config/server.py

@@ -17,6 +17,8 @@
 import logging
 import os.path
 
+from netaddr import IPSet
+
 from synapse.http.endpoint import parse_and_validate_server_name
 from synapse.python_dependencies import DependencyException, check_requirements
 
@@ -137,6 +139,24 @@ class ServerConfig(Config):
             for domain in federation_domain_whitelist:
                 self.federation_domain_whitelist[domain] = True
 
+        self.federation_ip_range_blacklist = config.get(
+            "federation_ip_range_blacklist", [],
+        )
+
+        # Attempt to create an IPSet from the given ranges
+        try:
+            self.federation_ip_range_blacklist = IPSet(
+                self.federation_ip_range_blacklist
+            )
+
+            # Always blacklist 0.0.0.0, ::
+            self.federation_ip_range_blacklist.update(["0.0.0.0", "::"])
+        except Exception as e:
+            raise ConfigError(
+                "Invalid range(s) provided in "
+                "federation_ip_range_blacklist: %s" % e
+            )
+
         if self.public_baseurl is not None:
             if self.public_baseurl[-1] != '/':
                 self.public_baseurl += '/'
@@ -386,6 +406,24 @@ class ServerConfig(Config):
         #  - nyc.example.com
         #  - syd.example.com
 
+        # Prevent federation requests from being sent to the following
+        # blacklist IP address CIDR ranges. If this option is not specified, or
+        # specified with an empty list, no ip range blacklist will be enforced.
+        #
+        # (0.0.0.0 and :: are always blacklisted, whether or not they are explicitly
+        # listed here, since they correspond to unroutable addresses.)
+        #
+        federation_ip_range_blacklist:
+          - '127.0.0.0/8'
+          - '10.0.0.0/8'
+          - '172.16.0.0/12'
+          - '192.168.0.0/16'
+          - '100.64.0.0/10'
+          - '169.254.0.0/16'
+          - '::1/128'
+          - 'fe80::/64'
+          - 'fc00::/7'
+
         # List of ports that Synapse should listen on, their purpose and their
         # configuration.
         #

synapse/http/client.py

@@ -165,7 +165,8 @@ class BlacklistingAgentWrapper(Agent):
                 ip_address, self._ip_whitelist, self._ip_blacklist
             ):
                 logger.info(
-                    "Blocking access to %s because of blacklist" % (ip_address,)
+                    "Blocking access to %s due to blacklist" %
+                    (ip_address,)
                 )
                 e = SynapseError(403, "IP address blocked by IP blacklist entry")
                 return defer.fail(Failure(e))
@@ -263,9 +264,6 @@ class SimpleHttpClient(object):
             uri (str): URI to query.
             data (bytes): Data to send in the request body, if applicable.
             headers (t.w.http_headers.Headers): Request headers.
-
-        Raises:
-            SynapseError: If the IP is blacklisted.
         """
         # A small wrapper around self.agent.request() so we can easily attach
         # counters to it

synapse/http/matrixfederationclient.py

@@ -27,9 +27,11 @@ import treq
 from canonicaljson import encode_canonical_json
 from prometheus_client import Counter
 from signedjson.sign import sign_json
+from zope.interface import implementer
 
 from twisted.internet import defer, protocol
 from twisted.internet.error import DNSLookupError
+from twisted.internet.interfaces import IReactorPluggableNameResolver
 from twisted.internet.task import _EPSILON, Cooperator
 from twisted.web._newclient import ResponseDone
 from twisted.web.http_headers import Headers
@@ -44,6 +46,7 @@ from synapse.api.errors import (
     SynapseError,
 )
 from synapse.http import QuieterFileBodyProducer
+from synapse.http.client import BlacklistingAgentWrapper, IPBlacklistingResolver
 from synapse.http.federation.matrix_federation_agent import MatrixFederationAgent
 from synapse.util.async_helpers import timeout_deferred
 from synapse.util.logcontext import make_deferred_yieldable
@@ -172,19 +175,44 @@ class MatrixFederationHttpClient(object):
         self.hs = hs
         self.signing_key = hs.config.signing_key[0]
         self.server_name = hs.hostname
-        reactor = hs.get_reactor()
+
+        real_reactor = hs.get_reactor()
+
+        # We need to use a DNS resolver which filters out blacklisted IP
+        # addresses, to prevent DNS rebinding.
+        nameResolver = IPBlacklistingResolver(
+            real_reactor, None, hs.config.federation_ip_range_blacklist,
+        )
+
+        @implementer(IReactorPluggableNameResolver)
+        class Reactor(object):
+            def __getattr__(_self, attr):
+                if attr == "nameResolver":
+                    return nameResolver
+                else:
+                    return getattr(real_reactor, attr)
+
+        self.reactor = Reactor()
 
         self.agent = MatrixFederationAgent(
-            hs.get_reactor(),
+            self.reactor,
             tls_client_options_factory,
         )
+
+        # Use a BlacklistingAgentWrapper to prevent circumventing the IP
+        # blacklist via IP literals in server names
+        self.agent = BlacklistingAgentWrapper(
+            self.agent, self.reactor,
+            ip_blacklist=hs.config.federation_ip_range_blacklist,
+        )
+
         self.clock = hs.get_clock()
         self._store = hs.get_datastore()
         self.version_string_bytes = hs.version_string.encode('ascii')
         self.default_timeout = 60
 
         def schedule(x):
-            reactor.callLater(_EPSILON, x)
+            self.reactor.callLater(_EPSILON, x)
 
         self._cooperator = Cooperator(scheduler=schedule)
 
@@ -370,7 +398,7 @@ class MatrixFederationHttpClient(object):
                             request_deferred = timeout_deferred(
                                 request_deferred,
                                 timeout=_sec_timeout,
-                                reactor=self.hs.get_reactor(),
+                                reactor=self.reactor,
                             )
 
                             response = yield request_deferred
@@ -397,7 +425,7 @@ class MatrixFederationHttpClient(object):
                         d = timeout_deferred(
                             d,
                             timeout=_sec_timeout,
-                            reactor=self.hs.get_reactor(),
+                            reactor=self.reactor,
                         )
 
                         try:
@@ -586,7 +614,7 @@ class MatrixFederationHttpClient(object):
         )
 
         body = yield _handle_json_response(
-            self.hs.get_reactor(), self.default_timeout, request, response,
+            self.reactor, self.default_timeout, request, response,
         )
 
         defer.returnValue(body)
@@ -645,7 +673,7 @@ class MatrixFederationHttpClient(object):
             _sec_timeout = self.default_timeout
 
         body = yield _handle_json_response(
-            self.hs.get_reactor(), _sec_timeout, request, response,
+            self.reactor, _sec_timeout, request, response,
         )
         defer.returnValue(body)
 
@@ -704,7 +732,7 @@ class MatrixFederationHttpClient(object):
         )
 
         body = yield _handle_json_response(
-            self.hs.get_reactor(), self.default_timeout, request, response,
+            self.reactor, self.default_timeout, request, response,
         )
 
         defer.returnValue(body)
@@ -753,7 +781,7 @@ class MatrixFederationHttpClient(object):
         )
 
         body = yield _handle_json_response(
-            self.hs.get_reactor(), self.default_timeout, request, response,
+            self.reactor, self.default_timeout, request, response,
         )
         defer.returnValue(body)
 
@@ -801,7 +829,7 @@ class MatrixFederationHttpClient(object):
 
         try:
             d = _readBodyToFile(response, output_stream, max_size)
-            d.addTimeout(self.default_timeout, self.hs.get_reactor())
+            d.addTimeout(self.default_timeout, self.reactor)
             length = yield make_deferred_yieldable(d)
         except Exception as e:
             logger.warn(

tests/http/test_fedclient.py

@@ -15,6 +15,8 @@
 
 from mock import Mock
 
+from netaddr import IPSet
+
 from twisted.internet import defer
 from twisted.internet.defer import TimeoutError
 from twisted.internet.error import ConnectingCancelledError, DNSLookupError
@@ -209,6 +211,75 @@ class FederationClientTests(HomeserverTestCase):
         self.assertIsInstance(f.value, RequestSendFailed)
         self.assertIsInstance(f.value.inner_exception, ResponseNeverReceived)
 
+    def test_client_ip_range_blacklist(self):
+        """Ensure that Synapse does not try to connect to blacklisted IPs"""
+
+        # Set up the ip_range blacklist
+        self.hs.config.federation_ip_range_blacklist = IPSet([
+            "127.0.0.0/8",
+            "fe80::/64",
+        ])
+        self.reactor.lookups["internal"] = "127.0.0.1"
+        self.reactor.lookups["internalv6"] = "fe80:0:0:0:0:8a2e:370:7337"
+        self.reactor.lookups["fine"] = "10.20.30.40"
+        cl = MatrixFederationHttpClient(self.hs, None)
+
+        # Try making a GET request to a blacklisted IPv4 address
+        # ------------------------------------------------------
+        # Make the request
+        d = cl.get_json("internal:8008", "foo/bar", timeout=10000)
+
+        # Nothing happened yet
+        self.assertNoResult(d)
+
+        self.pump(1)
+
+        # Check that it was unable to resolve the address
+        clients = self.reactor.tcpClients
+        self.assertEqual(len(clients), 0)
+
+        f = self.failureResultOf(d)
+        self.assertIsInstance(f.value, RequestSendFailed)
+        self.assertIsInstance(f.value.inner_exception, DNSLookupError)
+
+        # Try making a POST request to a blacklisted IPv6 address
+        # -------------------------------------------------------
+        # Make the request
+        d = cl.post_json("internalv6:8008", "foo/bar", timeout=10000)
+
+        # Nothing has happened yet
+        self.assertNoResult(d)
+
+        # Move the reactor forwards
+        self.pump(1)
+
+        # Check that it was unable to resolve the address
+        clients = self.reactor.tcpClients
+        self.assertEqual(len(clients), 0)
+
+        # Check that it was due to a blacklisted DNS lookup
+        f = self.failureResultOf(d, RequestSendFailed)
+        self.assertIsInstance(f.value.inner_exception, DNSLookupError)
+
+        # Try making a GET request to a non-blacklisted IPv4 address
+        # ----------------------------------------------------------
+        # Make the request
+        d = cl.post_json("fine:8008", "foo/bar", timeout=10000)
+
+        # Nothing has happened yet
+        self.assertNoResult(d)
+
+        # Move the reactor forwards
+        self.pump(1)
+
+        # Check that it was able to resolve the address
+        clients = self.reactor.tcpClients
+        self.assertNotEqual(len(clients), 0)
+
+        # Connection will still fail as this IP address does not resolve to anything
+        f = self.failureResultOf(d, RequestSendFailed)
+        self.assertIsInstance(f.value.inner_exception, ConnectingCancelledError)
+
     def test_client_gets_headers(self):
         """
         Once the client gets the headers, _request returns successfully.