Merge pull request #893 from dweinstein/sni-ignore

Add ignore based on TLS ClientHello SNI

4 files changed, 98 insertions, 58 deletions

diff --git a/docs/features/passthrough.rst b/docs/features/passthrough.rst
index 80521393..b7b5df84 100644
--- a/docs/features/passthrough.rst
+++ b/docs/features/passthrough.rst
@@ -31,9 +31,9 @@ mitmproxy allows you to specify a regex which is matched against a ``host:port``
 
 There are two important quirks to consider:
 
-- **In transparent mode, the ignore pattern is matched against the IP.** While we usually infer the
+- **In transparent mode, the ignore pattern is matched against the IP and ClientHello SNI host.** While we usually infer the
   hostname from the Host header if the :option:`--host` argument is passed to mitmproxy, we do not
-  have access to this information before the SSL handshake.
+  have access to this information before the SSL handshake. If the client uses SNI however, then we treat the SNI host as an ignore target.
 - In regular mode, explicit HTTP requests are never ignored. [#explicithttp]_ The ignore pattern is
   applied on CONNECT requests, which initiate HTTPS or clear-text WebSocket connections.
 
diff --git a/libmproxy/protocol/__init__.py b/libmproxy/protocol/__init__.py
index 0d624fd7..d8ebd4f0 100644
--- a/libmproxy/protocol/__init__.py
+++ b/libmproxy/protocol/__init__.py
@@ -28,12 +28,12 @@ as late as possible; this makes server replay without any outgoing connections p
 from __future__ import (absolute_import, print_function, division)
 from .base import Layer, ServerConnectionMixin, Kill
 from .http import Http1Layer, UpstreamConnectLayer, Http2Layer
-from .tls import TlsLayer, is_tls_record_magic
+from .tls import TlsLayer, is_tls_record_magic, TlsClientHello
 from .rawtcp import RawTCPLayer
 
 __all__ = [
     "Layer", "ServerConnectionMixin", "Kill",
     "Http1Layer", "UpstreamConnectLayer", "Http2Layer",
-    "TlsLayer", "is_tls_record_magic",
+    "TlsLayer", "is_tls_record_magic", "TlsClientHello"
     "RawTCPLayer"
 ]
diff --git a/libmproxy/protocol/tls.py b/libmproxy/protocol/tls.py
index ed747643..6d4cac85 100644
--- a/libmproxy/protocol/tls.py
+++ b/libmproxy/protocol/tls.py
@@ -221,6 +221,80 @@ def is_tls_record_magic(d):
         d[2] in ('\x00', '\x01', '\x02', '\x03')
     )
 
+def get_client_hello(client_conn):
+    """
+    Peek into the socket and read all records that contain the initial client hello message.
+
+    client_conn:
+        The :py:class:`client connection <libmproxy.models.ClientConnection>`.
+
+    Returns:
+        The raw handshake packet bytes, without TLS record header(s).
+    """
+    client_hello = ""
+    client_hello_size = 1
+    offset = 0
+    while len(client_hello) < client_hello_size:
+        record_header = client_conn.rfile.peek(offset + 5)[offset:]
+        if not is_tls_record_magic(record_header) or len(record_header) != 5:
+            raise TlsProtocolException('Expected TLS record, got "%s" instead.' % record_header)
+        record_size = struct.unpack("!H", record_header[3:])[0] + 5
+        record_body = client_conn.rfile.peek(offset + record_size)[offset + 5:]
+        if len(record_body) != record_size - 5:
+            raise TlsProtocolException("Unexpected EOF in TLS handshake: %s" % record_body)
+        client_hello += record_body
+        offset += record_size
+        client_hello_size = struct.unpack("!I", '\x00' + client_hello[1:4])[0] + 4
+    return client_hello
+
+class TlsClientHello(object):
+    def __init__(self, raw_client_hello):
+        self._client_hello = ClientHello.parse(raw_client_hello)
+
+    def raw(self):
+        return self._client_hello
+
+    @property
+    def client_cipher_suites(self):
+        return self._client_hello.cipher_suites.cipher_suites
+
+    @property
+    def client_sni(self):
+        for extension in self._client_hello.extensions:
+            if (extension.type == 0x00 and len(extension.server_names) == 1
+                    and extension.server_names[0].type == 0):
+                return extension.server_names[0].name
+
+    @property
+    def client_alpn_protocols(self):
+        for extension in self._client_hello.extensions:
+            if extension.type == 0x10:
+                return list(extension.alpn_protocols)
+
+    @classmethod
+    def from_client_conn(cls, client_conn):
+        """
+        Peek into the connection, read the initial client hello and parse it to obtain ALPN values.
+        client_conn:
+            The :py:class:`client connection <libmproxy.models.ClientConnection>`.
+        Returns:
+            :py:class:`client hello <libmproxy.protocol.tls.TlsClientHello>`.
+        """
+        try:
+            raw_client_hello = get_client_hello(client_conn)[4:]  # exclude handshake header.
+        except ProtocolException as e:
+            raise TlsProtocolException('Cannot read raw Client Hello: %s' % repr(e))
+
+        try:
+            return cls(raw_client_hello)
+        except ConstructError as e:
+            raise TlsProtocolException('Cannot parse Client Hello: %s, Raw Client Hello: %s' % \
+                            (repr(e), raw_client_hello.encode("hex")))
+
+    def __repr__(self):
+        return "TlsClientHello( sni: %s alpn_protocols: %s,  cipher_suites: %s)" % \
+                (self.client_sni, self.client_alpn_protocols, self.client_cipher_suites)
+
 
 class TlsLayer(Layer):
     def __init__(self, ctx, client_tls, server_tls):
@@ -281,60 +355,18 @@ class TlsLayer(Layer):
         else:
             return "TlsLayer(inactive)"
 
-    def _get_client_hello(self):
-        """
-        Peek into the socket and read all records that contain the initial client hello message.
-
-        Returns:
-            The raw handshake packet bytes, without TLS record header(s).
-        """
-        client_hello = ""
-        client_hello_size = 1
-        offset = 0
-        while len(client_hello) < client_hello_size:
-            record_header = self.client_conn.rfile.peek(offset + 5)[offset:]
-            if not is_tls_record_magic(record_header) or len(record_header) != 5:
-                raise TlsProtocolException('Expected TLS record, got "%s" instead.' % record_header)
-            record_size = struct.unpack("!H", record_header[3:])[0] + 5
-            record_body = self.client_conn.rfile.peek(offset + record_size)[offset + 5:]
-            if len(record_body) != record_size - 5:
-                raise TlsProtocolException("Unexpected EOF in TLS handshake: %s" % record_body)
-            client_hello += record_body
-            offset += record_size
-            client_hello_size = struct.unpack("!I", '\x00' + client_hello[1:4])[0] + 4
-        return client_hello
 
     def _parse_client_hello(self):
         """
         Peek into the connection, read the initial client hello and parse it to obtain ALPN values.
         """
         try:
-            raw_client_hello = self._get_client_hello()[4:]  # exclude handshake header.
-        except ProtocolException as e:
-            self.log("Cannot parse Client Hello: %s" % repr(e), "error")
-            return
-
-        try:
-            client_hello = ClientHello.parse(raw_client_hello)
-        except ConstructError as e:
+            parsed = TlsClientHello.from_client_conn(self.client_conn)
+            self.client_sni = parsed.client_sni
+            self.client_alpn_protocols = parsed.client_alpn_protocols
+            self.client_ciphers  = parsed.client_cipher_suites
+        except TlsProtocolException as e:
             self.log("Cannot parse Client Hello: %s" % repr(e), "error")
-            self.log("Raw Client Hello: %s" % raw_client_hello.encode("hex"), "debug")
-            return
-
-        self.client_ciphers = client_hello.cipher_suites.cipher_suites
-
-        for extension in client_hello.extensions:
-            if extension.type == 0x00:
-                if len(extension.server_names) != 1 or extension.server_names[0].type != 0:
-                    self.log("Unknown Server Name Indication: %s" % extension.server_names, "error")
-                self.client_sni = extension.server_names[0].name
-            elif extension.type == 0x10:
-                self.client_alpn_protocols = list(extension.alpn_protocols)
-
-        self.log(
-            "Parsed Client Hello: sni=%s, alpn=%s" % (self.client_sni, self.client_alpn_protocols),
-            "debug"
-        )
 
     def connect(self):
         if not self.server_conn:
diff --git a/libmproxy/proxy/root_context.py b/libmproxy/proxy/root_context.py
index f62b0c8e..d70fc299 100644
--- a/libmproxy/proxy/root_context.py
+++ b/libmproxy/proxy/root_context.py
@@ -4,15 +4,14 @@ import sys
 
 import six
 
-from libmproxy.exceptions import ProtocolException
+from libmproxy.exceptions import ProtocolException, TlsProtocolException
 from netlib.exceptions import TcpException
 from ..protocol import (
     RawTCPLayer, TlsLayer, Http1Layer, Http2Layer, is_tls_record_magic, ServerConnectionMixin,
-    UpstreamConnectLayer
+    UpstreamConnectLayer, TlsClientHello
 )
 from .modes import HttpProxy, HttpUpstreamProxy, ReverseProxy
 
-
 class RootContext(object):
     """
     The outermost context provided to the root layer.
@@ -48,16 +47,25 @@ class RootContext(object):
         return self.channel.ask("next_layer", layer)
 
     def _next_layer(self, top_layer):
-        # 1. Check for --ignore.
-        if self.config.check_ignore(top_layer.server_conn.address):
-            return RawTCPLayer(top_layer, logging=False)
-
         try:
             d = top_layer.client_conn.rfile.peek(3)
         except TcpException as e:
             six.reraise(ProtocolException, ProtocolException(str(e)), sys.exc_info()[2])
         client_tls = is_tls_record_magic(d)
 
+        # 1. check for --ignore
+        if self.config.check_ignore:
+            ignore = self.config.check_ignore(top_layer.server_conn.address)
+            if not ignore and client_tls:
+                try:
+                    client_hello = TlsClientHello.from_client_conn(self.client_conn)
+                except TlsProtocolException as e:
+                    self.log("Cannot parse Client Hello: %s" % repr(e), "error")
+                else:
+                    ignore = self.config.check_ignore((client_hello.client_sni, 443))
+            if ignore:
+                return RawTCPLayer(top_layer, logging=False)
+
         # 2. Always insert a TLS layer, even if there's neither client nor server tls.
         # An inline script may upgrade from http to https,
         # in which case we need some form of TLS layer.