fixed crash on invalid input in http_parser

ChangeLog

@@ -1,5 +1,6 @@
 1.1.1 release
 
+	* fixed crash on invalid input in http_parser
 	* upgraded to libtommath 1.0
 	* fixed parsing of IPv6 endpoint with invalid port character separator
 	* added limited support for new x.pe parameter from BEP 9

include/libtorrent/add_torrent_params.hpp

@@ -306,7 +306,7 @@ namespace libtorrent
 		// also be stored in resume data. If you want the save path saved in
 		// the resume data to be used, you need to set the
 		// flag_use_resume_save_path flag.
-		//
+		// 
 		// .. note::
 		// 	On windows this path (and other paths) are interpreted as UNC
 		// 	paths. This means they must use backslashes as directory separators
@@ -377,6 +377,7 @@ namespace libtorrent
 
 		// flags controlling aspects of this torrent and how it's added. See
 		// flags_t for details.
+		// 
 		// .. note::
 		// 	The ``flags`` field is initialized with default flags by the
 		// 	constructor. In order to preserve default behavior when clearing or

src/http_parser.cpp

@@ -174,6 +174,7 @@ namespace libtorrent
 		if (m_state == read_status)
 		{
 			TORRENT_ASSERT(!m_finished);
+			TORRENT_ASSERT(pos <= recv_buffer.end);
 			char const* newline = std::find(pos, recv_buffer.end, '\n');
 			// if we don't have a full line yet, wait.
 			if (newline == recv_buffer.end)
@@ -194,6 +195,7 @@ namespace libtorrent
 
 			char const* line = pos;
 			++newline;
+			TORRENT_ASSERT(newline >= pos);
 			int incoming = int(newline - pos);
 			m_recv_pos += incoming;
 			boost::get<1>(ret) += newline - (m_recv_buffer.begin + start_pos);
@@ -227,6 +229,7 @@ namespace libtorrent
 		if (m_state == read_header)
 		{
 			TORRENT_ASSERT(!m_finished);
+			TORRENT_ASSERT(pos <= recv_buffer.end);
 			char const* newline = std::find(pos, recv_buffer.end, '\n');
 			std::string line;
 
@@ -277,6 +280,12 @@ namespace libtorrent
 				if (name == "content-length")
 				{
 					m_content_length = strtoll(value.c_str(), 0, 10);
+					if (m_content_length < 0)
+					{
+						m_state = error_state;
+						error = true;
+						return ret;
+					}
 				}
 				else if (name == "connection")
 				{
@@ -294,12 +303,24 @@ namespace libtorrent
 					if (string_begins_no_case("bytes ", ptr)) ptr += 6;
 					char* end;
 					m_range_start = strtoll(ptr, &end, 10);
+					if (m_range_start < 0)
+					{
+						m_state = error_state;
+						error = true;
+						return ret;
+					}
 					if (end == ptr) success = false;
 					else if (*end != '-') success = false;
 					else
 					{
 						ptr = end + 1;
 						m_range_end = strtoll(ptr, &end, 10);
+						if (m_range_end < 0)
+						{
+							m_state = error_state;
+							error = true;
+							return ret;
+						}
 						if (end == ptr) success = false;
 					}
 
@@ -318,6 +339,7 @@ namespace libtorrent
 				}
 
 				TORRENT_ASSERT(m_recv_pos <= recv_buffer.left());
+				TORRENT_ASSERT(pos <= recv_buffer.end);
 				newline = std::find(pos, recv_buffer.end, '\n');
 			}
 			boost::get<1>(ret) += newline - (m_recv_buffer.begin + start_pos);
@@ -347,6 +369,12 @@ namespace libtorrent
 					int header_size;
 					if (parse_chunk_header(buf, &chunk_size, &header_size))
 					{
+						if (chunk_size < 0)
+						{
+							m_state = error_state;
+							error = true;
+							return ret;
+						}
 						if (chunk_size > 0)
 						{
 							std::pair<boost::int64_t, boost::int64_t> chunk_range(m_cur_chunk_end + header_size
@@ -419,6 +447,7 @@ namespace libtorrent
 	bool http_parser::parse_chunk_header(buffer::const_interval buf
 		, boost::int64_t* chunk_size, int* header_size)
 	{
+		TORRENT_ASSERT(buf.begin <= buf.end);
 		char const* pos = buf.begin;
 
 		// ignore one optional new-line. This is since each chunk
@@ -429,6 +458,7 @@ namespace libtorrent
 		if (pos < buf.end && pos[0] == '\n') ++pos;
 		if (pos == buf.end) return false;
 
+		TORRENT_ASSERT(pos <= buf.end);
 		char const* newline = std::find(pos, buf.end, '\n');
 		if (newline == buf.end) return false;
 		++newline;
@@ -441,6 +471,8 @@ namespace libtorrent
 
 		// first, read the chunk length
 		*chunk_size = strtoll(pos, 0, 16);
+		if (*chunk_size < 0) return true;
+
 		if (*chunk_size != 0)
 		{
 			*header_size = newline - buf.begin;

test/test_http_parser.cpp

@@ -361,29 +361,6 @@ TORRENT_TEST(http_parser)
 		TEST_EQUAL(parser.headers().find("test2")->second, "bar");
 	}
 
-	// test chunked encoding
-
-	parser.reset();
-
-	char const* chunked_input =
-		"HTTP/1.1 200 OK\r\n"
-		"Transfer-Encoding: chunked\r\n"
-		"Content-Type: text/plain\r\n"
-		"\r\n"
-		"4\r\ntest\r\n4\r\n1234\r\n10\r\n0123456789abcdef\r\n"
-		"0\r\n\r\n";
-	received = feed_bytes(parser, chunked_input);
-
-	TEST_EQUAL(strlen(chunked_input), 24 + 94)
-	TEST_CHECK(received == make_tuple(24, 94, false));
-	TEST_CHECK(parser.finished());
-
-	char mutable_buffer[100];
-	memcpy(mutable_buffer, parser.get_body().begin, parser.get_body().left());
-	int len = parser.collapse_chunk_headers(mutable_buffer, parser.get_body().left());
-
-	TEST_CHECK(std::equal(mutable_buffer, mutable_buffer + len, "test12340123456789abcdef"));
-
 	// test url parsing
 
 	error_code ec;
@@ -476,3 +453,119 @@ TORRENT_TEST(http_parser)
 	TEST_EQUAL(is_redirect(400), false);
 }
 
+TORRENT_TEST(chunked_encoding)
+{
+	char const* chunked_input =
+		"HTTP/1.1 200 OK\r\n"
+		"Transfer-Encoding: chunked\r\n"
+		"Content-Type: text/plain\r\n"
+		"\r\n"
+		"4\r\ntest\r\n4\r\n1234\r\n10\r\n0123456789abcdef\r\n"
+		"0\r\n\r\n";
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, chunked_input);
+
+	TEST_EQUAL(strlen(chunked_input), 24 + 94)
+	TEST_CHECK(received == make_tuple(24, 94, false));
+	TEST_CHECK(parser.finished());
+
+	char mutable_buffer[100];
+	memcpy(mutable_buffer, parser.get_body().begin, parser.get_body().left());
+	int len = parser.collapse_chunk_headers(mutable_buffer, parser.get_body().left());
+
+	TEST_CHECK(std::equal(mutable_buffer, mutable_buffer + len, "test12340123456789abcdef"));
+}
+
+TORRENT_TEST(invalid_content_length)
+{
+	char const* chunked_input =
+		"HTTP/1.1 200 OK\r\n"
+		"Transfer-Encoding: chunked\r\n"
+		"Content-Length: -45345\r\n"
+		"\r\n";
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, chunked_input);
+
+	TEST_CHECK(boost::get<2>(received) == true);
+}
+
+TORRENT_TEST(invalid_chunked)
+{
+	char const* chunked_input =
+		"HTTP/1.1 200 OK\r\n"
+		"Transfer-Encoding: chunked\r\n"
+		"\r\n"
+		"-53465234545\r\n"
+		"foobar";
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, chunked_input);
+
+	TEST_CHECK(boost::get<2>(received) == true);
+}
+
+TORRENT_TEST(invalid_content_range_start)
+{
+	char const* chunked_input =
+		"HTTP/1.1 206 OK\n"
+		"Content-Range: bYTes -3-4\n"
+		"\n";
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, chunked_input);
+
+	TEST_CHECK(boost::get<2>(received) == true);
+}
+
+TORRENT_TEST(invalid_content_range_end)
+{
+	char const* chunked_input =
+		"HTTP/1.1 206 OK\n"
+		"Content-Range: bYTes 3--434\n"
+		"\n";
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, chunked_input);
+
+	TEST_CHECK(boost::get<2>(received) == true);
+}
+
+TORRENT_TEST(invalid_chunk_afl)
+{
+	boost::uint8_t const invalid_chunked_input[] = {
+		0x48, 0x6f, 0x54, 0x50, 0x2f, 0x31, 0x2e, 0x31, // HoTP/1.1 200 OK
+		0x20, 0x32, 0x30, 0x30, 0x20, 0x4f, 0x4b, 0x0d, // Cont-Length: 20
+		0x0a, 0x43, 0x6f, 0x6e, 0x74, 0x2d, 0x4c, 0x65, // Contente: tn
+		0x6e, 0x67, 0x74, 0x68, 0x3a, 0x20, 0x32, 0x30, // Transfer-Encoding: chunked
+		0x0d, 0x0a, 0x43, 0x6f, 0x6e, 0x74, 0x65, 0x6e, //
+		0x74, 0x65, 0x3a, 0x20, 0x74, 0x6e, 0x0d, 0x0a, //
+		0x54, 0x72, 0x61, 0x6e, 0x73, 0x66, 0x65, 0x72, //
+		0x2d, 0x45, 0x6e, 0x63, 0x6f, 0x64, 0x69, 0x6e, // -89abc9abcdef
+		0x67, 0x3a, 0x20, 0x63, 0x68, 0x75, 0x6e, 0x6b, // �
+		0x65, 0x64, 0x0d, 0x0a, 0x0d, 0x0d, 0x0a, 0x0d, // T����������def
+		0x0a, 0x0a, 0x2d, 0x38, 0x39, 0x61, 0x62, 0x63, // �
+		0x39, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x0d, // T�����������est-headyr: foobar
+		0x0a, 0xd6, 0x0d, 0x0a, 0x54, 0xbd, 0xbd, 0xbd,
+		0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0x64,
+		0x65, 0x66, 0x0d, 0x0a, 0xd6, 0x0d, 0x0a, 0x54,
+		0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0xbd, 0xbd,
+		0xbd, 0xbd, 0xbd, 0x65, 0x73, 0x74, 0x2d, 0x68,
+		0x65, 0x61, 0x64, 0x79, 0x72, 0x3a, 0x20, 0x66,
+		0x6f, 0x6f, 0x62, 0x61, 0x72, 0x0d, 0x0a, 0x0d,
+		0x0a, 0x00
+	};
+
+	http_parser parser;
+	boost::tuple<int, int, bool> const received
+		= feed_bytes(parser, reinterpret_cast<char const*>(invalid_chunked_input));
+
+	TEST_CHECK(boost::get<2>(received) == true);
+}
+