|
@@ -1898,8 +1898,8 @@ rend_service_parse_intro_for_v3(
|
|
|
}
|
|
|
}
|
|
|
|
|
|
-
|
|
|
- if (plaintext_len < (size_t)(ts_offset)) {
|
|
|
+
|
|
|
+ if (plaintext_len < (size_t)(ts_offset)+4) {
|
|
|
if (err_msg_out) {
|
|
|
tor_asprintf(err_msg_out,
|
|
|
"truncated plaintext of encrypted parted of "
|
|
@@ -1922,12 +1922,6 @@ rend_service_parse_intro_for_v3(
|
|
|
memcpy(intro->u.v3.auth_data, buf + 4, intro->u.v3.auth_len);
|
|
|
}
|
|
|
|
|
|
-
|
|
|
- * Apparently we don't use the timestamp any more, but might as well copy
|
|
|
- * over just in case we ever care about it.
|
|
|
- */
|
|
|
- intro->u.v3.timestamp = ntohl(get_uint32(buf + ts_offset));
|
|
|
-
|
|
|
|
|
|
* From here on, the format is as in v2, so we call the v2 parser with
|
|
|
* adjusted buffer and length. We are 4 + ts_offset octets in, but the
|