https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=5846
--- Comment #8 from Alexis La Goutte <alexis.lagoutte@xxxxxxxxx> 2011-04-20 01:32:47 PDT ---
(In reply to comment #6)
>
> OK: I've looked at the libutp utp.cpp and see that the V2 (and greater) header
> format is quite different: somewhat similar fields but different sizes/offsets.
>
> The V2 format matches your capture so I'll update the dissector tomorrow to
> dissect the header according to the version (unless you've already started).
Hi Bill,
I not found a V2 (and greader) header, only a v0(?) Header
struct PacketFormat {
// connection ID
uint32_big connid;
uint32_big tv_sec;
...
}
and there is no field for type and version in v0 Header.
Also the Libutp don't respect the spec ? (about type and version)
BEP-0029 Spec :
0 4 8 16 24 32
+-------+-------+---------------+---------------+---------------+
| ver | type | extension | connection_id |
+-------+-------+---------------+---------------+---------------+
It is indicated : All fields are in network byte order (big endian).
LibUTP Source :
struct PacketFormatV1 {
// packet_type (4 high bits)
// protocol version (4 low bits)
byte ver_type;
byte version() const { return ver_type & 0xf; }
byte type() const { return ver_type >> 4; }
void set_version(byte v) { ver_type = (ver_type & 0xf0) | (v & 0xf); }
void set_type(byte t) { ver_type = (ver_type & 0xf) | (t << 4); }
The field is inverted ?
The problem comes from following commit
https://github.com/bittorrent/libutp/commit/931f01252ef66e8263bfda09a629ff07e40538c0#utp.cpp
(It not in Expert in Big/Little Endian...)
--
Configure bugmail: https://bugs.wireshark.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
