On Tue, Jan 06, 2004 at 10:25:56PM -0800, Guy Harris wrote:
> On Thu, Dec 18, 2003 at 01:09:24AM +0100, Jan Kiszka wrote:
> > based on the work of Shaun Jackman, I started some months ago a new
> > approach to implement a dissector plugin which is able to analyze
> > infrared communication according to the IrDA specification.
>
> The IrLAP V1.1 specification gives, for the XID command, 0x2F (with the
> P/F bit ORed in); for the XID response, it gives 0xAF (again, with the
> P/F bit ORed in).
>
> However, that spec says IrLAP is "based on pre-existing standard
> asynchronous HDLC and SDLC half duplex protocols as used on multi-drop
> links (see references)", and one of the references is to ISO 4335
> (HDLC).
>
> 4335 has been withdrawn and replaced by ISO/IEC 13239:2002, which gives
> as the encoding for XID commands and responses 0xAF (with the P/F bit ORed)
> in.
>
> Does IrLAP really use a different encoding for XID commands than does
> HDLC? (0x2F, in HDLC, is SABM.) Or is that a typo in the spec? It's
> not mentioned in the IrLAP V1.1 Errata for March 2003. Were it not for
> that, the IrDA dissector could, I think, use the
> "dissect_xdlc_control()" routine in xdlc.c (the current CVS version,
> that is, which supports named fields for the subfields of the control
> field).
As far as I can see, the IrDA protocol really uses 0x2F as the
XID command. This is easy to verify, as XID is used for basic
discovery. The log below is my Palm-Pilot discovering my Linux box :
-------------------------------------------------------------
17:23:08.866936 xid:cmd ffffffff < d616546e S=6 s=0 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 00 00
17:23:08.955917 xid:cmd ffffffff < d616546e S=6 s=1 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 01 00
17:23:09.034901 xid:cmd ffffffff < d616546e S=6 s=2 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 02 00
17:23:09.185871 xid:cmd ffffffff < d616546e S=6 s=3 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 03 00
17:23:09.186239 xid:rsp 36a47307 > d616546e S=6 s=3 lagaffe hint=0400 [ Computer ] (23)
fe bf 01 07 73 a4 36 6e 54 16 d6 01 03 00 04 00
6c 61 67 61 66 66 65
17:23:09.266852 xid:cmd ffffffff < d616546e S=6 s=4 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 04 00
17:23:09.355833 xid:cmd ffffffff < d616546e S=6 s=5 (14)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 05 00
17:23:09.464810 xid:cmd ffffffff < d616546e S=6 s=* Jean Tourrilhes hint=8220 [ PDA/Palmtop IrOBEX ] (32)
ff 3f 01 6e 54 16 d6 ff ff ff ff 01 ff 00 82 20
00 4a 65 61 6e 20 54 6f 75 72 72 69 6c 68 65 73
-------------------------------------------------------------
With respect to reusing "dissect_xdlc_control()", I can't
comment on that, because I didn't check that code.
Have fun...
Jean
