On Mon, Jun 25, 2001 at 06:59:41AM +0900, Jun-ichiro itojun Hagino wrote:
>       there seem to be some differences in print-llc.c with netbsd.
>       any idea why?

Most of them are either "we fixed bugs" or "we added enhancements".

For example:

> ***************
> *** 34,48 ****
>   struct llc {
> !     u_int8_t dsap;
> !     u_int8_t ssap;

        ...

> --- 36,50 ----
>   struct llc {
> !     u_char dsap;
> !     u_char ssap;

That was changed in revision 1.7; to quote the change comment for that
revision:

  revision 1.7
  date: 2000/10/03 02:54:57;  author: itojun;  state: Exp;  lines: +10 -10
  always use u_intXX_t for protocol format declaration.  char/short/int may not
  come with exact size.  while at it, correct signedness of ip/udp header field.
  nuke most of the use of bitfield.
  TODO: bitfield in namser.h

Note that our changes may also involve changes to other files, e.g.
handling NetBEUI frames.

>  for example, LLC_IS_POLL has a different vlaue
>       (0x0100 and 0x0001), but the code looks to be equivalent.

        ...

That's revision 1.8 to "print-llc.h":

  revision 1.8
  date: 2000/12/18 07:55:36;  author: guy;  state: Exp;  lines: +8 -8
  Pass only I frames and UI frames to the NetBEUI printer; thanks and a
  tip of the Hatlo hat to Pascal Hennequin <[EMAIL PROTECTED]>
  for catching this one.

  Also, the control field in an S or I frame is little-endian; treat it as
  such, fixing various #defines in "llc.h" to match.

and revision 1.32 to "print-llc.c".  I think those may have been related
to revision 1.30 to "print-llc.c":

  revision 1.30
  date: 2000/12/05 06:42:48;  author: guy;  state: Exp;  lines: +28 -2
  The control field of an LLC frame is 1 byte for a U frame and 2 bytes
  for an S or I frame; have the LLC dissector pass the control field value
  as an argument to the NetBEUI dissector, rather than having the NetBEUI
  dissector extract the control field value itself, and have it pass a
  pointer to the beginning of the LLC *payload* rather than to the control
  field.

  Keep the NetBEUI dissector from running past the end of the buffer.

  Add support for the NetBEUI SESSION ALIVE frame.

which is a bug I found when looking at a bunch of captures with NetBEUI
frames in them.
-
This is the TCPDUMP workers list. It is archived at
http://www.tcpdump.org/lists/workers/index.html
To unsubscribe use mailto:[EMAIL PROTECTED]?body=unsubscribe

Reply via email to