Hi,
Since I've often got it wrong myself, I added a check (patch attached) to
make sure that the bitmasks supplied to header fields look sane, i.e. that
they don't accidentally contain any 'holes' (I assumed no correct bitmask
ever would).
Here is the output I see now when ethereal starts up:
lapd.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
lapd.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
iax2.video.subclass: has non-contiguous bitmask (0x000000bf)
lapb.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
lapb.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
lapd.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
lapd.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
llc.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
llc.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
sdlc.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
sdlc.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
telnet.auth.mod.enc: has non-contiguous bitmask (0x00000014)
v120.control.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
v120.control.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
x11.do-not-propagate-mask.erroneous-bits: has non-contiguous bitmask
(0xffffc0b0)
x11.pointer-event-mask.erroneous-bits: has non-contiguous bitmask
(0x00008003)
irlap.c.u_modifier_cmd: has non-contiguous bitmask (0x000000ec)
irlap.c.u_modifier_resp: has non-contiguous bitmask (0x000000ec)
irlmp.dst.lsap: has non-contiguous bitmask (0xffffff7f)
irlmp.src.lsap: has non-contiguous bitmask (0xffffff7f)
ttp.icredit: has non-contiguous bitmask (0xffffff7f)
ttp.dcredit: has non-contiguous bitmask (0xffffff7f)
- Do all or any of the above look like errors?
- Maybe something like this test should be added to Gilbert's script (see
http://www.ethereal.com/lists/ethereal-dev/200505/msg00491.html ?)
Best regards,
Martin
Attachment:
proto.c.diff
Description: Binary data
