stamp in the packet; bugs in the OS kernel or the WinPcap driver, or
just forcibly setting the system time backwards, can cause time stamps
in packet traces not to monotonously increase. That can cause infinite
loops when picking the scale for the graph.
svn path=/trunk/; revision=6800
have had one ages ago, but they call their C/C++ compiler "Visual C++".
Use the right name, so as not to confuse people into thinking that the
instructions are only for "Visual C" and can't be used for Visual C++
(yes, this really did appear to happen).
svn path=/trunk/; revision=6799
don't initialize variables that are set elsewhere before they're
used;
don't call "tvb_get_ptr()" to set a variable if you're not going
to use that variable;
make the two character-processing loops have the same structure;
put the result of the XORing into an unsigned character, so it
can be handed to "isprint()" without running the risk of bogus
behavior if the 8th bit is set.
svn path=/trunk/; revision=6796
use : 42:f9:02:34:12:66:22:88 instead of 42:d2:00:34:12:66:22:88
We should accept both (perhaps bytes 2 and 3 are a version number ?)
- the code which looks for the "capture start time" is wrong.
Apparently, we should look for the string "Active Time" in the file.
The "frame_date" structure which contains the capture start time is
found 32 bytes before this string.
svn path=/trunk/; revision=6794
subdissectors, so that we try to dissect the first frame of a
multi-frame message as, for example, SMB, as that's the frame that
contains the SMB header. (That also means we try to dissect middle
frames, but that probably won't work unless one happens to begin with
0xff S M B.)
svn path=/trunk/; revision=6791
The MD5 is copyrighted by L. Peter Deutsch, and released under the same
license as zlib. It is GPL-compatible, and should NOT have the GPL
applied to it.
svn path=/trunk/; revision=6790
until we know that we have the entire PDU - we might not have all of it,
as some of it might be in, for example, a later TCP segment.
svn path=/trunk/; revision=6785
routines except for the enumprinterdata values. Note the display of
strings inside the protocol tree is broken due to lack of a unicode
string frametype.
svn path=/trunk/; revision=6784
output for a USER_LEVEL_1 it looks like the info level and container
pointer are transposed. I'm not even sure this structure is a
container
svn path=/trunk/; revision=6783
we're using is relative to the beginning of that tvbuff, not relative to
the beginning of the containing tvbuff; that also lets us use -1 when in
"proto_tree_add_text()" calls when we mean "to the end of the buffer.
Fix the comment for one field.
svn path=/trunk/; revision=6781
Minor change to the connection oriented DCE/RPC function calls.
Now the offset is provided in the call, instead of having a
hard-coded value in each function. Also makes the calling
convention consistent with the datagram equivalents for the
functions.
Didn't do it for dissect_dcerpc_cn_auth() yet, as that is a
special case (and I am in the process of restructuring it to
make verifier decryption work properly).
svn path=/trunk/; revision=6778
will be dropped), RANAP will be an upper layer of SUA.
- register as upper layer of SCCP with the ITU assigned subsystem number
svn path=/trunk/; revision=6776
changes to the actual argument lists in the calls (putting the byte
string length after the byte string pointer).
Make the byte string length actually be the length, not the length - 1.
Use a #define for the longest byte string it can handle, and put in a
"g_assert()" to check the sanity of that length.
svn path=/trunk/; revision=6774
dissector, and have the LLC dissector register itself as the dissector
for IP-over-FC frames, as the payload is just an LLC 802.2 header plus
payload for the protocol specified by that header.
In the dissector for IP-over-FC as a Wiretap link-layer type, have its
top-level item be a protocol item rather than a text item, and don't
register it as the dissector for IP-over-FC frames from the FC dissector
- it assumes what it's handed includes the Network_Header, but for full
FC frames, the FC dissector has already consumed the Network_Header.
Move the definitions of the value_string tables out of the header file;
most of them are used only in one file, so define them in that file, and
for "fc_fc4_val", define it in "packet-fc.c", make it not static, and
declare it in "packet-fc.h".
Use FALSE, rather than 0, as the last argument to
"proto_tree_add_item()" calls that add a big-endian value.
Fix one "proto_tree_add_uint()" call that was supposed to be a
"proto_tree_add_item()" call.
Use "%u", not "%d", to display unsigned values.
svn path=/trunk/; revision=6773
channel number, X.25 logical channel number).
Clean up white space and the like, and get rid of unnecessary arguments
to "col_set_port()".
svn path=/trunk/; revision=6772
"ether_to_str_punct()" no longer deals only with Ethernet-style
addresses, as it now takes a length argument, rename it
"bytestring_to_str()" - and make it static, as it's not used outside
"to_str.c".
Get rid of unused "fc_to_str_buf()" routine.
svn path=/trunk/; revision=6770
be installed - don't explicitly search
"/usr/lib/ethereal/plugins/{version}" or
"/usr/local/lib/ethereal/plugins/{version}", so that if there's more
than one version of Ethereal installed, we don't end up picking up
plugins from the wrong version.
svn path=/trunk/; revision=6766
Get rid of some probably-unnecessary #includes.
Register the IP-over-FC dissector as the dissector to call for
IP-over-FC captures.
svn path=/trunk/; revision=6764
often get TCP to collapse multiple RPC PDUs into a MSS TCP segment.
This changes the RPC dissector so that it will put one entry on COL_INFO
for each PDU in the segment, (as the SMB dissector does for multiple AndX calls in one SMB PDU)
and just one entry for the first/last RPC PDU.
svn path=/trunk/; revision=6751