path: root/doc/wsdg_src/wsdg_dissection.adoc

// WSDG Chapter Dissection

[#ChapterDissection]

== Packet Dissection

[#ChDissectWorks]

=== How packet dissection works

Each dissector decodes its part of the protocol and then hands off
decoding to subsequent dissectors for an encapsulated protocol.

Every dissection starts with the Frame dissector which dissects the
details of the capture file itself (e.g. timestamps). From there it passes the
data on to the lowest-level data dissector, e.g. the Ethernet dissector for
the Ethernet header. The payload is then passed on to the next dissector (e.g.
IP) and so on. At each stage, details of the packet are decoded and
displayed.

Dissectors can either be built-in to Wireshark or written as a self-registering
plugin (a shared library or DLL).
There is little difference in having your dissector as either a plugin
or built-in.  You have limited function access through the ABI exposed
by functions declared as WS_DLL_PUBLIC.

The big benefit of writing a dissector as a plugin is that rebuilding
a plugin is much faster than rebuilding wireshark after editing a built-in
dissector.
As such, starting with a plugin often makes initial development quicker, while
the finished code may make more sense as a built-in dissector.

[NOTE]
.Read README.dissector
====
The file _doc/README.dissector_ contains detailed information about writing
a dissector. In many cases it is more up to date than this document.
====

[#ChDissectAdd]

=== Adding a basic dissector

Let’s step through adding a basic dissector. We'll start with the made up "foo"
protocol. It consists of the following basic items.

* A packet type - 8 bits. Possible values: 1 - initialisation, 2 - terminate, 3 - data.

* A set of flags stored in 8 bits. 0x01 - start packet, 0x02 - end packet, 0x04 - priority packet.

* A sequence number - 16 bits.

* An IPv4 address.

[#ChDissectSetup]

==== Setting up the dissector

The first decision you need to make is if this dissector will be a
built-in dissector and included in the main program, or a plugin.

Plugins are easier to write initially, so let’s start with that.
With a little care, the plugin can be converted into a built-in dissector.

.Dissector Initialisation.
[source,c]
----
#include "config.h"
#include <epan/packet.h>

#define FOO_PORT 1234

static int proto_foo;

void
proto_register_foo(void)
{
    proto_foo = proto_register_protocol (
        "FOO Protocol", /* name        */
        "FOO",          /* short name  */
        "foo"           /* filter_name */
        );
}
----

Let’s go through this a bit at a time. First we have some boilerplate
include files. These will be pretty constant to start with.

Then a `#define` for the UDP port that carries _foo_ traffic.

Next we have `proto_foo`, an int that stores our protocol handle and is
initialised to `-1`.
This handle will be set when the dissector is registered within the main program.
It’s good practice to make all variables and functions that aren't exported
static to minimize name space pollution. This normally isn't a problem unless your
dissector gets so big that it spans multiple files.

Now that we have the basics in place to interact with the main program, we'll
start with two protocol dissector setup functions: `proto_register_XXX` and
`proto_reg_handoff_XXX`.

Each protocol must have a register function with the form "proto_register_XXX".
This function is used to register the protocol in Wireshark.
The code to call the register routines is generated automatically and is
called when Wireshark starts. In this example, the function is named
`proto_register_foo`.

`proto_register_foo` calls `proto_register_protocol()`, which takes a `name`,
`short name`, and `filter_name`. The
name and short name are used in the "Preferences" and "Enabled protocols"
dialogs and the documentation's generated field name list. The
`filter_name` is used as the display filter name. `proto_register_protocol()`
returns a protocol handle, which can be used to refer to the protocol and
obtain a handle to the protocol's dissector.


Next we need a handoff routine.

.Dissector Handoff.
[source,c]
----
void
proto_reg_handoff_foo(void)
{
    static dissector_handle_t foo_handle;

    foo_handle = create_dissector_handle(dissect_foo, proto_foo);
    dissector_add_uint("udp.port", FOO_PORT, foo_handle);
}
----

A handoff routine associates a protocol handler with the protocol's
traffic. It consists of two major steps: The first step is to create a
dissector handle, which is a handle associated with the protocol and the
function called to do the actual dissecting.
The second step is to register the dissector handle so that traffic
associated with the protocol calls the dissector.

In this example, `proto_reg_handoff_foo()` calls `create_dissector_handle()`
to obtain a dissector handle for the foo protocol. It then uses
`dissector_add_uint()` to associate traffic on UDP port FOO_PORT (1234)
with the foo protocol, so that Wireshark will call `dissect_foo()` when
it receives UDP traffic on port 1234.

Wireshark's dissector convention is to put `proto_register_foo()` and
`proto_reg_handoff_foo()` as the last two functions in the dissector source.

The next step is to write the dissecting function, `dissect_foo()`.
We'll start with a basic placeholder.

.Dissection.
[source,c]
----
static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree _U_, void *data _U_)
{
    col_set_str(pinfo->cinfo, COL_PROTOCOL, "FOO");
    /* Clear the info column */
    col_clear(pinfo->cinfo,COL_INFO);

    return tvb_captured_length(tvb);
}
----

`dissect_foo()` is called to dissect the packets presented to it. The packet data
is held in a special buffer referenced here as `tvb`. The packet_info structure
contains general data about the protocol and we can update
information here. The tree parameter is where the detail dissection takes place.
Note that the `\_U_` following `tree` and `data` signals to the compiler that the
parameters are unused, so that the compiler does not print a warning.

For now we'll do the minimum we can get away with. `col_set_str()` is used to set
Wireshark's Protocol column to "FOO" so everyone can see it’s being
recognised. The
only other thing we do is to clear out any data in the INFO column if it’s being
displayed.

At this point we have a basic dissector ready to compile and install.
The dissector doesn't do anything other than identify the protocol and label it.
Here is the dissector's complete code:

.Complete _packet-foo.c_:.
[source,c]
----
#include "config.h"
#include <epan/packet.h>

#define FOO_PORT 1234

static int proto_foo;

static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree _U_, void *data _U_)
{
    col_set_str(pinfo->cinfo, COL_PROTOCOL, "FOO");
    /* Clear the info column */
    col_clear(pinfo->cinfo,COL_INFO);

    return tvb_captured_length(tvb);
}

void
proto_register_foo(void)
{
    proto_foo = proto_register_protocol (
        "FOO Protocol", /* name        */
        "FOO",          /* short_name  */
        "foo"           /* filter_name */
        );
}

void
proto_reg_handoff_foo(void)
{
    static dissector_handle_t foo_handle;

    foo_handle = create_dissector_handle(dissect_foo, proto_foo);
    dissector_add_uint("udp.port", FOO_PORT, foo_handle);
}

----

To compile this dissector and create a plugin a few support files
are required, besides the dissector source in _packet-foo.c_:

* _CMakeLists.txt_ - Contains the CMake file and version info for this plugin.

* _packet-foo.c_ - Your dissector source.

* _plugin.rc.in_ - Contains the DLL resource template for Windows. (optional)

Samples of these files are available in the gryphon plugin directory
(plugins/epan/gryphon).
If you copy the files from the gryphon plugin, _CMakeLists.txt_ will need
to be updated with the correct plugin name, version
info, and the relevant files to compile.

In the main top-level source directory, copy _CMakeListsCustom.txt.example_ to
_CMakeListsCustom.txt_ and add the path of your plugin to the list in
`CUSTOM_PLUGIN_SRC_DIR`.

Compile the dissector to a DLL or shared library and either run Wireshark from
the build directory as detailed in <<ChSrcRunFirstTime>> or copy the plugin
binary into the plugin directory of your Wireshark installation and run that.

[#ChDissectDetails]

==== Dissecting the protocol's details

Now that we have our basic dissector up and running, let’s do something with it.
The simplest thing to start with is labeling the payload. We can label the
payload by building a subtree to decode our results into.
This subtree will hold all the protocol's details and
helps keep things looking nice in the detailed display.

We add the new subtree with `proto_tree_add_item()`, as is depicted below:

.Plugin Packet Dissection.
[source,c]
----
static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
    col_set_str(pinfo->cinfo, COL_PROTOCOL, "FOO");
    /* Clear out stuff in the info column */
    col_clear(pinfo->cinfo,COL_INFO);

    proto_item *ti = proto_tree_add_item(tree, proto_foo, tvb, 0, -1, ENC_NA);

    return tvb_captured_length(tvb);
}
----

As the `FOO` protocol does not encapsulate another protocol, we
consume all of the tvb's data, from `0` to the end (`-1`).

The final parameter specifies the "encoding" and is set to
`ENC_NA` ("not applicable"), as the protocol doesn't specifically
use big endian (`ENC_BIG_ENDIAN`) or little endian (`ENC_LITTLE_ENDIAN`).

After adding the call to
`proto_tree_add_item()`
, there should be a label `FOO` in the protocol's detailed display.
Selecting this label will highlight the remaining contents of the packet.

Now let’s go to the next step and add some protocol dissection. To do this
we'll need to construct tables to define which fields will be present in the
packet and to store the opened/closed state of the subtree. We'll
add these statically allocated arrays to the beginning of the file
and name them
`hf_register_info` ('hf' is short for 'header field') and `ett`.
The arrays wil then registered after the call to
`proto_register_protocol()` by calling `proto_register_field_array()`
and `proto_register_subtree_array()`:

.Registering data structures.
[source,c]
----
static int hf_foo_pdu_type;
static int ett_foo;

/* ... */

void
proto_register_foo(void)
{
    static hf_register_info hf[] = {
        { &hf_foo_pdu_type,
            { "FOO PDU Type", "foo.type",
            FT_UINT8, BASE_DEC,
            NULL, 0x0,
            NULL, HFILL }
        }
    };

    /* Setup protocol subtree array */
    static int *ett[] = {
        &ett_foo
    };

    proto_foo = proto_register_protocol (
        "FOO Protocol", /* name       */
        "FOO",          /* short_name */
        "foo"           /* filter_name*/
        );

    proto_register_field_array(proto_foo, hf, array_length(hf));
    proto_register_subtree_array(ett, array_length(ett));
}
----

As you can see, a field `foo.type` was defined inside the array of
header fields.

Now we can dissect the `FOO PDU Type` (referenced as `foo.type`)
field in `dissect_foo()` by adding
the FOO Protocol's subtree with `proto_item_add_subtree()` and
then calling `proto_tree_add_item()` to add the field:

.Dissector starting to dissect the packets.
[source,c]
----
    proto_item *ti = proto_tree_add_item(tree, proto_foo, tvb, 0, -1, ENC_NA);
    proto_tree *foo_tree = proto_item_add_subtree(ti, ett_foo);
    proto_tree_add_item(foo_tree, hf_foo_pdu_type, tvb, 0, 1, ENC_BIG_ENDIAN);
----

As mentioned earlier, the foo protocol begins with an 8-bit `packet type`
which can have three possible values: 1 - initialisation, 2 - terminate, 3 - data.
Here's how we can add the packet details:

The `proto_item_add_subtree()` call has added a child node
to the protocol tree which is where we will do our detail dissection.
The expansion of this node is controlled by the `ett_foo`
variable. This remembers if the node should be expanded or not as you move
between packets. All subsequent dissection will be added to this tree,
as you can see from the next call.
A call to `proto_tree_add_item()` in the foo_tree,
this time using the `hf_foo_pdu_type` to control the formatting
of the item. The pdu type is one byte of data, starting at 0. We assume it is
in network order (also called big endian), so that is why we use `ENC_BIG_ENDIAN`.
For a 1-byte quantity, there is no order issue, but it is good practice to
make this the same as any multibyte fields that may be present, and as we will
see in the next section, this particular protocol uses network order.

If we look in detail at the `hf_foo_pdu_type` declaration in
the static array we can see the details of the definition.

----
static hf_register_info hf[] = {
    { &hf_foo_pdu_type,
        { "FOO PDU Type", "foo.type",
        FT_UINT8, BASE_DEC,
        NULL, 0x0,
        NULL, HFILL }
    }
};
----

* _hf_foo_pdu_type_ - The node's index.

* _FOO PDU Type_ - The item's label.

* _foo.type_ - The item's abbreviated name, for use in the display filter
(e.g., `foo.type=1`).

* _FT_UINT8_ - The item's type: An 8bit unsigned integer.
This tallies with our call above where we tell it to only look at one byte.

* _BASE_DEC_ - For an integer type, this tells it to be printed as a decimal
number. It could be hexadecimal (BASE_HEX) or octal (BASE_OCT) if that made more sense.

We'll ignore the rest of the structure for now.

If you install this plugin and try it out, you'll see something that begins to look
useful.

Now let’s finish off dissecting the simple protocol. We need to add a few
more variables to the hfarray, and a couple more procedure calls.

.Wrapping up the packet dissection.
[source,c]
----
...
static int hf_foo_flags;
static int hf_foo_sequenceno;
static int hf_foo_initialip;
...

static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
    int offset = 0;

    ...
    proto_item *ti = proto_tree_add_item(tree, proto_foo, tvb, 0, -1, ENC_NA);
    proto_tree *foo_tree = proto_item_add_subtree(ti, ett_foo);
    proto_tree_add_item(foo_tree, hf_foo_pdu_type, tvb, offset, 1, ENC_BIG_ENDIAN);
    offset += 1;
    proto_tree_add_item(foo_tree, hf_foo_flags, tvb, offset, 1, ENC_BIG_ENDIAN);
    offset += 1;
    proto_tree_add_item(foo_tree, hf_foo_sequenceno, tvb, offset, 2, ENC_BIG_ENDIAN);
    offset += 2;
    proto_tree_add_item(foo_tree, hf_foo_initialip, tvb, offset, 4, ENC_BIG_ENDIAN);
    offset += 4;
    ...

    return tvb_captured_length(tvb);
}

void
proto_register_foo(void) {
    ...
        ...
        { &hf_foo_flags,
            { "FOO PDU Flags", "foo.flags",
            FT_UINT8, BASE_HEX,
            NULL, 0x0,
            NULL, HFILL }
        },
        { &hf_foo_sequenceno,
            { "FOO PDU Sequence Number", "foo.seqn",
            FT_UINT16, BASE_DEC,
            NULL, 0x0,
            NULL, HFILL }
        },
        { &hf_foo_initialip,
            { "FOO PDU Initial IP", "foo.initialip",
            FT_IPv4, BASE_NONE,
            NULL, 0x0,
            NULL, HFILL }
        },
        ...
    ...
}
...
----

This dissects all the bits of this simple hypothetical protocol. We've
introduced a new variable offsetinto the mix to help keep track of where we are
in the packet dissection. With these extra bits in place, the whole protocol is
now dissected.

==== Improving the dissection information

We can certainly improve the display of the protocol with a bit of extra data.
The first step is to add some text labels. Let’s start by labeling the packet
types. There is some useful support for this sort of thing by adding a couple of
extra things. First we add a simple table of type to name.


.Naming the packet types.
[source,c]
----
static const value_string packettypenames[] = {
    { 1, "Initialise" },
    { 2, "Terminate" },
    { 3, "Data" },
    { 0, NULL }
};
----

This is a handy data structure that can be used to look up a name for a value.
There are routines to directly access this lookup table, but we don't need to
do that, as the support code already has that added in. We just have to give
these details to the appropriate part of the data, using the `VALS` macro.

.Adding Names to the protocol.
[source,c]
----
   { &hf_foo_pdu_type,
        { "FOO PDU Type", "foo.type",
        FT_UINT8, BASE_DEC,
        VALS(packettypenames), 0x0,
        NULL, HFILL }
    }
----

This helps in deciphering the packets, and we can do a similar thing for the
flags structure. For this we need to add some more data to the table though.

.Adding Flags to the protocol.
[source,c]
----
#define FOO_START_FLAG      0x01
#define FOO_END_FLAG        0x02
#define FOO_PRIORITY_FLAG   0x04

static int hf_foo_startflag;
static int hf_foo_endflag;
static int hf_foo_priorityflag;

static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
    ...
        ...
        static int* const bits[] = {
            &hf_foo_startflag,
            &hf_foo_endflag,
            &hf_foo_priorityflag,
            NULL
        };

        proto_tree_add_bitmask(foo_tree, tvb, offset, hf_foo_flags, ett_foo, bits, ENC_BIG_ENDIAN);
        offset += 1;
        ...
    ...
    return tvb_captured_length(tvb);
}

void
proto_register_foo(void) {
    ...
        ...
        { &hf_foo_startflag,
            { "FOO PDU Start Flags", "foo.flags.start",
            FT_BOOLEAN, 8,
            NULL, FOO_START_FLAG,
            NULL, HFILL }
        },
        { &hf_foo_endflag,
            { "FOO PDU End Flags", "foo.flags.end",
            FT_BOOLEAN, 8,
            NULL, FOO_END_FLAG,
            NULL, HFILL }
        },
        { &hf_foo_priorityflag,
            { "FOO PDU Priority Flags", "foo.flags.priority",
            FT_BOOLEAN, 8,
            NULL, FOO_PRIORITY_FLAG,
            NULL, HFILL }
        },
        ...
    ...
}
...
----

Some things to note here. For the flags, as each bit is a different flag, we use
the type `FT_BOOLEAN`, as the flag is either on or off. Second, we include the flag
mask in the 7th field of the data, which allows the system to mask the relevant bit.
We've also changed the 5th field to 8, to indicate that we are looking at an 8 bit
quantity when the flags are extracted. Then finally we add the extra constructs
to the dissection routine.

This is starting to look fairly full featured now, but there are a couple of
other things we can do to make things look even more pretty. At the moment our
dissection shows the packets as "Foo Protocol" which whilst correct is a little
uninformative. We can enhance this by adding a little more detail. First, let’s
get hold of the actual value of the protocol type. We can use the handy function
`tvb_get_uint8()` to do this. With this value in hand, there are a couple of
things we can do. First we can set the INFO column of the non-detailed view to
show what sort of PDU it is - which is extremely helpful when looking at
protocol traces. Second, we can also display this information in the dissection
window.

.Enhancing the display.
[source,c]
----
static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
    int offset = 0;
    uint8_t packet_type = tvb_get_uint8(tvb, 0);

    col_set_str(pinfo->cinfo, COL_PROTOCOL, "FOO");
    /* Clear out stuff in the info column */
    col_clear(pinfo->cinfo,COL_INFO);
    col_add_fstr(pinfo->cinfo, COL_INFO, "Type %s",
             val_to_str(packet_type, packettypenames, "Unknown (0x%02x)"));

    proto_item *ti = proto_tree_add_item(tree, proto_foo, tvb, 0, -1, ENC_NA);
    proto_item_append_text(ti, ", Type %s",
        val_to_str(packet_type, packettypenames, "Unknown (0x%02x)"));
    proto_tree *foo_tree = proto_item_add_subtree(ti, ett_foo);
    proto_tree_add_item(foo_tree, hf_foo_pdu_type, tvb, offset, 1, ENC_BIG_ENDIAN);
    offset += 1;

    return tvb_captured_length(tvb);
}
----

So here, after grabbing the value of the first 8 bits, we use it with one of the
built-in utility routines `val_to_str()`, to lookup the value. If the value
isn't found we provide a fallback which just prints the value in hex. We use
this twice, once in the INFO field of the columns -- if it’s displayed, and
similarly we append this data to the base of our dissecting tree.

[#ChDissectExpertInfo]

=== How to add an expert item

A dissector showing the protocol fields and interpretation of their values is
very informative. It can be even more helpful if the dissector can draw your
attention to fields where something noteworthy can be seen. This can be something
as simple as the start flag of a session, or something more severe as an invalid
value.

Here we take our dissector for `FOO` and add an expert item for the sequence
number being zero (assuming that's a noteworthy thing for this protocol).

.Expert item setup.
[source,c]
----
#include <epan/expert.h>

static expert_field ei_foo_seqn_zero;

/* ... */

void
proto_register_foo(void)
{
    /* ... */
    expert_module_t* expert_foo;

    /* ... */
    static ei_register_info ei[] = {
        {
            &ei_foo_seqn_zero,
            { "foo.seqn_zero", PI_SEQUENCE, PI_CHAT,
              "Sequence number is zero", EXPFILL }
        }
    };

    /* ... */
    expert_foo = expert_register_protocol(proto_foo);
    expert_register_field_array(expert_foo, ei, array_length(ei));
}
----

Let's go through this step by step. The data structures and functions needed for
expert items are found in epan/expert.h, so we have to include that file.

Next we have to allocate an `expert_field` structure for every type of expert item
we would like to add to the dissection. This structure is initialised with `EI_INIT`.

Now we have to register with the protocol we are providing expert info for. Since
we already have a function to register our protocol, we add the expert info
registration there too. This is done by calling `expert_register_protocol()` with
the handle for the protocol we received earlier in this function.

Next we need to register an array of definitions of expert items that we would
like to add to the dissection. This array, not unlike the array of header fields
before, contains all the data the dissection engine needs to create and handle
the expert items.

The expert item definition consists of a pointer to the `expert_field` structure
we defined before and a structure with data elements of the expert item itself.

* _"foo.seqn_zero"_ - The expert items display filter

* _PI_SEQUENCE_ - The group to which the expert item belongs

* _PI_CHAT_ - The severity of the expert item

* _"Sequence number is zero"_ - The text string added to the dissection

We'll ignore the rest of the structure for now.

To keep an overview of lots of expert items it helps to categorize them into groups.
Currently there are several types of groups defined, e.g. `checksum`, `sequence`,
`protocol`, etc. All these are defined in the epan/proto.h header file.

Not every noteworthy field value is of equal severity. The start of a session
is nice to know, while an invalid value may be significant error in the protocol.
To differentiate between these severties the expert item is assigned one of them:
`comment`, `chat`, `note`, `warn` or `error`. Try to choose the lowest one which
is suitable. The topic you're currently working on seems probably more important
than it will look like in a few weeks.

With the expert item array setup, we add this to the dissection engine with a
call to `expert_register_field_array()`.

Now that all information of the expert item is defined and registered it's time
to actually add the expert item to the dissection.

.Expert item use.
[source,c]
----
static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
    uint32_t sequenceno = 0xFFFF;

    /* ... */

    ti = proto_tree_add_item_ret_uint(foo_tree, hf_foo_sequenceno,
                                      tvb, offset, 2, ENC_BIG_ENDIAN, &sequenceno);
    if (sequenceno == 0) {
        expert_add_info(pinfo, ti, &ei_foo_seqn_zero);
    }

    /* ... */
}
----

There's been a slight alteration to the function used to add the sequence number
dissection. First the proto_item created by the function is saved in previously
defined variable `ti`, and the actual value of the field is stored in the variable
`sequenceno`. We can now use the value of this field to determine wether to add
the expert item.

Adding the expert item is simply done by calling `expert_add_info()` with reference
to the `packet_info` structure, the proto item `ti` to add the expert item to and
the previously defined and registered expert item information.

[#ChDissectTransformed]

=== How to handle transformed data

Some protocols do clever things with data. They might possibly
encrypt the data, or compress data, or part of it. If you know
how these steps are taken it is possible to reverse them within the
dissector.

As encryption can be tricky, let’s consider the case of compression.
These techniques can also work for other transformations of data,
where some step is required before the data can be examined.

What basically needs to happen here, is to identify the data that needs
conversion, take that data and transform it into a new stream, and then call a
dissector on it. Often this needs to be done "on-the-fly" based on clues in the
packet. Sometimes this needs to be used in conjunction with other techniques,
such as packet reassembly. The following shows a technique to achieve this
effect.

.Decompressing data packets for dissection.
[source,c]
----
    uint8_t flags = tvb_get_uint8(tvb, offset);
    offset ++;
    if (flags & FLAG_COMPRESSED) { /* the remainder of the packet is compressed */
        uint16_t orig_size = tvb_get_ntohs(tvb, offset);
        unsigned char *decompressed_buffer = (unsigned char*)wmem_alloc(pinfo->pool, orig_size);
        offset += 2;
        decompress_packet(tvb_get_ptr(tvb, offset, -1),
                tvb_captured_length_remaining(tvb, offset),
                decompressed_buffer, orig_size);
        /* Now re-setup the tvb buffer to have the new data */
        next_tvb = tvb_new_child_real_data(tvb, decompressed_buffer, orig_size, orig_size);
        add_new_data_source(pinfo, next_tvb, "Decompressed Data");
    } else {
        next_tvb = tvb_new_subset_remaining(tvb, offset);
    }
    offset = 0;
    /* process next_tvb from here on */
----

The first steps here are to recognise the compression. In this case a flag byte
alerts us to the fact the remainder of the packet is compressed. Next we
retrieve the original size of the packet, which in this case is conveniently
within the protocol. If it’s not, it may be part of the compression routine to
work it out for you, in which case the logic would be different.

So armed with the size, a buffer is allocated to receive the uncompressed data
using `wmem_alloc()` in pinfo->pool memory, and the packet is decompressed into
it. The `tvb_get_ptr()` function is useful to get a pointer to the raw data of
the packet from the offset onwards. In this case the decompression routine also
needs to know the length, which is given by the
`tvb_captured_length_remaining()` function.

Next we build a new tvb buffer from this data, using the
`tvb_new_child_real_data()` call. This data is a child of our original data, so
calling this function also acknowledges that. No need to call
`tvb_set_free_cb()` as the pinfo->pool was used (the memory block will be
automatically freed when the pinfo pool lifetime expires). Finally we add this
tvb as a new data source, so that the detailed display can show the
decompressed bytes as well as the original.

After this has been set up the remainder of the dissector can dissect the buffer
next_tvb, as it’s a new buffer the offset needs to be 0 as we start again from
the beginning of this buffer. To make the rest of the dissector work regardless
of whether compression was involved or not, in the case that compression was not
signaled, we use `tvb_new_subset_remaining()` to deliver us a new buffer based
on the old one but starting at the current offset, and extending to the end.
This makes dissecting the packet from this point on exactly the same regardless
of compression.

[#ChDissectReassemble]

=== How to reassemble split packets

Some protocols have times when they have to split a large packet across
multiple other packets. In this case the dissection can't be carried out correctly
until you have all the data. The first packet doesn't have enough data,
and the subsequent packets don't have the expect format.
To dissect these packets you need to wait until all the parts have
arrived and then start the dissection.

The following sections will guide you through two common cases. For a
description of all possible functions, structures and parameters, see
_epan/reassemble.h_.

[#ChDissectReassembleUdp]

==== How to reassemble split UDP packets

As an example, let’s examine a protocol that is layered on top of UDP that
splits up its own data stream. If a packet is bigger than some given size, it
will be split into chunks, and somehow signaled within its protocol.

To deal with such streams, we need several things to trigger from. We need to
know that this packet is part of a multi-packet sequence. We need to know how
many packets are in the sequence. We also need to know when we have all the
packets.

For this example we'll assume there is a simple in-protocol signaling mechanism
to give details. A flag byte that signals the presence of a multi-packet
sequence and also the last packet, followed by an ID of the sequence and a
packet sequence number.

----
msg_pkt ::= SEQUENCE {
    .....
    flags ::= SEQUENCE {
        fragment    BOOLEAN,
        last_fragment   BOOLEAN,
    .....
    }
    msg_id  INTEGER(0..65535),
    frag_id INTEGER(0..65535),
    .....
}
----

.Reassembling fragments - Part 1
[source,c]
----
#include <epan/reassemble.h>
   ...
save_fragmented = pinfo->fragmented;
flags = tvb_get_uint8(tvb, offset); offset++;
if (flags & FL_FRAGMENT) { /* fragmented */
    tvbuff_t* new_tvb = NULL;
    fragment_data *frag_msg = NULL;
    uint16_t msg_seqid = tvb_get_ntohs(tvb, offset); offset += 2;
    uint16_t msg_num = tvb_get_ntohs(tvb, offset); offset += 2;

    pinfo->fragmented = true;
    frag_msg = fragment_add_seq_check(msg_reassembly_table,
        tvb, offset, pinfo,
        msg_seqid, NULL, /* ID for fragments belonging together */
        msg_num, /* fragment sequence number */
        tvb_captured_length_remaining(tvb, offset), /* fragment length - to the end */
        flags & FL_FRAG_LAST); /* More fragments? */
----

We start by saving the fragmented state of this packet, so we can restore it
later. Next comes some protocol specific stuff, to dig the fragment data out of
the stream if it’s present. Having decided it is present, we let the function
`fragment_add_seq_check()` do its work. We need to provide this with a certain
amount of parameters:

* The `msg_reassembly_table` table is for bookkeeping and is described later.

* The tvb buffer we are dissecting.

* The offset where the partial packet starts.

* The provided packet info.

* The sequence number of the fragment stream. There may be several streams of
  fragments in flight, and this is used to key the relevant one to be used for
  reassembly.

* Optional additional data for identifying the fragment. Can be set to `NULL`
  (as is done in the example) for most dissectors.

* msg_num is the packet number within the sequence.

* The length here is specified as the rest of the tvb as we want the rest of the packet data.

* Finally a parameter that signals if this is the last fragment or not. This
  might be a flag as in this case, or there may be a counter in the protocol.

.Reassembling fragments part 2
[source,c]
----
    new_tvb = process_reassembled_data(tvb, offset, pinfo,
        "Reassembled Message", frag_msg, &msg_frag_items,
        NULL, msg_tree);

    if (frag_msg) { /* Reassembled */
        col_append_str(pinfo->cinfo, COL_INFO,
                " (Message Reassembled)");
    } else { /* Not last packet of reassembled Short Message */
        col_append_fstr(pinfo->cinfo, COL_INFO,
                " (Message fragment %u)", msg_num);
    }

    if (new_tvb) { /* take it all */
        next_tvb = new_tvb;
    } else { /* make a new subset */
        next_tvb = tvb_new_subset_remaining(tvb, offset);
    }
}
else { /* Not fragmented */
    next_tvb = tvb_new_subset_remaining(tvb, offset);
}

.....
pinfo->fragmented = save_fragmented;
----

Having passed the fragment data to the reassembly handler, we can now check if
we have the whole message. If there is enough information, this routine will
return the newly reassembled data buffer.

After that, we add a couple of informative messages to the display to show that
this is part of a sequence. Then a bit of manipulation of the buffers and the
dissection can proceed. Normally you will probably not bother dissecting further
unless the fragments have been reassembled as there won't be much to find.
Sometimes the first packet in the sequence can be partially decoded though if
you wish.

Now the mysterious data we passed into the `fragment_add_seq_check()`.

.Reassembling fragments - Initialisation
[source,c]
----
static reassembly_table reassembly_table;

static void
proto_register_msg(void)
{
    reassembly_table_register(&msg_reassemble_table,
        &addresses_ports_reassembly_table_functions);
}
----

First a `reassembly_table` structure is declared and initialised in the protocol
initialisation routine. The second parameter specifies the functions that should
be used for identifying fragments. We will use
`addresses_ports_reassembly_table_functions` in order to identify fragments by
the given sequence number (`msg_seqid`), the source and destination addresses
and ports from the packet.

Following that, a `fragment_items` structure is allocated and filled in with a
series of ett items, hf data items, and a string tag. The ett and hf values
should be included in the relevant tables like all the other variables your
protocol may use. The hf variables need to be placed in the structure something
like the following. Of course the names may need to be adjusted.

.Reassembling fragments - Data
[source,c]
----
...
static int hf_msg_fragments;
static int hf_msg_fragment;
static int hf_msg_fragment_overlap;
static int hf_msg_fragment_overlap_conflicts;
static int hf_msg_fragment_multiple_tails;
static int hf_msg_fragment_too_long_fragment;
static int hf_msg_fragment_error;
static int hf_msg_fragment_count;
static int hf_msg_reassembled_in;
static int hf_msg_reassembled_length;
...
static int ett_msg_fragment;
static int ett_msg_fragments;
...
static const fragment_items msg_frag_items = {
    /* Fragment subtrees */
    &ett_msg_fragment,
    &ett_msg_fragments,
    /* Fragment fields */
    &hf_msg_fragments,
    &hf_msg_fragment,
    &hf_msg_fragment_overlap,
    &hf_msg_fragment_overlap_conflicts,
    &hf_msg_fragment_multiple_tails,
    &hf_msg_fragment_too_long_fragment,
    &hf_msg_fragment_error,
    &hf_msg_fragment_count,
    /* Reassembled in field */
    &hf_msg_reassembled_in,
    /* Reassembled length field */
    &hf_msg_reassembled_length,
    /* Tag */
    "Message fragments"
};
...
static hf_register_info hf[] =
{
...
{&hf_msg_fragments,
    {"Message fragments", "msg.fragments",
    FT_NONE, BASE_NONE, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment,
    {"Message fragment", "msg.fragment",
    FT_FRAMENUM, BASE_NONE, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_overlap,
    {"Message fragment overlap", "msg.fragment.overlap",
    FT_BOOLEAN, 0, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_overlap_conflicts,
    {"Message fragment overlapping with conflicting data",
    "msg.fragment.overlap.conflicts",
    FT_BOOLEAN, 0, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_multiple_tails,
    {"Message has multiple tail fragments",
    "msg.fragment.multiple_tails",
    FT_BOOLEAN, 0, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_too_long_fragment,
    {"Message fragment too long", "msg.fragment.too_long_fragment",
    FT_BOOLEAN, 0, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_error,
    {"Message defragmentation error", "msg.fragment.error",
    FT_FRAMENUM, BASE_NONE, NULL, 0x00, NULL, HFILL } },
{&hf_msg_fragment_count,
    {"Message fragment count", "msg.fragment.count",
    FT_UINT32, BASE_DEC, NULL, 0x00, NULL, HFILL } },
{&hf_msg_reassembled_in,
    {"Reassembled in", "msg.reassembled.in",
    FT_FRAMENUM, BASE_NONE, NULL, 0x00, NULL, HFILL } },
{&hf_msg_reassembled_length,
    {"Reassembled length", "msg.reassembled.length",
    FT_UINT32, BASE_DEC, NULL, 0x00, NULL, HFILL } },
...
static int *ett[] =
{
...
&ett_msg_fragment,
&ett_msg_fragments
...
----

These hf variables are used internally within the reassembly routines to make
useful links, and to add data to the dissection. It produces links from one
packet to another, such as a partial packet having a link to the fully
reassembled packet. Likewise there are back pointers to the individual packets
from the reassembled one. The other variables are used for flagging up errors.

[#TcpDissectPdus]

==== How to reassemble split TCP Packets

A dissector gets a `tvbuff_t` pointer which holds the payload
of a TCP packet. This payload contains the header and data
of your application layer protocol.

When dissecting an application layer protocol you cannot assume
that each TCP packet contains exactly one application layer message.
One application layer message can be split into several TCP packets.

You also cannot assume that a TCP packet contains only one application layer message
and that the message header is at the start of your TCP payload.
More than one messages can be transmitted in one TCP packet,
so that a message can start at an arbitrary position.

This sounds complicated, but there is a simple solution.
`tcp_dissect_pdus()` does all this tcp packet reassembling for you.
This function is implemented in _epan/dissectors/packet-tcp.h_.

.Reassembling TCP fragments
[source,c]
----
#include "config.h"

#include <epan/packet.h>
#include <epan/prefs.h>
#include "packet-tcp.h"

...

#define FRAME_HEADER_LEN 8

/* This method dissects fully reassembled messages */
static int
dissect_foo_message(tvbuff_t *tvb, packet_info *pinfo _U_, proto_tree *tree _U_, void *data _U_)
{
    /* TODO: implement your dissecting code */
    return tvb_captured_length(tvb);
}

/* determine PDU length of protocol foo */
static unsigned
get_foo_message_len(packet_info *pinfo _U_, tvbuff_t *tvb, int offset, void *data _U_)
{
    /* TODO: change this to your needs */
    return (unsigned)tvb_get_ntohl(tvb, offset+4); /* e.g. length is at offset 4 */
}

/* The main dissecting routine */
static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data)
{
    tcp_dissect_pdus(tvb, pinfo, tree, true, FRAME_HEADER_LEN,
                     get_foo_message_len, dissect_foo_message, data);
    return tvb_captured_length(tvb);
}

...
----

As you can see this is really simple. Just call `tcp_dissect_pdus()` in your
main dissection routine and move you message parsing code into another function.
This function gets called whenever a message has been reassembled.

The parameters tvb, pinfo, tree and data are just handed over to
`tcp_dissect_pdus()`. The 4th parameter is a flag to indicate if the data should
be reassembled or not. This could be set according to a dissector preference as
well. Parameter 5 indicates how much data has at least to be available to be
able to determine the length of the foo message. Parameter 6 is a function
pointer to a method that returns this length. It gets called when at least the
number of bytes given in the previous parameter is available. Parameter 7 is a
function pointer to your real message dissector. Parameter 8 is the data
passed in from parent dissector.

Protocols which need more data before the message length can be determined can
return zero. Other values smaller than the fixed length will result in an
exception.

[#ChDissectTap]

=== How to tap protocols

Adding a Tap interface to a protocol allows it to do some useful things.
In particular you can produce protocol statistics from the tap interface.

A tap is basically a way of allowing other items to see what’s happening as
a protocol is dissected. A tap is registered with the main program, and
then called on each dissection. Some arbitrary protocol specific data
is provided with the routine that can be used.

To create a tap, you first need to register a tap. A tap is registered with an
integer handle, and registered with the routine `register_tap()`. This takes a
string name with which to find it again.

.Initialising a tap
[source,c]
----
#include <epan/packet.h>
#include <epan/tap.h>

static int foo_tap;

void proto_register_foo(void)
{
       ...
    foo_tap = register_tap("foo");
----

Whilst you can program a tap without protocol specific data, it is generally not
very useful. Therefore it’s a good idea to declare a structure that can be
passed through the tap. This needs to be allocated in packet scope as it will be used
after the dissection routine has returned. It’s generally best to pick out some
generic parts of the protocol you are dissecting into the tap data. A packet
type, a priority or a status code maybe. The structure really needs to be
included in a header file so that it can be included by other components that
want to listen in to the tap.

Once you have these defined, it’s simply a case of populating the protocol
specific structure and then calling `tap_queue_packet`, probably as the last part
of the dissector.

.Calling a protocol tap
[source,c]
----
struct FooTap {
    int packet_type;
    int priority;
       ...
};

static int
dissect_foo(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree, void *data _U_)
{
       ...
    struct FooTap *fooinfo = wmem_new0(pinfo->pool, struct FooTap);
    fooinfo->packet_type = tvb_get_uint8(tvb, 0);
    fooinfo->priority = tvb_get_ntohs(tvb, 8);
       ...
    tap_queue_packet(foo_tap, pinfo, fooinfo);

    return tvb_captured_length(tvb);
}
----

[TIP]
====
Allocate your structure using `wmem_new0()`, so it sets all values of your structure
to zero. This way, if you add members later but forget to initialize them, they will
have a consistent value, making troubleshooting easier.
====

This now enables those interested parties to listen in on the details
of this protocol conversation.

[#ChDissectStats]

=== How to produce protocol statistics (stats)

Given that you have a tap interface for the protocol, you can use this
to produce some interesting statistics (well presumably interesting!) from
protocol traces.

This can be done in a separate plugin, or in the same plugin that is
doing the dissection. The latter scheme is better, as the tap and stats
module typically rely on sharing protocol specific data, which might get out
of step between two different plugins.

Here is a mechanism to produce statistics from the above TAP interface.

.Initialising a stats interface
[source,c]
----
#include <epan/stats_tree.h>

void proto_reg_handoff_foo(void) {
       ...
    stats_tree_register("foo", "foo", "Foo" STATS_TREE_MENU_SEPARATOR "Packet Types", 0,
        foo_stats_tree_packet, foo_stats_tree_init, NULL);
}
----

The interface entry point, `proto_reg_handoff_foo()`,
calls the `stats_tree_register()` function, which takes three
strings, an integer, and three callback functions:

. This is the tap name that was registered using `register_tap()`.

. An abbreviation of the stats name.

. The name of the stats module. `STATS_TREE_MENU_SEPARATOR` can be used to make sub menus.

. Flags for per-packet callback, taken from `epan/stats_tree.h`.

. The function that will called to generate the stats.

. A function that can be called to initialise the stats data.

. A function that will be called to clean up the stats data.

In this case we only need the first two functions, as there is nothing specific to clean up.

[NOTE]
====
If you are registering statistics from a plugin, then your plugin should have
a plugin interface entry point called `plugin_register_tap_listener()`,
which should call `stats_tree_register_plugin()` instead of `stats_tree_register()`.
====

.Initialising a stats session
[source,c]
----
static const uint8_t* st_str_packets = "Total Packets";
static const uint8_t* st_str_packet_types = "FOO Packet Types";
static int st_node_packets = -1;
static int st_node_packet_types = -1;

static void foo_stats_tree_init(stats_tree* st)
{
    st_node_packets = stats_tree_create_node(st, st_str_packets, 0, STAT_DT_INT, true);
    st_node_packet_types = stats_tree_create_pivot(st, st_str_packet_types, st_node_packets);
}
----

In this case we create a new tree node, to handle the total packets,
and as a child of that we create a pivot table to handle the stats about
different packet types.


.Generating the stats
[source,c]
----
static tap_packet_status foo_stats_tree_packet(stats_tree* st, packet_info* pinfo, epan_dissect_t* edt, const void* p, tap_flags_t flags)
{
    struct FooTap *pi = (struct FooTap *)p;
    tick_stat_node(st, st_str_packets, 0, false);
    stats_tree_tick_pivot(st, st_node_packet_types,
            val_to_str(pi->packet_type, packettypenames, "Unknown packet type (%d)"));
    return TAP_PACKET_REDRAW;
}
----

In this case the processing of the stats is quite simple. First we call the
`tick_stat_node` for the `st_str_packets` packet node, to count packets. Then a
call to `stats_tree_tick_pivot()` on the `st_node_packet_types` subtree allows
us to record statistics by packet type.

[NOTE]
====
Notice that stats trees and pivots are identified by their name string,
_not_ by the identifier returned by
`stats_tree_create_node()`/`stats_tree_create_pivot()`.
====

[#ChDissectConversation]

=== How to use conversations

Some info about how to use conversations in a dissector can be found in the file
_doc/README.dissector_, chapter 2.2.

[#ChDissectIdl2wrs]

=== __idl2wrs__: Creating dissectors from CORBA IDL files

Many of Wireshark’s dissectors are automatically generated. This section shows
how to generate one from a CORBA IDL file.

==== What is it?

As you have probably guessed from the name, `idl2wrs` takes a user specified IDL
file and attempts to build a dissector that can decode the IDL traffic over
GIOP. The resulting file is “C” code, that should compile okay as a Wireshark
dissector.

`idl2wrs` parses the data struct given to it by the `omniidl` compiler,
and using the GIOP API available in packet-giop.[ch], generates get_CDR_xxx
calls to decode the CORBA traffic on the wire.

It consists of 4 main files.

README.idl2wrs::
This document

wireshark_be.py::
The main compiler backend

wireshark_gen.py::
A helper class, that generates the C code.

idl2wrs::
A simple shell script wrapper that the end user should use to generate the
dissector from the IDL file(s).

==== Why do this?

It is important to understand what CORBA traffic looks like over GIOP/IIOP, and
to help build a tool that can assist in troubleshooting CORBA interworking. This
was especially the case after seeing a lot of discussions about how particular
IDL types are represented inside an octet stream.

I have also had comments/feedback that this tool would be good for say a CORBA
class when teaching students what CORBA traffic looks like “on the wire”.

It is also COOL to work on a great Open Source project such as the case with
“Wireshark” ({wireshark-main-url}).


==== How to use idl2wrs

To use the idl2wrs to generate Wireshark dissectors, you need the following:

* Python must be installed.  See link:https://python.org/[]

* `omniidl` from the omniORB package must be available. See link:http://omniorb.sourceforge.net/[]

* Of course you need Wireshark installed to compile the code and tweak it if
required. idl2wrs is part of the standard Wireshark distribution

To use idl2wrs to generate an Wireshark dissector from an idl file use the following procedure:

* To write the C code to stdout.
+
--
----
$ idl2wrs <your_file.idl>
----

e.g.:

----
$ idl2wrs echo.idl
----
--

* To write to a file, just redirect the output.
+
--
----
$ idl2wrs echo.idl > packet-test-idl.c
----

You may wish to comment out the register_giop_user_module() code and that will
leave you with heuristic dissection.

If you don't want to use the shell script wrapper, then try steps 3 or 4 instead.
--

* To write the C code to stdout.
+
--
----
$ omniidl  -p ./ -b wireshark_be <your file.idl>
----

e.g.:

----
$ omniidl  -p ./ -b wireshark_be echo.idl
----
--

* To write to a file, just redirect the output.
+
--
----
$ omniidl  -p ./ -b wireshark_be echo.idl > packet-test-idl.c
----

You may wish to comment out the register_giop_user_module() code and that will
leave you with heuristic dissection.
--

* Copy the resulting C code to subdirectory epan/dissectors/ inside your
Wireshark source directory.
+
--
----
$ cp packet-test-idl.c /dir/where/wireshark/lives/epan/dissectors/
----

The new dissector has to be added to CMakeLists.txt in the same directory. Look
for the declaration DISSECTOR_SRC and add the new dissector there. For
example,

----
DISSECTOR_SRC = \
        ${CMAKE_CURRENT_SOURCE_DIR}/packet-2dparityfec.c
        ${CMAKE_CURRENT_SOURCE_DIR}/packet-3com-njack.c
        ...
----

becomes

----
DISSECTOR_SRC = \
        ${CMAKE_CURRENT_SOURCE_DIR}/packet-test-idl.c       \
        ${CMAKE_CURRENT_SOURCE_DIR}/packet-2dparityfec.c    \
        ${CMAKE_CURRENT_SOURCE_DIR}/packet-3com-njack.c     \
        ...
----
--

For the next steps, go up to the top of your Wireshark source directory.

* Create a build dir
+
--
----
$ mkdir build && cd build
----
--

* Run cmake
+
--
----
$ cmake ..
----
--

* Build the code
+
--
----
$ make
----
--

* Good Luck !!

==== TODO

* Exception code not generated  (yet), but can be added manually.

* Enums not converted to symbolic values (yet), but can be added manually.

* Add command line options etc

* More I am sure :-)

==== Limitations

See the TODO list inside _packet-giop.c_

==== Notes

The `-p ./` option passed to omniidl indicates that the wireshark_be.py and
wireshark_gen.py are residing in the current directory. This may need tweaking
if you place these files somewhere else.

If it complains about being unable to find some modules (e.g. tempfile.py), you
may want to check if PYTHONPATH is set correctly.


// End of WSDG Chapter Dissection