5 This file is a HOWTO for Wireshark developers. It describes how to start coding
6 a Wireshark protocol dissector and the use of some of the important functions
9 This file is compiled to give in depth information on Wireshark.
10 It is by no means all inclusive and complete. Please feel free to send
11 remarks and patches to the developer mailing list.
15 Before starting to develop a new dissector, a "running" Wireshark build
16 environment is required - there's no such thing as a standalone "dissector
19 How to setup such an environment is platform dependent, detailed information
20 about these steps can be found in the "Developer's Guide" (available from:
21 http://www.wireshark.org) and in the INSTALL and README files of the sources
24 0.1. General README files.
26 You'll find additional information in the following README files:
28 - README.capture - the capture engine internals
29 - README.design - Wireshark software design - incomplete
30 - README.developer - this file
31 - README.display_filter - Display Filter Engine
32 - README.idl2wrs - CORBA IDL converter
33 - README.packaging - how to distribute a software package containing WS
34 - README.regression - regression testing of WS and TS
35 - README.stats_tree - a tree statistics counting specific packets
36 - README.tapping - "tap" a dissector to get protocol specific events
37 - README.xml-output - how to work with the PDML exported output
38 - wiretap/README.developer - how to add additional capture file types to
41 0.2. Dissector related README files.
43 You'll find additional dissector related information in the following README
46 - README.binarytrees - fast access to large data collections
47 - README.malloc - how to obtain "memory leak free" memory
48 - README.plugins - how to "pluginize" a dissector
49 - README.request_response_tracking - how to track req./resp. times and such
53 James Coe <jammer[AT]cin.net>
54 Gilbert Ramirez <gram[AT]alumni.rice.edu>
55 Jeff Foster <jfoste[AT]woodward.com>
56 Olivier Abad <oabad[AT]cybercable.fr>
57 Laurent Deniel <laurent.deniel[AT]free.fr>
58 Gerald Combs <gerald[AT]wireshark.org>
59 Guy Harris <guy[AT]alum.mit.edu>
60 Ulf Lamping <ulf.lamping[AT]web.de>
62 1. Setting up your protocol dissector code.
64 This section provides skeleton code for a protocol dissector. It also explains
65 the basic functions needed to enter values in the traffic summary columns,
66 add to the protocol tree, and work with registered header fields.
72 Wireshark runs on many platforms, and can be compiled with a number of
73 different compilers; here are some rules for writing code that will work
74 on multiple platforms.
76 Don't use C++-style comments (comments beginning with "//" and running
77 to the end of the line); Wireshark's dissectors are written in C, and
78 thus run through C rather than C++ compilers, and not all C compilers
79 support C++-style comments (GCC does, but IBM's C compiler for AIX, for
80 example, doesn't do so by default).
82 Don't initialize variables in their declaration with non-constant
83 values. Not all compilers support this. E.g. don't use
84 guint32 i = somearray[2];
90 Don't use zero-length arrays; not all compilers support them. If an
91 array would have no members, just leave it out.
93 Don't declare variables in the middle of executable code; not all C
94 compilers support that. Variables should be declared outside a
95 function, or at the beginning of a function or compound statement.
97 Don't use anonymous unions; not all compilers support it.
104 } u; /* have a name here */
107 Don't use "uchar", "u_char", "ushort", "u_short", "uint", "u_int",
108 "ulong", "u_long" or "boolean"; they aren't defined on all platforms.
109 If you want an 8-bit unsigned quantity, use "guint8"; if you want an
110 8-bit character value with the 8th bit not interpreted as a sign bit,
111 use "guchar"; if you want a 16-bit unsigned quantity, use "guint16";
112 if you want a 32-bit unsigned quantity, use "guint32"; and if you want
113 an "int-sized" unsigned quantity, use "guint"; if you want a boolean,
114 use "gboolean". Use "%d", "%u", "%x", and "%o" to print those types;
115 don't use "%ld", "%lu", "%lx", or "%lo", as longs are 64 bits long on
116 many platforms, but "guint32" is 32 bits long.
118 Don't use "long" to mean "signed 32-bit integer", and don't use
119 "unsigned long" to mean "unsigned 32-bit integer"; "long"s are 64 bits
120 long on many platforms. Use "gint32" for signed 32-bit integers and use
121 "guint32" for unsigned 32-bit integers.
123 Don't use "long" to mean "signed 64-bit integer" and don't use "unsigned
124 long" to mean "unsigned 64-bit integer"; "long"s are 32 bits long on
125 many other platforms. Don't use "long long" or "unsigned long long",
126 either, as not all platforms support them; use "gint64" or "guint64",
127 which will be defined as the appropriate types for 64-bit signed and
130 When printing or displaying the values of 64-bit integral data types,
131 don't use "%lld", "%llu", "%llx", or "%llo" - not all platforms
132 support "%ll" for printing 64-bit integral data types. Instead, for
133 GLib routines, and routines that use them, such as all the routines in
134 Wireshark that take format arguments, use G_GINT64_MODIFIER, for example:
136 proto_tree_add_text(tree, tvb, offset, 8,
137 "Sequence Number: %" G_GINT64_MODIFIER "u",
140 When using standard C routines, such as printf and scanf, use
141 PRId64, PRIu64, PRIx64, PRIX64, and PRIo64, for example:
143 printf("Sequence Number: %" PRIu64 "\n", sequence_number);
145 When specifying an integral constant that doesn't fit in 32 bits, don't
146 use "LL" at the end of the constant - not all compilers use "LL" for
147 that. Instead, put the constant in a call to the "G_GINT64_CONSTANT()"
150 G_GINT64_CONSTANT(11644473600U)
156 Don't use a label without a statement following it. For example,
166 will not work with all compilers - you have to do
176 with some statement, even if it's a null statement, after the label.
178 Don't use "bzero()", "bcopy()", or "bcmp()"; instead, use the ANSI C
181 "memset()" (with zero as the second argument, so that it sets
182 all the bytes to zero);
184 "memcpy()" or "memmove()" (note that the first and second
185 arguments to "memcpy()" are in the reverse order to the
186 arguments to "bcopy()"; note also that "bcopy()" is typically
187 guaranteed to work on overlapping memory regions, while
188 "memcpy()" isn't, so if you may be copying from one region to a
189 region that overlaps it, use "memmove()", not "memcpy()" - but
190 "memcpy()" might be faster as a result of not guaranteeing
191 correct operation on overlapping memory regions);
193 and "memcmp()" (note that "memcmp()" returns 0, 1, or -1, doing
194 an ordered comparison, rather than just returning 0 for "equal"
195 and 1 for "not equal", as "bcmp()" does).
197 Not all platforms necessarily have "bzero()"/"bcopy()"/"bcmp()", and
198 those that do might not declare them in the header file on which they're
199 declared on your platform.
201 Don't use "index()" or "rindex()"; instead, use the ANSI C equivalents,
202 "strchr()" and "strrchr()". Not all platforms necessarily have
203 "index()" or "rindex()", and those that do might not declare them in the
204 header file on which they're declared on your platform.
206 Don't fetch data from packets by getting a pointer to data in the packet
207 with "tvb_get_ptr()", casting that pointer to a pointer to a structure,
208 and dereferencing that pointer. That pointer won't necessarily be aligned
209 on the proper boundary, which can cause crashes on some platforms (even
210 if it doesn't crash on an x86-based PC); furthermore, the data in a
211 packet is not necessarily in the byte order of the machine on which
212 Wireshark is running. Use the tvbuff routines to extract individual
213 items from the packet, or use "proto_tree_add_item()" and let it extract
216 Don't use structures that overlay packet data, or into which you copy
217 packet data; the C programming language does not guarantee any
218 particular alignment of fields within a structure, and even the
219 extensions that try to guarantee that are compiler-specific and not
220 necessarily supported by all compilers used to build Wireshark. Using
221 bitfields in those structures is even worse; the order of bitfields
224 Don't use "ntohs()", "ntohl()", "htons()", or "htonl()"; the header
225 files required to define or declare them differ between platforms, and
226 you might be able to get away with not including the appropriate header
227 file on your platform but that might not work on other platforms.
228 Instead, use "g_ntohs()", "g_ntohl()", "g_htons()", and "g_htonl()";
229 those are declared by <glib.h>, and you'll need to include that anyway,
230 as Wireshark header files that all dissectors must include use stuff from
233 Don't fetch a little-endian value using "tvb_get_ntohs() or
234 "tvb_get_ntohl()" and then using "g_ntohs()", "g_htons()", "g_ntohl()",
235 or "g_htonl()" on the resulting value - the g_ routines in question
236 convert between network byte order (big-endian) and *host* byte order,
237 not *little-endian* byte order; not all machines on which Wireshark runs
238 are little-endian, even though PCs are. Fetch those values using
239 "tvb_get_letohs()" and "tvb_get_letohl()".
241 Don't put a comma after the last element of an enum - some compilers may
242 either warn about it (producing extra noise) or refuse to accept it.
244 Don't include <unistd.h> without protecting it with
252 and, if you're including it to get routines such as "open()", "close()",
253 "read()", and "write()" declared, also include <io.h> if present:
259 in order to declare the Windows C library routines "_open()",
260 "_close()", "_read()", and "_write()". Your file must include <glib.h>
261 - which many of the Wireshark header files include, so you might not have
262 to include it explicitly - in order to get "open()", "close()",
263 "read()", "write()", etc. mapped to "_open()", "_close()", "_read()",
266 Do not use "open()", "rename()", "mkdir()", "stat()", "unlink()", "remove()",
267 "fopen()", "freopen()" directly. Instead use "ws_open()", "ws_rename()",
268 "ws_mkdir()", "ws_stat()", "ws_unlink()", "ws_remove()", "ws_fopen()",
269 "ws_freopen()": these wrapper functions change the path and file name from
270 UTF8 to UTF16 on Windows allowing the functions to work correctly when the
271 path or file name contain non-ASCII characters.
273 When opening a file with "ws_fopen()", "ws_freopen()", or "ws_fdopen()", if
274 the file contains ASCII text, use "r", "w", "a", and so on as the open mode
275 - but if it contains binary data, use "rb", "wb", and so on. On
276 Windows, if a file is opened in a text mode, writing a byte with the
277 value of octal 12 (newline) to the file causes two bytes, one with the
278 value octal 15 (carriage return) and one with the value octal 12, to be
279 written to the file, and causes bytes with the value octal 15 to be
280 discarded when reading the file (to translate between C's UNIX-style
281 lines that end with newline and Windows' DEC-style lines that end with
282 carriage return/line feed).
284 In addition, that also means that when opening or creating a binary
285 file, you must use "ws_open()" (with O_CREAT and possibly O_TRUNC if the
286 file is to be created if it doesn't exist), and OR in the O_BINARY flag.
287 That flag is not present on most, if not all, UNIX systems, so you must
294 to properly define it for UNIX (it's not necessary on UNIX).
296 Don't use forward declarations of static arrays without a specified size
297 in a fashion such as this:
299 static const value_string foo_vals[];
303 static const value_string foo_vals[] = {
310 as some compilers will reject the first of those statements. Instead,
311 initialize the array at the point at which it's first declared, so that
314 Don't put a comma after the last tuple of an initializer of an array.
316 For #define names and enum member names, prefix the names with a tag so
317 as to avoid collisions with other names - this might be more of an issue
318 on Windows, as it appears to #define names such as DELETE and
321 Don't use the "numbered argument" feature that many UNIX printf's
324 g_snprintf(add_string, 30, " - (%1$d) (0x%1$04x)", value);
326 as not all UNIX printf's implement it, and Windows printf doesn't appear
327 to implement it. Use something like
329 g_snprintf(add_string, 30, " - (%d) (0x%04x)", value, value);
333 Don't use "variadic macros", such as
335 #define DBG(format, args...) fprintf(stderr, format, ## args)
337 as not all C compilers support them. Use macros that take a fixed
338 number of arguments, such as
340 #define DBG0(format) fprintf(stderr, format)
341 #define DBG1(format, arg1) fprintf(stderr, format, arg1)
342 #define DBG2(format, arg1, arg2) fprintf(stderr, format, arg1, arg2)
348 #define DBG(args) printf args
354 as that's not supported by all compilers.
356 snprintf() -> g_snprintf()
357 snprintf() is not available on all platforms, so it's a good idea to use the
358 g_snprintf() function declared by <glib.h> instead.
360 tmpnam() -> mkstemp()
361 tmpnam is insecure and should not be used any more. Wireshark brings its
362 own mkstemp implementation for use on platforms that lack mkstemp.
363 Note: mkstemp does not accept NULL as a parameter.
365 The pointer returned by a call to "tvb_get_ptr()" is not guaranteed to be
366 aligned on any particular byte boundary; this means that you cannot
367 safely cast it to any data type other than a pointer to "char",
368 "unsigned char", "guint8", or other one-byte data types. You cannot,
369 for example, safely cast it to a pointer to a structure, and then access
370 the structure members directly; on some systems, unaligned accesses to
371 integral data types larger than 1 byte, and floating-point data types,
372 cause a trap, which will, at best, result in the OS slowly performing an
373 unaligned access for you, and will, on at least some platforms, cause
374 the program to be terminated.
376 Wireshark supports platforms with GLib 2.4[.x]/GTK+ 2.4[.x] or newer.
377 If a Glib/GTK+ mechanism is available only in Glib/GTK+ versions
378 newer than 2.4/2.4 then use "#if GTK_CHECK_VERSION(...)" to conditionally
379 compile code using that mechanism.
381 When different code must be used on UN*X and Win32, use a #if or #ifdef
382 that tests _WIN32, not WIN32. Try to write code portably whenever
383 possible, however; note that there are some routines in Wireshark with
384 platform-dependent implementations and platform-independent APIs, such
385 as the routines in epan/filesystem.c, allowing the code that calls it to
386 be written portably without #ifdefs.
388 1.1.2 String handling
390 Do not use functions such as strcat() or strcpy().
391 A lot of work has been done to remove the existing calls to these functions and
392 we do not want any new callers of these functions.
394 Instead use g_snprintf() since that function will if used correctly prevent
395 buffer overflows for large strings.
397 When using a buffer to create a string, do not use a buffer stored on the stack.
398 I.e. do not use a buffer declared as
400 instead allocate a buffer dynamically using the emem routines (see
401 README.malloc) such as
404 #define MAX_BUFFER 1024
405 buffer=ep_alloc(MAX_BUFFER);
408 g_snprintf(buffer, MAX_BUFFER, ...
410 This avoids the stack from being corrupted in case there is a bug in your code
411 that accidentally writes beyond the end of the buffer.
414 If you write a routine that will create and return a pointer to a filled in
415 string and if that buffer will not be further processed or appended to after
416 the routine returns (except being added to the proto tree),
417 do not preallocate the buffer to fill in and pass as a parameter instead
418 pass a pointer to a pointer to the function and return a pointer to an
419 emem allocated buffer that will be automatically freed. (see README.malloc)
421 I.e. do not write code such as
423 foo_to_str(char *string, ... ){
429 foo_to_str(buffer, ...
430 proto_tree_add_text(... buffer ...
432 instead write the code as
434 foo_to_str(char **buffer, ...
436 *buffer=ep_alloc(MAX_BUFFER);
442 foo_to_str(&buffer, ...
443 proto_tree_add_text(... *buffer ...
445 Use ep_ allocated buffers. They are very fast and nice. These buffers are all
446 automatically free()d when the dissection of the current packet ends so you
447 don't have to worry about free()ing them explicitly in order to not leak memory.
448 Please read README.malloc.
452 Wireshark is not guaranteed to read only network traces that contain correctly-
453 formed packets. Wireshark is commonly used to track down networking
454 problems, and the problems might be due to a buggy protocol implementation
455 sending out bad packets.
457 Therefore, protocol dissectors not only have to be able to handle
458 correctly-formed packets without, for example, crashing or looping
459 infinitely, they also have to be able to handle *incorrectly*-formed
460 packets without crashing or looping infinitely.
462 Here are some suggestions for making dissectors more robust in the face
463 of incorrectly-formed packets:
465 Do *NOT* use "g_assert()" or "g_assert_not_reached()" in dissectors.
466 *NO* value in a packet's data should be considered "wrong" in the sense
467 that it's a problem with the dissector if found; if it cannot do
468 anything else with a particular value from a packet's data, the
469 dissector should put into the protocol tree an indication that the
470 value is invalid, and should return. You can use the DISSECTOR_ASSERT
471 macro for that purpose.
473 If you are allocating a chunk of memory to contain data from a packet,
474 or to contain information derived from data in a packet, and the size of
475 the chunk of memory is derived from a size field in the packet, make
476 sure all the data is present in the packet before allocating the buffer.
479 1) Wireshark won't leak that chunk of memory if an attempt to
480 fetch data not present in the packet throws an exception
484 2) it won't crash trying to allocate an absurdly-large chunk of
485 memory if the size field has a bogus large value.
487 If you're fetching into such a chunk of memory a string from the buffer,
488 and the string has a specified size, you can use "tvb_get_*_string()",
489 which will check whether the entire string is present before allocating
490 a buffer for the string, and will also put a trailing '\0' at the end of
493 If you're fetching into such a chunk of memory a 2-byte Unicode string
494 from the buffer, and the string has a specified size, you can use
495 "tvb_get_ephemeral_faked_unicode()", which will check whether the entire
496 string is present before allocating a buffer for the string, and will also
497 put a trailing '\0' at the end of the buffer. The resulting string will be
498 a sequence of single-byte characters; the only Unicode characters that
499 will be handled correctly are those in the ASCII range. (Wireshark's
500 ability to handle non-ASCII strings is limited; it needs to be
503 If you're fetching into such a chunk of memory a sequence of bytes from
504 the buffer, and the sequence has a specified size, you can use
505 "tvb_memdup()", which will check whether the entire sequence is present
506 before allocating a buffer for it.
508 Otherwise, you can check whether the data is present by using
509 "tvb_ensure_bytes_exist()" or by getting a pointer to the data by using
510 "tvb_get_ptr()", although note that there might be problems with using
511 the pointer from "tvb_get_ptr()" (see the item on this in the
512 Portability section above, and the next item below).
514 Note also that you should only fetch string data into a fixed-length
515 buffer if the code ensures that no more bytes than will fit into the
516 buffer are fetched ("the protocol ensures" isn't good enough, as
517 protocol specifications can't ensure only packets that conform to the
518 specification will be transmitted or that only packets for the protocol
519 in question will be interpreted as packets for that protocol by
520 Wireshark). If there's no maximum length of string data to be fetched,
521 routines such as "tvb_get_*_string()" are safer, as they allocate a buffer
522 large enough to hold the string. (Note that some variants of this call
523 require you to free the string once you're finished with it.)
525 If you have gotten a pointer using "tvb_get_ptr()", you must make sure
526 that you do not refer to any data past the length passed as the last
527 argument to "tvb_get_ptr()"; while the various "tvb_get" routines
528 perform bounds checking and throw an exception if you refer to data not
529 available in the tvbuff, direct references through a pointer gotten from
530 "tvb_get_ptr()" do not do any bounds checking.
532 If you have a loop that dissects a sequence of items, each of which has
533 a length field, with the offset in the tvbuff advanced by the length of
534 the item, then, if the length field is the total length of the item, and
535 thus can be zero, you *MUST* check for a zero-length item and abort the
536 loop if you see one. Otherwise, a zero-length item could cause the
537 dissector to loop infinitely. You should also check that the offset,
538 after having the length added to it, is greater than the offset before
539 the length was added to it, if the length field is greater than 24 bits
540 long, so that, if the length value is *very* large and adding it to the
541 offset causes an overflow, that overflow is detected.
543 If you are fetching a length field from the buffer, corresponding to the
544 length of a portion of the packet, and subtracting from that length a
545 value corresponding to the length of, for example, a header in the
546 packet portion in question, *ALWAYS* check that the value of the length
547 field is greater than or equal to the length you're subtracting from it,
548 and report an error in the packet and stop dissecting the packet if it's
549 less than the length you're subtracting from it. Otherwise, the
550 resulting length value will be negative, which will either cause errors
551 in the dissector or routines called by the dissector, or, if the value
552 is interpreted as an unsigned integer, will cause the value to be
553 interpreted as a very large positive value.
555 Any tvbuff offset that is added to as processing is done on a packet
556 should be stored in a 32-bit variable, such as an "int"; if you store it
557 in an 8-bit or 16-bit variable, you run the risk of the variable
560 sprintf() -> g_snprintf()
561 Prevent yourself from using the sprintf() function, as it does not test the
562 length of the given output buffer and might be writing into unintended memory
563 areas. This function is one of the main causes of security problems like buffer
564 exploits and many other bugs that are very hard to find. It's much better to
565 use the g_snprintf() function declared by <glib.h> instead.
567 You should test your dissector against incorrectly-formed packets. This
568 can be done using the randpkt and editcap utilities that come with the
569 Wireshark distribution. Testing using randpkt can be done by generating
570 output at the same layer as your protocol, and forcing Wireshark/TShark
571 to decode it as your protocol, e.g. if your protocol sits on top of UDP:
573 randpkt -c 50000 -t dns randpkt.pcap
574 tshark -nVr randpkt.pcap -d udp.port==53,<myproto>
576 Testing using editcap can be done using preexisting capture files and the
577 "-E" flag, which introduces errors in a capture file. E.g.:
579 editcap -E 0.03 infile.pcap outfile.pcap
580 tshark -nVr outfile.pcap
582 The script fuzz-test.sh is available to help automate these tests.
584 1.1.4 Name convention.
586 Wireshark uses the underscore_convention rather than the InterCapConvention for
587 function names, so new code should probably use underscores rather than
588 intercaps for functions and variable names. This is especially important if you
589 are writing code that will be called from outside your code. We are just
590 trying to keep things consistent for other developers.
592 1.1.5 White space convention.
594 Avoid using tab expansions different from 8 column widths, as not all
595 text editors in use by the developers support this. For a detailed
596 discussion of tabs, spaces, and indentation, see
598 http://www.jwz.org/doc/tabs-vs-spaces.html
600 When creating a new file, you are free to choose an indentation logic.
601 Most of the files in Wireshark tend to use 2-space or 4-space
602 indentation. You are encouraged to write a short comment on the
603 indentation logic at the beginning of this new file, especially if
604 you're using non-mod-8 tabs. The tabs-vs-spaces document above provides
605 examples of Emacs and vi modelines for this purpose.
607 When editing an existing file, try following the existing indentation
608 logic and even if it very tempting, never ever use a restyler/reindenter
609 utility on an existing file. If you run across wildly varying
610 indentation styles within the same file, it might be helpful to send a
611 note to wireshark-dev for guidance.
613 1.1.6 Compiler warnings
615 You should write code that is free of compiler warnings. Such warnings will
616 often indicate questionable code and sometimes even real bugs, so it's best
617 to avoid warnings at all.
619 The compiler flags in the Makefiles are set to "treat warnings as errors",
620 so your code won't even compile when warnings occur.
624 Wireshark requires certain things when setting up a protocol dissector.
625 Below is skeleton code for a dissector that you can copy to a file and
626 fill in. Your dissector should follow the naming convention of packet-
627 followed by the abbreviated name for the protocol. It is recommended
628 that where possible you keep to the IANA abbreviated name for the
629 protocol, if there is one, or a commonly-used abbreviation for the
632 Usually, you will put your newly created dissector file into the directory
633 epan/dissectors, just like all the other packet-....c files already in there.
635 Also, please add your dissector file to the corresponding makefile,
636 described in section "1.9 Editing Makefile.common to add your dissector" below.
638 Dissectors that use the dissector registration to register with a lower level
639 dissector don't need to define a prototype in the .h file. For other
640 dissectors the main dissector routine should have a prototype in a header
641 file whose name is "packet-", followed by the abbreviated name for the
642 protocol, followed by ".h"; any dissector file that calls your dissector
643 should be changed to include that file.
645 You may not need to include all the headers listed in the skeleton
646 below, and you may need to include additional headers. For example, the
655 is needed only if you are using a function from libpcre, e.g. the
656 "pcre_compile()" function.
658 The "$Id$" in the comment will be updated by Subversion when the file is
661 When creating a new file, it is fine to just write "$Id$" as Subversion will
662 automatically fill in the identifier at the time the file will be added to the
663 SVN repository (committed).
665 ------------------------------------Cut here------------------------------------
666 /* packet-PROTOABBREV.c
667 * Routines for PROTONAME dissection
668 * Copyright 200x, YOUR_NAME <YOUR_EMAIL_ADDRESS>
672 * Wireshark - Network traffic analyzer
673 * By Gerald Combs <gerald@wireshark.org>
674 * Copyright 1998 Gerald Combs
676 * Copied from WHATEVER_FILE_YOU_USED (where "WHATEVER_FILE_YOU_USED"
677 * is a dissector file; if you just copied this from README.developer,
678 * don't bother with the "Copied from" - you don't even need to put
679 * in a "Copied from" if you copied an existing dissector, especially
680 * if the bulk of the code in the new dissector is your code)
682 * This program is free software; you can redistribute it and/or
683 * modify it under the terms of the GNU General Public License
684 * as published by the Free Software Foundation; either version 2
685 * of the License, or (at your option) any later version.
687 * This program is distributed in the hope that it will be useful,
688 * but WITHOUT ANY WARRANTY; without even the implied warranty of
689 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
690 * GNU General Public License for more details.
692 * You should have received a copy of the GNU General Public License
693 * along with this program; if not, write to the Free Software
694 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
707 #include <epan/packet.h>
708 #include <epan/prefs.h>
710 /* IF PROTO exposes code to other dissectors, then it must be exported
711 in a header file. If not, a header file is not needed at all. */
712 #include "packet-PROTOABBREV.h"
714 /* Forward declaration we need below */
715 void proto_reg_handoff_PROTOABBREV(void);
717 /* Initialize the protocol and registered fields */
718 static int proto_PROTOABBREV = -1;
719 static int hf_PROTOABBREV_FIELDABBREV = -1;
721 /* Global sample preference ("controls" display of numbers) */
722 static gboolean gPREF_HEX = FALSE;
724 /* Initialize the subtree pointers */
725 static gint ett_PROTOABBREV = -1;
727 /* Code to actually dissect the packets */
729 dissect_PROTOABBREV(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree)
732 /* Set up structures needed to add the protocol subtree and manage it */
734 proto_tree *PROTOABBREV_tree;
736 /* First, if at all possible, do some heuristics to check if the packet cannot
737 * possibly belong to your protocol. This is especially important for
738 * protocols directly on top of TCP or UDP where port collisions are
739 * common place (e.g., even though your protocol uses a well known port,
740 * someone else may set up, for example, a web server on that port which,
741 * if someone analyzed that web server's traffic in Wireshark, would result
742 * in Wireshark handing an HTTP packet to your dissector). For example:
744 /* Check that there's enough data */
745 if (tvb_length(tvb) < /* your protocol's smallest packet size */)
748 /* Get some values from the packet header, probably using tvb_get_*() */
749 if ( /* these values are not possible in PROTONAME */ )
750 /* This packet does not appear to belong to PROTONAME.
751 * Return 0 to give another dissector a chance to dissect it.
755 /* Make entries in Protocol column and Info column on summary display */
756 if (check_col(pinfo->cinfo, COL_PROTOCOL))
757 col_set_str(pinfo->cinfo, COL_PROTOCOL, "PROTOABBREV");
759 /* This field shows up as the "Info" column in the display; you should use
760 it, if possible, to summarize what's in the packet, so that a user looking
761 at the list of packets can tell what type of packet it is. See section 1.5
762 for more information.
764 Before changing the contents of a column you should make sure the column is
765 active by calling "check_col(pinfo->cinfo, COL_*)". If it is not active
766 don't bother setting it.
768 If you are setting the column to a constant string, use "col_set_str()",
769 as it's more efficient than the other "col_set_XXX()" calls.
771 If you're setting it to a string you've constructed, or will be
772 appending to the column later, use "col_add_str()".
774 "col_add_fstr()" can be used instead of "col_add_str()"; it takes
775 "printf()"-like arguments. Don't use "col_add_fstr()" with a format
776 string of "%s" - just use "col_add_str()" or "col_set_str()", as it's
777 more efficient than "col_add_fstr()".
779 If you will be fetching any data from the packet before filling in
780 the Info column, clear that column first, in case the calls to fetch
781 data from the packet throw an exception because they're fetching data
782 past the end of the packet, so that the Info column doesn't have data
783 left over from the previous dissector; do
785 if (check_col(pinfo->cinfo, COL_INFO))
786 col_clear(pinfo->cinfo, COL_INFO);
790 if (check_col(pinfo->cinfo, COL_INFO))
791 col_set_str(pinfo->cinfo, COL_INFO, "XXX Request");
793 /* A protocol dissector can be called in 2 different ways:
795 (a) Operational dissection
797 In this mode, Wireshark is only interested in the way protocols
798 interact, protocol conversations are created, packets are
799 reassembled and handed over to higher-level protocol dissectors.
800 In this mode Wireshark does not build a so-called "protocol
803 (b) Detailed dissection
805 In this mode, Wireshark is also interested in all details of
806 a given protocol, so a "protocol tree" is created.
808 Wireshark distinguishes between the 2 modes with the proto_tree pointer:
812 In the interest of speed, if "tree" is NULL, avoid building a
813 protocol tree and adding stuff to it, or even looking at any packet
814 data needed only if you're building the protocol tree, if possible.
816 Note, however, that you must fill in column information, create
817 conversations, reassemble packets, build any other persistent state
818 needed for dissection, and call subdissectors regardless of whether
819 "tree" is NULL or not. This might be inconvenient to do without
820 doing most of the dissection work; the routines for adding items to
821 the protocol tree can be passed a null protocol tree pointer, in
822 which case they'll return a null item pointer, and
823 "proto_item_add_subtree()" returns a null tree pointer if passed a
824 null item pointer, so, if you're careful not to dereference any null
825 tree or item pointers, you can accomplish this by doing all the
826 dissection work. This might not be as efficient as skipping that
827 work if you're not building a protocol tree, but if the code would
828 have a lot of tests whether "tree" is null if you skipped that work,
829 you might still be better off just doing all that work regardless of
830 whether "tree" is null or not. */
833 /* NOTE: The offset and length values in the call to
834 "proto_tree_add_item()" define what data bytes to highlight in the hex
835 display window when the line in the protocol tree display
836 corresponding to that item is selected.
838 Supplying a length of -1 is the way to highlight all data from the
839 offset to the end of the packet. */
841 /* create display subtree for the protocol */
842 ti = proto_tree_add_item(tree, proto_PROTOABBREV, tvb, 0, -1, FALSE);
844 PROTOABBREV_tree = proto_item_add_subtree(ti, ett_PROTOABBREV);
846 /* add an item to the subtree, see section 1.6 for more information */
847 proto_tree_add_item(PROTOABBREV_tree,
848 hf_PROTOABBREV_FIELDABBREV, tvb, offset, len, FALSE);
851 /* Continue adding tree items to process the packet here */
856 /* If this protocol has a sub-dissector call it here, see section 1.8 */
858 /* Return the amount of data this dissector was able to dissect */
859 return tvb_length(tvb);
863 /* Register the protocol with Wireshark */
865 /* this format is require because a script is used to build the C function
866 that calls all the protocol registration.
870 proto_register_PROTOABBREV(void)
872 module_t *PROTOABBREV_module;
874 /* Setup list of header fields See Section 1.6.1 for details*/
875 static hf_register_info hf[] = {
876 { &hf_PROTOABBREV_FIELDABBREV,
877 { "FIELDNAME", "PROTOABBREV.FIELDABBREV",
878 FIELDTYPE, FIELDBASE, FIELDCONVERT, BITMASK,
879 "FIELDDESCR", HFILL }
883 /* Setup protocol subtree array */
884 static gint *ett[] = {
888 /* Register the protocol name and description */
889 proto_PROTOABBREV = proto_register_protocol("PROTONAME",
890 "PROTOSHORTNAME", "PROTOABBREV");
892 /* Required function calls to register the header fields and subtrees used */
893 proto_register_field_array(proto_PROTOABBREV, hf, array_length(hf));
894 proto_register_subtree_array(ett, array_length(ett));
896 /* Register preferences module (See Section 2.6 for more on preferences) */
897 PROTOABBREV_module = prefs_register_protocol(proto_PROTOABBREV,
898 proto_reg_handoff_PROTOABBREV);
900 /* Register a sample preference */
901 prefs_register_bool_preference(PROTOABBREV_module, "show_hex",
902 "Display numbers in Hex",
903 "Enable to display numerical values in hexadecimal.",
908 /* If this dissector uses sub-dissector registration add a registration routine.
909 This exact format is required because a script is used to find these
910 routines and create the code that calls these routines.
912 This function is also called by preferences whenever "Apply" is pressed
913 (see prefs_register_protocol above) so it should accommodate being called
917 proto_reg_handoff_PROTOABBREV(void)
919 static gboolean inited = FALSE;
923 dissector_handle_t PROTOABBREV_handle;
925 /* Use new_create_dissector_handle() to indicate that dissect_PROTOABBREV()
926 * returns the number of bytes it dissected (or 0 if it thinks the packet
927 * does not belong to PROTONAME).
929 PROTOABBREV_handle = new_create_dissector_handle(dissect_PROTOABBREV,
931 dissector_add("PARENT_SUBFIELD", ID_VALUE, PROTOABBREV_handle);
937 If you perform registration functions which are dependent upon
938 prefs the you should de-register everything which was associated
939 with the previous settings and re-register using the new prefs
940 settings here. In general this means you need to keep track of what
941 value the preference had at the time you registered using a local
942 static in this function. ie.
944 static int currentPort = -1;
946 if (currentPort != -1) {
947 dissector_delete("tcp.port", currentPort, PROTOABBREV_handle);
950 currentPort = gPortPref;
952 dissector_add("tcp.port", currentPort, PROTOABBREV_handle);
957 ------------------------------------Cut here------------------------------------
959 1.3 Explanation of needed substitutions in code skeleton.
961 In the above code block the following strings should be substituted with
964 YOUR_NAME Your name, of course. You do want credit, don't you?
965 It's the only payment you will receive....
966 YOUR_EMAIL_ADDRESS Keep those cards and letters coming.
967 WHATEVER_FILE_YOU_USED Add this line if you are using another file as a
969 PROTONAME The name of the protocol; this is displayed in the
970 top-level protocol tree item for that protocol.
971 PROTOSHORTNAME An abbreviated name for the protocol; this is displayed
972 in the "Preferences" dialog box if your dissector has
973 any preferences, in the dialog box of enabled protocols,
974 and in the dialog box for filter fields when constructing
976 PROTOABBREV A name for the protocol for use in filter expressions;
977 it shall contain only lower-case letters, digits, and
979 FIELDNAME The displayed name for the header field.
980 FIELDABBREV The abbreviated name for the header field. (NO SPACES)
981 FIELDTYPE FT_NONE, FT_BOOLEAN, FT_UINT8, FT_UINT16, FT_UINT24,
982 FT_UINT32, FT_UINT64, FT_INT8, FT_INT16, FT_INT24, FT_INT32,
983 FT_INT64, FT_FLOAT, FT_DOUBLE, FT_ABSOLUTE_TIME,
984 FT_RELATIVE_TIME, FT_STRING, FT_STRINGZ, FT_EBCDIC,
985 FT_UINT_STRING, FT_ETHER, FT_BYTES, FT_IPv4, FT_IPv6, FT_IPXNET,
986 FT_FRAMENUM, FT_PROTOCOL, FT_GUID, FT_OID
987 FIELDBASE BASE_NONE, BASE_DEC, BASE_HEX, BASE_OCT, BASE_DEC_HEX,
988 BASE_HEX_DEC, BASE_RANGE_STRING
989 FIELDCONVERT VALS(x), RVALS(x), TFS(x), NULL
990 BITMASK Usually 0x0 unless using the TFS(x) field conversion.
991 FIELDDESCR A brief description of the field, or NULL.
992 PARENT_SUBFIELD Lower level protocol field used for lookup, i.e. "tcp.port"
993 ID_VALUE Lower level protocol field value that identifies this protocol
994 For example the TCP or UDP port number
996 If, for example, PROTONAME is "Internet Bogosity Discovery Protocol",
997 PROTOSHORTNAME would be "IBDP", and PROTOABBREV would be "ibdp". Try to
998 conform with IANA names.
1000 1.4 The dissector and the data it receives.
1005 This is only needed if the dissector doesn't use self-registration to
1006 register itself with the lower level dissector, or if the protocol dissector
1007 wants/needs to expose code to other subdissectors.
1009 The dissector must be declared exactly as follows in the file
1010 packet-PROTOABBREV.h:
1013 dissect_PROTOABBREV(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree);
1016 1.4.2 Extracting data from packets.
1018 NOTE: See the file /epan/tvbuff.h for more details.
1020 The "tvb" argument to a dissector points to a buffer containing the raw
1021 data to be analyzed by the dissector; for example, for a protocol
1022 running atop UDP, it contains the UDP payload (but not the UDP header,
1023 or any protocol headers above it). A tvbuffer is an opaque data
1024 structure, the internal data structures are hidden and the data must be
1025 accessed via the tvbuffer accessors.
1029 Bit accessors for a maximum of 8-bits, 16-bits 32-bits and 64-bits:
1031 guint8 tvb_get_bits8(tvbuff_t *tvb, gint bit_offset, gint no_of_bits);
1032 guint16 tvb_get_bits16(tvbuff_t *tvb, gint bit_offset, gint no_of_bits,gboolean little_endian);
1033 guint32 tvb_get_bits32(tvbuff_t *tvb, gint bit_offset, gint no_of_bits,gboolean little_endian);
1034 guint64 tvb_get_bits64(tvbuff_t *tvb, gint bit_offset, gint no_of_bits,gboolean little_endian);
1036 Single-byte accessor:
1038 guint8 tvb_get_guint8(tvbuff_t*, gint offset);
1040 Network-to-host-order accessors for 16-bit integers (guint16), 24-bit
1041 integers, 32-bit integers (guint32), and 64-bit integers (guint64):
1043 guint16 tvb_get_ntohs(tvbuff_t*, gint offset);
1044 guint32 tvb_get_ntoh24(tvbuff_t*, gint offset);
1045 guint32 tvb_get_ntohl(tvbuff_t*, gint offset);
1046 guint64 tvb_get_ntoh64(tvbuff_t*, gint offset);
1048 Network-to-host-order accessors for single-precision and
1049 double-precision IEEE floating-point numbers:
1051 gfloat tvb_get_ntohieee_float(tvbuff_t*, gint offset);
1052 gdouble tvb_get_ntohieee_double(tvbuff_t*, gint offset);
1054 Little-Endian-to-host-order accessors for 16-bit integers (guint16),
1055 24-bit integers, 32-bit integers (guint32), and 64-bit integers
1058 guint16 tvb_get_letohs(tvbuff_t*, gint offset);
1059 guint32 tvb_get_letoh24(tvbuff_t*, gint offset);
1060 guint32 tvb_get_letohl(tvbuff_t*, gint offset);
1061 guint64 tvb_get_letoh64(tvbuff_t*, gint offset);
1063 Little-Endian-to-host-order accessors for single-precision and
1064 double-precision IEEE floating-point numbers:
1066 gfloat tvb_get_letohieee_float(tvbuff_t*, gint offset);
1067 gdouble tvb_get_letohieee_double(tvbuff_t*, gint offset);
1069 Accessors for IPv4 and IPv6 addresses:
1071 guint32 tvb_get_ipv4(tvbuff_t*, gint offset);
1072 void tvb_get_ipv6(tvbuff_t*, gint offset, struct e_in6_addr *addr);
1074 NOTE: IPv4 addresses are not to be converted to host byte order before
1075 being passed to "proto_tree_add_ipv4()". You should use "tvb_get_ipv4()"
1076 to fetch them, not "tvb_get_ntohl()" *OR* "tvb_get_letohl()" - don't,
1077 for example, try to use "tvb_get_ntohl()", find that it gives you the
1078 wrong answer on the PC on which you're doing development, and try
1079 "tvb_get_letohl()" instead, as "tvb_get_letohl()" will give the wrong
1080 answer on big-endian machines.
1084 void tvb_get_ntohguid(tvbuff_t *, gint offset, e_guid_t *guid);
1085 void tvb_get_letohguid(tvbuff_t *, gint offset, e_guid_t *guid);
1089 guint8 *tvb_get_string(tvbuff_t*, gint offset, gint length);
1090 guint8 *tvb_get_ephemeral_string(tvbuff_t*, gint offset, gint length);
1092 Returns a null-terminated buffer containing data from the specified
1093 tvbuff, starting at the specified offset, and containing the specified
1094 length worth of characters (the length of the buffer will be length+1,
1095 as it includes a null character to terminate the string).
1097 tvb_get_string() returns a buffer allocated by g_malloc() so you must
1098 g_free() it when you are finished with the string. Failure to g_free() this
1099 buffer will lead to memory leaks.
1100 tvb_get_ephemeral_string() returns a buffer allocated from a special heap
1101 with a lifetime until the next packet is dissected. You do not need to
1102 free() this buffer, it will happen automatically once the next packet is
1106 guint8 *tvb_get_stringz(tvbuff_t *tvb, gint offset, gint *lengthp);
1107 guint8 *tvb_get_ephemeral_stringz(tvbuff_t *tvb, gint offset, gint *lengthp);
1109 Returns a null-terminated buffer, allocated with "g_malloc()",
1110 containing data from the specified tvbuff, starting at the
1111 specified offset, and containing all characters from the tvbuff up to
1112 and including a terminating null character in the tvbuff. "*lengthp"
1113 will be set to the length of the string, including the terminating null.
1115 tvb_get_stringz() returns a buffer allocated by g_malloc() so you must
1116 g_free() it when you are finished with the string. Failure to g_free() this
1117 buffer will lead to memory leaks.
1118 tvb_get_ephemeral_stringz() returns a buffer allocated from a special heap
1119 with a lifetime until the next packet is dissected. You do not need to
1120 free() this buffer, it will happen automatically once the next packet is
1124 guint8 *tvb_fake_unicode(tvbuff_t*, gint offset, gint length, gboolean little_endian);
1125 guint8 *tvb_get_ephemeral_faked_unicode(tvbuff_t*, gint offset, gint length, gboolean little_endian);
1127 Converts a 2-byte unicode string to an ASCII string.
1128 Returns a null-terminated buffer containing data from the specified
1129 tvbuff, starting at the specified offset, and containing the specified
1130 length worth of characters (the length of the buffer will be length+1,
1131 as it includes a null character to terminate the string).
1133 tvb_fake_unicode() returns a buffer allocated by g_malloc() so you must
1134 g_free() it when you are finished with the string. Failure to g_free() this
1135 buffer will lead to memory leaks.
1136 tvb_get_ephemeral_faked_unicode() returns a buffer allocated from a special
1137 heap with a lifetime until the next packet is dissected. You do not need to
1138 free() this buffer, it will happen automatically once the next packet is
1143 guint8* tvb_memcpy(tvbuff_t*, guint8* target, gint offset, gint length);
1145 Copies into the specified target the specified length's worth of data
1146 from the specified tvbuff, starting at the specified offset.
1148 guint8* tvb_memdup(tvbuff_t*, gint offset, gint length);
1149 guint8* ep_tvb_memdup(tvbuff_t*, gint offset, gint length);
1151 Returns a buffer, allocated with "g_malloc()", containing the specified
1152 length's worth of data from the specified tvbuff, starting at the
1153 specified offset. The ephemeral variant is freed automatically after the
1154 packet is dissected.
1157 /* WARNING! This function is possibly expensive, temporarily allocating
1158 * another copy of the packet data. Furthermore, it's dangerous because once
1159 * this pointer is given to the user, there's no guarantee that the user will
1160 * honor the 'length' and not overstep the boundaries of the buffer.
1162 guint8* tvb_get_ptr(tvbuff_t*, gint offset, gint length);
1164 The reason that tvb_get_ptr() might have to allocate a copy of its data
1165 only occurs with TVBUFF_COMPOSITES, data that spans multiple tvbuffers.
1166 If the user requests a pointer to a range of bytes that span the member
1167 tvbuffs that make up the TVBUFF_COMPOSITE, the data will have to be
1168 copied to another memory region to assure that all the bytes are
1173 1.5 Functions to handle columns in the traffic summary window.
1175 The topmost pane of the main window is a list of the packets in the
1176 capture, possibly filtered by a display filter.
1178 Each line corresponds to a packet, and has one or more columns, as
1179 configured by the user.
1181 Many of the columns are handled by code outside individual dissectors;
1182 most dissectors need only specify the value to put in the "Protocol" and
1185 Columns are specified by COL_ values; the COL_ value for the "Protocol"
1186 field, typically giving an abbreviated name for the protocol (but not
1187 the all-lower-case abbreviation used elsewhere) is COL_PROTOCOL, and the
1188 COL_ value for the "Info" field, giving a summary of the contents of the
1189 packet for that protocol, is COL_INFO.
1191 A value for a column should only be added if the user specified that it
1192 be displayed; to check whether a given column is to be displayed, call
1193 'check_col' with the COL_ value for that field as an argument - it will
1194 return TRUE if the column is to be displayed and FALSE if it is not to
1197 The value for a column can be specified with one of several functions,
1198 all of which take the 'fd' argument to the dissector as their first
1199 argument, and the COL_ value for the column as their second argument.
1201 1.5.1 The col_set_str function.
1203 'col_set_str' takes a string as its third argument, and sets the value
1204 for the column to that value. It assumes that the pointer passed to it
1205 points to a string constant or a static "const" array, not to a
1206 variable, as it doesn't copy the string, it merely saves the pointer
1207 value; the argument can itself be a variable, as long as it always
1208 points to a string constant or a static "const" array.
1210 It is more efficient than 'col_add_str' or 'col_add_fstr'; however, if
1211 the dissector will be using 'col_append_str' or 'col_append_fstr" to
1212 append more information to the column, the string will have to be copied
1213 anyway, so it's best to use 'col_add_str' rather than 'col_set_str' in
1216 For example, to set the "Protocol" column
1219 if (check_col(pinfo->cinfo, COL_PROTOCOL))
1220 col_set_str(pinfo->cinfo, COL_PROTOCOL, "PROTOABBREV");
1223 1.5.2 The col_add_str function.
1225 'col_add_str' takes a string as its third argument, and sets the value
1226 for the column to that value. It takes the same arguments as
1227 'col_set_str', but copies the string, so that if the string is, for
1228 example, an automatic variable that won't remain in scope when the
1229 dissector returns, it's safe to use.
1232 1.5.3 The col_add_fstr function.
1234 'col_add_fstr' takes a 'printf'-style format string as its third
1235 argument, and 'printf'-style arguments corresponding to '%' format
1236 items in that string as its subsequent arguments. For example, to set
1237 the "Info" field to "<XXX> request, <N> bytes", where "reqtype" is a
1238 string containing the type of the request in the packet and "n" is an
1239 unsigned integer containing the number of bytes in the request:
1241 if (check_col(pinfo->cinfo, COL_INFO))
1242 col_add_fstr(pinfo->cinfo, COL_INFO, "%s request, %u bytes",
1245 Don't use 'col_add_fstr' with a format argument of just "%s" -
1246 'col_add_str', or possibly even 'col_set_str' if the string that matches
1247 the "%s" is a static constant string, will do the same job more
1251 1.5.4 The col_clear function.
1253 If the Info column will be filled with information from the packet, that
1254 means that some data will be fetched from the packet before the Info
1255 column is filled in. If the packet is so small that the data in
1256 question cannot be fetched, the routines to fetch the data will throw an
1257 exception (see the comment at the beginning about tvbuffers improving
1258 the handling of short packets - the tvbuffers keep track of how much
1259 data is in the packet, and throw an exception on an attempt to fetch
1260 data past the end of the packet, so that the dissector won't process
1261 bogus data), causing the Info column not to be filled in.
1263 This means that the Info column will have data for the previous
1264 protocol, which would be confusing if, for example, the Protocol column
1265 had data for this protocol.
1267 Therefore, before a dissector fetches any data whatsoever from the
1268 packet (unless it's a heuristic dissector fetching data to determine
1269 whether the packet is one that it should dissect, in which case it
1270 should check, before fetching the data, whether there's any data to
1271 fetch; if there isn't, it should return FALSE), it should set the
1272 Protocol column and the Info column.
1274 If the Protocol column will ultimately be set to, for example, a value
1275 containing a protocol version number, with the version number being a
1276 field in the packet, the dissector should, before fetching the version
1277 number field or any other field from the packet, set it to a value
1278 without a version number, using 'col_set_str', and should later set it
1279 to a value with the version number after it's fetched the version
1282 If the Info column will ultimately be set to a value containing
1283 information from the packet, the dissector should, before fetching any
1284 fields from the packet, clear the column using 'col_clear' (which is
1285 more efficient than clearing it by calling 'col_set_str' or
1286 'col_add_str' with a null string), and should later set it to the real
1287 string after it's fetched the data to use when doing that.
1290 1.5.5 The col_append_str function.
1292 Sometimes the value of a column, especially the "Info" column, can't be
1293 conveniently constructed at a single point in the dissection process;
1294 for example, it might contain small bits of information from many of the
1295 fields in the packet. 'col_append_str' takes, as arguments, the same
1296 arguments as 'col_add_str', but the string is appended to the end of the
1297 current value for the column, rather than replacing the value for that
1298 column. (Note that no blank separates the appended string from the
1299 string to which it is appended; if you want a blank there, you must add
1300 it yourself as part of the string being appended.)
1303 1.5.6 The col_append_fstr function.
1305 'col_append_fstr' is to 'col_add_fstr' as 'col_append_str' is to
1306 'col_add_str' - it takes, as arguments, the same arguments as
1307 'col_add_fstr', but the formatted string is appended to the end of the
1308 current value for the column, rather than replacing the value for that
1311 1.5.7 The col_append_sep_str and col_append_sep_fstr functions.
1313 In specific situations the developer knows that a column's value will be
1314 created in a stepwise manner, where the appended values are listed. Both
1315 'col_append_sep_str' and 'col_append_sep_fstr' functions will add an item
1316 separator between two consecutive items, and will not add the separator at the
1317 beginning of the column. The remainder of the work both functions do is
1318 identical to what 'col_append_str' and 'col_append_fstr' do.
1320 1.5.8 The col_set_fence and col_prepend_fence_fstr functions.
1322 Sometimes a dissector may be called multiple times for different PDUs in the
1323 same frame (for example in the case of SCTP chunk bundling: several upper
1324 layer data packets may be contained in one SCTP packet). If the upper layer
1325 dissector calls 'col_set_str()' or 'col_clear()' on the Info column when it
1326 begins dissecting each of those PDUs then when the frame is fully dissected
1327 the Info column would contain only the string from the last PDU in the frame.
1328 The 'col_set_fence' function erects a "fence" in the column that prevents
1329 subsequent 'col_...' calls from clearing the data currently in that column.
1330 For example, the SCTP dissector calls 'col_set_fence' on the Info column
1331 after it has called any subdissectors for that chunk so that subdissectors
1332 of any subsequent chunks may only append to the Info column.
1333 'col_prepend_fence_fstr' prepends data before a fence (moving it if
1334 necessary). It will create a fence at the end of the prended data if the
1335 fence does not already exist.
1338 1.5.9 The col_set_time function.
1340 The 'col_set_time' function takes an nstime value as its third argument.
1341 This nstime value is a relative value and will be added as such to the
1342 column. The fourth argument is the filtername holding this value. This
1343 way, rightclicking on the column makes it possible to build a filter
1344 based on the time-value.
1348 if (check_col(pinfo->cinfo, COL_REL_CONV_TIME)) {
1349 nstime_delta(&ts, &pinfo->fd->abs_ts, &tcpd->ts_first);
1350 col_set_time(pinfo->cinfo, COL_REL_CONV_TIME, &ts, "tcp.time_relative");
1354 1.6 Constructing the protocol tree.
1356 The middle pane of the main window, and the topmost pane of a packet
1357 popup window, are constructed from the "protocol tree" for a packet.
1359 The protocol tree, or proto_tree, is a GNode, the N-way tree structure
1360 available within GLIB. Of course the protocol dissectors don't care
1361 what a proto_tree really is; they just pass the proto_tree pointer as an
1362 argument to the routines which allow them to add items and new branches
1365 When a packet is selected in the packet-list pane, or a packet popup
1366 window is created, a new logical protocol tree (proto_tree) is created.
1367 The pointer to the proto_tree (in this case, 'protocol tree'), is passed
1368 to the top-level protocol dissector, and then to all subsequent protocol
1369 dissectors for that packet, and then the GUI tree is drawn via
1372 The logical proto_tree needs to know detailed information about the protocols
1373 and fields about which information will be collected from the dissection
1374 routines. By strictly defining (or "typing") the data that can be attached to a
1375 proto tree, searching and filtering becomes possible. This means that for
1376 every protocol and field (which I also call "header fields", since they are
1377 fields in the protocol headers) which might be attached to a tree, some
1378 information is needed.
1380 Every dissector routine will need to register its protocols and fields
1381 with the central protocol routines (in proto.c). At first I thought I
1382 might keep all the protocol and field information about all the
1383 dissectors in one file, but decentralization seemed like a better idea.
1384 That one file would have gotten very large; one small change would have
1385 required a re-compilation of the entire file. Also, by allowing
1386 registration of protocols and fields at run-time, loadable modules of
1387 protocol dissectors (perhaps even user-supplied) is feasible.
1389 To do this, each protocol should have a register routine, which will be
1390 called when Wireshark starts. The code to call the register routines is
1391 generated automatically; to arrange that a protocol's register routine
1392 be called at startup:
1394 the file containing a dissector's "register" routine must be
1395 added to "DISSECTOR_SRC" in "epan/dissectors/Makefile.common";
1397 the "register" routine must have a name of the form
1398 "proto_register_XXX";
1400 the "register" routine must take no argument, and return no
1403 the "register" routine's name must appear in the source file
1404 either at the beginning of the line, or preceded only by "void "
1405 at the beginning of the line (that would typically be the
1406 definition) - other white space shouldn't cause a problem, e.g.:
1408 void proto_register_XXX(void) {
1417 proto_register_XXX( void )
1424 and so on should work.
1426 For every protocol or field that a dissector wants to register, a variable of
1427 type int needs to be used to keep track of the protocol. The IDs are
1428 needed for establishing parent/child relationships between protocols and
1429 fields, as well as associating data with a particular field so that it
1430 can be stored in the logical tree and displayed in the GUI protocol
1433 Some dissectors will need to create branches within their tree to help
1434 organize header fields. These branches should be registered as header
1435 fields. Only true protocols should be registered as protocols. This is
1436 so that a display filter user interface knows how to distinguish
1437 protocols from fields.
1439 A protocol is registered with the name of the protocol and its
1442 Here is how the frame "protocol" is registered.
1446 proto_frame = proto_register_protocol (
1448 /* short name */ "Frame",
1449 /* abbrev */ "frame" );
1451 A header field is also registered with its name and abbreviation, but
1452 information about its data type is needed. It helps to look at
1453 the header_field_info struct to see what information is expected:
1455 struct header_field_info {
1460 const void *strings;
1468 A string representing the name of the field. This is the name
1469 that will appear in the graphical protocol tree. It must be a non-empty
1474 A string with an abbreviation of the field. We concatenate the
1475 abbreviation of the parent protocol with an abbreviation for the field,
1476 using a period as a separator. For example, the "src" field in an IP packet
1477 would have "ip.src" as an abbreviation. It is acceptable to have
1478 multiple levels of periods if, for example, you have fields in your
1479 protocol that are then subdivided into subfields. For example, TRMAC
1480 has multiple error fields, so the abbreviations follow this pattern:
1481 "trmac.errors.iso", "trmac.errors.noniso", etc.
1483 The abbreviation is the identifier used in a display filter. If it is
1484 an empty string then the field will not be filterable.
1488 The type of value this field holds. The current field types are:
1490 FT_NONE No field type. Used for fields that
1491 aren't given a value, and that can only
1492 be tested for presence or absence; a
1493 field that represents a data structure,
1494 with a subtree below it containing
1495 fields for the members of the structure,
1496 or that represents an array with a
1497 subtree below it containing fields for
1498 the members of the array, might be an
1500 FT_PROTOCOL Used for protocols which will be placing
1501 themselves as top-level items in the
1502 "Packet Details" pane of the UI.
1503 FT_BOOLEAN 0 means "false", any other value means
1505 FT_FRAMENUM A frame number; if this is used, the "Go
1506 To Corresponding Frame" menu item can
1508 FT_UINT8 An 8-bit unsigned integer.
1509 FT_UINT16 A 16-bit unsigned integer.
1510 FT_UINT24 A 24-bit unsigned integer.
1511 FT_UINT32 A 32-bit unsigned integer.
1512 FT_UINT64 A 64-bit unsigned integer.
1513 FT_INT8 An 8-bit signed integer.
1514 FT_INT16 A 16-bit signed integer.
1515 FT_INT24 A 24-bit signed integer.
1516 FT_INT32 A 32-bit signed integer.
1517 FT_INT64 A 64-bit signed integer.
1518 FT_FLOAT A single-precision floating point number.
1519 FT_DOUBLE A double-precision floating point number.
1520 FT_ABSOLUTE_TIME Seconds (4 bytes) and nanoseconds (4 bytes)
1521 of time displayed as month name, month day,
1522 year, hours, minutes, and seconds with 9
1523 digits after the decimal point.
1524 FT_RELATIVE_TIME Seconds (4 bytes) and nanoseconds (4 bytes)
1525 of time displayed as seconds and 9 digits
1526 after the decimal point.
1527 FT_STRING A string of characters, not necessarily
1528 NUL-terminated, but possibly NUL-padded.
1529 This, and the other string-of-characters
1530 types, are to be used for text strings,
1531 not raw binary data.
1532 FT_STRINGZ A NUL-terminated string of characters.
1533 FT_EBCDIC A string of characters, not necessarily
1534 NUL-terminated, but possibly NUL-padded.
1535 The data from the packet is converted from
1536 EBCDIC to ASCII before displaying to the user.
1537 FT_UINT_STRING A counted string of characters, consisting
1538 of a count (represented as an integral value,
1539 of width given in the proto_tree_add_item()
1540 call) followed immediately by that number of
1542 FT_ETHER A six octet string displayed in
1543 Ethernet-address format.
1544 FT_BYTES A string of bytes with arbitrary values;
1545 used for raw binary data.
1546 FT_IPv4 A version 4 IP address (4 bytes) displayed
1547 in dotted-quad IP address format (4
1548 decimal numbers separated by dots).
1549 FT_IPv6 A version 6 IP address (16 bytes) displayed
1550 in standard IPv6 address format.
1551 FT_IPXNET An IPX address displayed in hex as a 6-byte
1552 network number followed by a 6-byte station
1554 FT_GUID A Globally Unique Identifier
1555 FT_OID An ASN.1 Object Identifier
1557 Some of these field types are still not handled in the display filter
1558 routines, but the most common ones are. The FT_UINT* variables all
1559 represent unsigned integers, and the FT_INT* variables all represent
1560 signed integers; the number on the end represent how many bits are used
1561 to represent the number.
1565 The display field has a couple of overloaded uses. This is unfortunate,
1566 but since we're using C as an application programming language, this sometimes
1567 makes for cleaner programs. Right now I still think that overloading
1568 this variable was okay.
1570 For integer fields (FT_UINT* and FT_INT*), this variable represents the
1571 base in which you would like the value displayed. The acceptable bases
1580 BASE_DEC, BASE_HEX, and BASE_OCT are decimal, hexadecimal, and octal,
1581 respectively. BASE_DEC_HEX and BASE_HEX_DEC display value in two bases
1582 (the 1st representation followed by the 2nd in parenthesis).
1584 For FT_BOOLEAN fields that are also bitfields, 'display' is used to tell
1585 the proto_tree how wide the parent bitfield is. With integers this is
1586 not needed since the type of integer itself (FT_UINT8, FT_UINT16,
1587 FT_UINT24, FT_UINT32, etc.) tells the proto_tree how wide the parent
1590 Additionally, BASE_NONE is used for 'display' as a NULL-value. That is,
1591 for non-integers and non-bitfield FT_BOOLEANs, you'll want to use BASE_NONE
1592 in the 'display' field. You may not use BASE_NONE for integers.
1594 It is possible that in the future we will record the endianness of
1595 integers. If so, it is likely that we'll use a bitmask on the display field
1596 so that integers would be represented as BEND|BASE_DEC or LEND|BASE_HEX.
1597 But that has not happened yet.
1601 Some integer fields, of type FT_UINT*, need labels to represent the true
1602 value of a field. You could think of those fields as having an
1603 enumerated data type, rather than an integral data type.
1605 A 'value_string' structure is a way to map values to strings.
1607 typedef struct _value_string {
1612 For fields of that type, you would declare an array of "value_string"s:
1614 static const value_string valstringname[] = {
1615 { INTVAL1, "Descriptive String 1" },
1616 { INTVAL2, "Descriptive String 2" },
1620 (the last entry in the array must have a NULL 'strptr' value, to
1621 indicate the end of the array). The 'strings' field would be set to
1622 'VALS(valstringname)'.
1624 If the field has a numeric rather than an enumerated type, the 'strings'
1625 field would be set to NULL.
1627 If the field has a numeric type that might logically fit in ranges of values
1628 one can use a range_string struct.
1630 Thus a 'range_string' structure is a way to map ranges to strings.
1632 typedef struct _range_string {
1635 const gchar *strptr;
1638 For fields of that type, you would declare an array of "range_string"s:
1640 static const range_string rvalstringname[] = {
1641 { INTVAL_MIN1, INTVALMAX1, "Descriptive String 1" },
1642 { INTVAL_MIN2, INTVALMAX2, "Descriptive String 2" },
1646 If INTVAL_MIN equals INTVAL_MAX for a given entry the range_string
1647 behavior collapses to the one of value_string.
1648 For FT_(U)INT* fields that need a 'range_string' struct, the 'strings' field
1649 would be set to 'RVALS(rvalstringname)'. Furthermore, 'display' field must be
1650 ORed with 'BASE_RANGE_STRING' (e.g. BASE_DEC|BASE_RANGE_STRING).
1652 FT_BOOLEANs have a default map of 0 = "False", 1 (or anything else) = "True".
1653 Sometimes it is useful to change the labels for boolean values (e.g.,
1654 to "Yes"/"No", "Fast"/"Slow", etc.). For these mappings, a struct called
1655 true_false_string is used.
1657 typedef struct true_false_string {
1660 } true_false_string;
1662 For Boolean fields for which "False" and "True" aren't the desired
1663 labels, you would declare a "true_false_string"s:
1665 static const true_false_string boolstringname = {
1670 Its two fields are pointers to the string representing truth, and the
1671 string representing falsehood. For FT_BOOLEAN fields that need a
1672 'true_false_string' struct, the 'strings' field would be set to
1673 'TFS(&boolstringname)'.
1675 If the Boolean field is to be displayed as "False" or "True", the
1676 'strings' field would be set to NULL.
1678 Wireshark predefines a whole range of ready made "true_false_string"s
1679 in tfs.h, included via packet.h.
1683 If the field is a bitfield, then the bitmask is the mask which will
1684 leave only the bits needed to make the field when ANDed with a value.
1685 The proto_tree routines will calculate 'bitshift' automatically
1686 from 'bitmask', by finding the rightmost set bit in the bitmask.
1687 This shift is applied before applying string mapping functions or
1689 If the field is not a bitfield, then bitmask should be set to 0.
1693 This is a string giving a proper description of the field. It should be
1694 at least one grammatically complete sentence, or NULL in which case the
1696 It is meant to provide a more detailed description of the field than the
1697 name alone provides. This information will be used in the man page, and
1698 in a future GUI display-filter creation tool. We might also add tooltips
1699 to the labels in the GUI protocol tree, in which case the blurb would
1700 be used as the tooltip text.
1703 1.6.1 Field Registration.
1705 Protocol registration is handled by creating an instance of the
1706 header_field_info struct (or an array of such structs), and
1707 calling the registration function along with the registration ID of
1708 the protocol that is the parent of the fields. Here is a complete example:
1710 static int proto_eg = -1;
1711 static int hf_field_a = -1;
1712 static int hf_field_b = -1;
1714 static hf_register_info hf[] = {
1717 { "Field A", "proto.field_a", FT_UINT8, BASE_HEX, NULL,
1718 0xf0, "Field A represents Apples", HFILL }},
1721 { "Field B", "proto.field_b", FT_UINT16, BASE_DEC, VALS(vs),
1722 0x0, "Field B represents Bananas", HFILL }}
1725 proto_eg = proto_register_protocol("Example Protocol",
1727 proto_register_field_array(proto_eg, hf, array_length(hf));
1729 Be sure that your array of hf_register_info structs is declared 'static',
1730 since the proto_register_field_array() function does not create a copy
1731 of the information in the array... it uses that static copy of the
1732 information that the compiler created inside your array. Here's the
1733 layout of the hf_register_info struct:
1735 typedef struct hf_register_info {
1736 int *p_id; /* pointer to parent variable */
1737 header_field_info hfinfo;
1740 Also be sure to use the handy array_length() macro found in packet.h
1741 to have the compiler compute the array length for you at compile time.
1743 If you don't have any fields to register, do *NOT* create a zero-length
1744 "hf" array; not all compilers used to compile Wireshark support them.
1745 Just omit the "hf" array, and the "proto_register_field_array()" call,
1748 It is OK to have header fields with a different format be registered with
1749 the same abbreviation. For instance, the following is valid:
1751 static hf_register_info hf[] = {
1753 { &hf_field_8bit, /* 8-bit version of proto.field */
1754 { "Field (8 bit)", "proto.field", FT_UINT8, BASE_DEC, NULL,
1755 0x00, "Field represents FOO", HFILL }},
1757 { &hf_field_32bit, /* 32-bit version of proto.field */
1758 { "Field (32 bit)", "proto.field", FT_UINT32, BASE_DEC, NULL,
1759 0x00, "Field represents FOO", HFILL }}
1762 This way a filter expression can match a header field, irrespective of the
1763 representation of it in the specific protocol context. This is interesting
1764 for protocols with variable-width header fields.
1766 The HFILL macro at the end of the struct will set reasonable default values
1767 for internally used fields.
1769 1.6.2 Adding Items and Values to the Protocol Tree.
1771 A protocol item is added to an existing protocol tree with one of a
1772 handful of proto_XXX_DO_YYY() functions.
1774 Remember that it only makes sense to add items to a protocol tree if its
1775 proto_tree pointer is not null. Should you add an item to a NULL tree, then
1776 the proto_XXX_DO_YYY() function will immediately return. The cost of this
1777 function call can be avoided by checking for the tree pointer.
1779 Subtrees can be made with the proto_item_add_subtree() function:
1781 item = proto_tree_add_item(....);
1782 new_tree = proto_item_add_subtree(item, tree_type);
1784 This will add a subtree under the item in question; a subtree can be
1785 created under an item made by any of the "proto_tree_add_XXX" functions,
1786 so that the tree can be given an arbitrary depth.
1788 Subtree types are integers, assigned by
1789 "proto_register_subtree_array()". To register subtree types, pass an
1790 array of pointers to "gint" variables to hold the subtree type values to
1791 "proto_register_subtree_array()":
1793 static gint ett_eg = -1;
1794 static gint ett_field_a = -1;
1796 static gint *ett[] = {
1801 proto_register_subtree_array(ett, array_length(ett));
1803 in your "register" routine, just as you register the protocol and the
1804 fields for that protocol.
1806 There are several functions that the programmer can use to add either
1807 protocol or field labels to the proto_tree:
1810 proto_tree_add_item(tree, id, tvb, start, length, little_endian);
1813 proto_tree_add_bits_item(tree, id, tvb, bit_offset, no_of_bits, little_endian);
1816 proto_tree_add_bits_ret_val(tree, id, tvb, bit_offset, no_of_bits, return_value, little_endian);
1819 proto_tree_add_none_format(tree, id, tvb, start, length, format, ...);
1822 proto_tree_add_protocol_format(tree, id, tvb, start, length,
1826 proto_tree_add_bytes(tree, id, tvb, start, length, start_ptr);
1829 proto_tree_add_bytes_format(tree, id, tvb, start, length, start_ptr,
1833 proto_tree_add_bytes_format_value(tree, id, tvb, start, length,
1834 start_ptr, format, ...);
1837 proto_tree_add_time(tree, id, tvb, start, length, value_ptr);
1840 proto_tree_add_time_format(tree, id, tvb, start, length, value_ptr,
1844 proto_tree_add_time_format_value(tree, id, tvb, start, length,
1845 value_ptr, format, ...);
1848 proto_tree_add_ipxnet(tree, id, tvb, start, length, value);
1851 proto_tree_add_ipxnet_format(tree, id, tvb, start, length, value,
1855 proto_tree_add_ipxnet_format_value(tree, id, tvb, start, length,
1856 value, format, ...);
1859 proto_tree_add_ipv4(tree, id, tvb, start, length, value);
1862 proto_tree_add_ipv4_format(tree, id, tvb, start, length, value,
1866 proto_tree_add_ipv4_format_value(tree, id, tvb, start, length,
1867 value, format, ...);
1870 proto_tree_add_ipv6(tree, id, tvb, start, length, value_ptr);
1873 proto_tree_add_ipv6_format(tree, id, tvb, start, length, value_ptr,
1877 proto_tree_add_ipv6_format_value(tree, id, tvb, start, length,
1878 value_ptr, format, ...);
1881 proto_tree_add_ether(tree, id, tvb, start, length, value_ptr);
1884 proto_tree_add_ether_format(tree, id, tvb, start, length, value_ptr,
1888 proto_tree_add_ether_format_value(tree, id, tvb, start, length,
1889 value_ptr, format, ...);
1892 proto_tree_add_string(tree, id, tvb, start, length, value_ptr);
1895 proto_tree_add_string_format(tree, id, tvb, start, length, value_ptr,
1899 proto_tree_add_string_format_value(tree, id, tvb, start, length,
1900 value_ptr, format, ...);
1903 proto_tree_add_boolean(tree, id, tvb, start, length, value);
1906 proto_tree_add_boolean_format(tree, id, tvb, start, length, value,
1910 proto_tree_add_boolean_format_value(tree, id, tvb, start, length,
1911 value, format, ...);
1914 proto_tree_add_float(tree, id, tvb, start, length, value);
1917 proto_tree_add_float_format(tree, id, tvb, start, length, value,
1921 proto_tree_add_float_format_value(tree, id, tvb, start, length,
1922 value, format, ...);
1925 proto_tree_add_double(tree, id, tvb, start, length, value);
1928 proto_tree_add_double_format(tree, id, tvb, start, length, value,
1932 proto_tree_add_double_format_value(tree, id, tvb, start, length,
1933 value, format, ...);
1936 proto_tree_add_uint(tree, id, tvb, start, length, value);
1939 proto_tree_add_uint_format(tree, id, tvb, start, length, value,
1943 proto_tree_add_uint_format_value(tree, id, tvb, start, length,
1944 value, format, ...);
1947 proto_tree_add_uint64(tree, id, tvb, start, length, value);
1950 proto_tree_add_uint64_format(tree, id, tvb, start, length, value,
1954 proto_tree_add_uint64_format_value(tree, id, tvb, start, length,
1955 value, format, ...);
1958 proto_tree_add_int(tree, id, tvb, start, length, value);
1961 proto_tree_add_int_format(tree, id, tvb, start, length, value,
1965 proto_tree_add_int_format_value(tree, id, tvb, start, length,
1966 value, format, ...);
1969 proto_tree_add_int64(tree, id, tvb, start, length, value);
1972 proto_tree_add_int64_format(tree, id, tvb, start, length, value,
1976 proto_tree_add_int64_format_value(tree, id, tvb, start, length,
1977 value, format, ...);
1980 proto_tree_add_text(tree, tvb, start, length, format, ...);
1983 proto_tree_add_text_valist(tree, tvb, start, length, format, ap);
1986 proto_tree_add_guid(tree, id, tvb, start, length, value_ptr);
1989 proto_tree_add_guid_format(tree, id, tvb, start, length, value_ptr,
1993 proto_tree_add_guid_format_value(tree, id, tvb, start, length,
1994 value_ptr, format, ...);
1997 proto_tree_add_oid(tree, id, tvb, start, length, value_ptr);
2000 proto_tree_add_oid_format(tree, id, tvb, start, length, value_ptr,
2004 proto_tree_add_oid_format_value(tree, id, tvb, start, length,
2005 value_ptr, format, ...);
2008 proto_tree_add_bitmask(tree, tvb, start, header, ett, **fields,
2011 The 'tree' argument is the tree to which the item is to be added. The
2012 'tvb' argument is the tvbuff from which the item's value is being
2013 extracted; the 'start' argument is the offset from the beginning of that
2014 tvbuff of the item being added, and the 'length' argument is the length,
2015 in bytes, of the item, bit_offset is the offset in bits and no_of_bits
2016 is the lenght in bits.
2018 The length of some items cannot be determined until the item has been
2019 dissected; to add such an item, add it with a length of -1, and, when the
2020 dissection is complete, set the length with 'proto_item_set_len()':
2023 proto_item_set_len(ti, length);
2025 The "ti" argument is the value returned by the call that added the item
2026 to the tree, and the "length" argument is the length of the item.
2028 proto_tree_add_item()
2029 ---------------------
2030 proto_tree_add_item is used when you wish to do no special formatting.
2031 The item added to the GUI tree will contain the name (as passed in the
2032 proto_register_*() function) and a value. The value will be fetched
2033 from the tvbuff by proto_tree_add_item(), based on the type of the field
2034 and, for integral and Boolean fields, the byte order of the value; the
2035 byte order is specified by the 'little_endian' argument, which is TRUE
2036 if the value is little-endian and FALSE if it is big-endian.
2038 Now that definitions of fields have detailed information about bitfield
2039 fields, you can use proto_tree_add_item() with no extra processing to
2040 add bitfield values to your tree. Here's an example. Take the Format
2041 Identifier (FID) field in the Transmission Header (TH) portion of the SNA
2042 protocol. The FID is the high nibble of the first byte of the TH. The
2043 FID would be registered like this:
2045 name = "Format Identifier"
2046 abbrev = "sna.th.fid"
2049 strings = sna_th_fid_vals
2052 The bitmask contains the value which would leave only the FID if bitwise-ANDed
2053 against the parent field, the first byte of the TH.
2055 The code to add the FID to the tree would be;
2057 proto_tree_add_item(bf_tree, hf_sna_th_fid, tvb, offset, 1, TRUE);
2059 The definition of the field already has the information about bitmasking
2060 and bitshifting, so it does the work of masking and shifting for us!
2061 This also means that you no longer have to create value_string structs
2062 with the values bitshifted. The value_string for FID looks like this,
2063 even though the FID value is actually contained in the high nibble.
2064 (You'd expect the values to be 0x0, 0x10, 0x20, etc.)
2066 /* Format Identifier */
2067 static const value_string sna_th_fid_vals[] = {
2068 { 0x0, "SNA device <--> Non-SNA Device" },
2069 { 0x1, "Subarea Node <--> Subarea Node" },
2070 { 0x2, "Subarea Node <--> PU2" },
2071 { 0x3, "Subarea Node or SNA host <--> Subarea Node" },
2074 { 0xf, "Adjaced Subarea Nodes" },
2078 The final implication of this is that display filters work the way you'd
2079 naturally expect them to. You'd type "sna.th.fid == 0xf" to find Adjacent
2080 Subarea Nodes. The user does not have to shift the value of the FID to
2081 the high nibble of the byte ("sna.th.fid == 0xf0") as was necessary
2084 proto_tree_add_bits_item()
2085 --------------------------
2086 Adds a number of bits to the protocol tree which does not have to be byte aligned.
2087 The offset and length is in bits.
2090 ..10 1010 10.. .... "value" (formated as FT_ indicates).
2092 proto_tree_add_bits_ret_val()
2093 -----------------------------
2094 Works in the same way but alo returns the value of the read bits.
2096 proto_tree_add_protocol_format()
2097 --------------------------------
2098 proto_tree_add_protocol_format is used to add the top-level item for the
2099 protocol when the dissector routine wants complete control over how the
2100 field and value will be represented on the GUI tree. The ID value for
2101 the protocol is passed in as the "id" argument; the rest of the
2102 arguments are a "printf"-style format and any arguments for that format.
2103 The caller must include the name of the protocol in the format; it is
2104 not added automatically as in proto_tree_add_item().
2106 proto_tree_add_none_format()
2107 ----------------------------
2108 proto_tree_add_none_format is used to add an item of type FT_NONE.
2109 The caller must include the name of the field in the format; it is
2110 not added automatically as in proto_tree_add_item().
2112 proto_tree_add_bytes()
2113 proto_tree_add_time()
2114 proto_tree_add_ipxnet()
2115 proto_tree_add_ipv4()
2116 proto_tree_add_ipv6()
2117 proto_tree_add_ether()
2118 proto_tree_add_string()
2119 proto_tree_add_boolean()
2120 proto_tree_add_float()
2121 proto_tree_add_double()
2122 proto_tree_add_uint()
2123 proto_tree_add_uint64()
2124 proto_tree_add_int()
2125 proto_tree_add_int64()
2126 proto_tree_add_guid()
2127 proto_tree_add_oid()
2128 ------------------------
2129 These routines are used to add items to the protocol tree if either:
2131 the value of the item to be added isn't just extracted from the
2132 packet data, but is computed from data in the packet;
2134 the value was fetched into a variable.
2136 The 'value' argument has the value to be added to the tree.
2138 NOTE: in all cases where the 'value' argument is a pointer, a copy is
2139 made of the object pointed to; if you have dynamically allocated a
2140 buffer for the object, that buffer will not be freed when the protocol
2141 tree is freed - you must free the buffer yourself when you don't need it
2144 For proto_tree_add_bytes(), the 'value_ptr' argument is a pointer to a
2147 For proto_tree_add_time(), the 'value_ptr' argument is a pointer to an
2148 "nstime_t", which is a structure containing the time to be added; it has
2149 'secs' and 'nsecs' members, giving the integral part and the fractional
2150 part of a time in units of seconds, with 'nsecs' being the number of
2151 nanoseconds. For absolute times, "secs" is a UNIX-style seconds since
2152 January 1, 1970, 00:00:00 GMT value.
2154 For proto_tree_add_ipxnet(), the 'value' argument is a 32-bit IPX
2157 For proto_tree_add_ipv4(), the 'value' argument is a 32-bit IPv4
2158 address, in network byte order.
2160 For proto_tree_add_ipv6(), the 'value_ptr' argument is a pointer to a
2161 128-bit IPv6 address.
2163 For proto_tree_add_ether(), the 'value_ptr' argument is a pointer to a
2166 For proto_tree_add_string(), the 'value_ptr' argument is a pointer to a
2169 For proto_tree_add_boolean(), the 'value' argument is a 32-bit integer.
2170 It is masked and shifted as defined by the field info after which zero
2171 means "false", and non-zero means "true".
2173 For proto_tree_add_float(), the 'value' argument is a 'float' in the
2174 host's floating-point format.
2176 For proto_tree_add_double(), the 'value' argument is a 'double' in the
2177 host's floating-point format.
2179 For proto_tree_add_uint(), the 'value' argument is a 32-bit unsigned
2180 integer value, in host byte order. (This routine cannot be used to add
2183 For proto_tree_add_uint64(), the 'value' argument is a 64-bit unsigned
2184 integer value, in host byte order.
2186 For proto_tree_add_int(), the 'value' argument is a 32-bit signed
2187 integer value, in host byte order. (This routine cannot be used to add
2190 For proto_tree_add_int64(), the 'value' argument is a 64-bit signed
2191 integer value, in host byte order.
2193 For proto_tree_add_guid(), the 'value_ptr' argument is a pointer to an
2196 For proto_tree_add_oid(), the 'value_ptr' argument is a pointer to an
2197 ASN.1 Object Identifier.
2199 proto_tree_add_bytes_format()
2200 proto_tree_add_time_format()
2201 proto_tree_add_ipxnet_format()
2202 proto_tree_add_ipv4_format()
2203 proto_tree_add_ipv6_format()
2204 proto_tree_add_ether_format()
2205 proto_tree_add_string_format()
2206 proto_tree_add_boolean_format()
2207 proto_tree_add_float_format()
2208 proto_tree_add_double_format()
2209 proto_tree_add_uint_format()
2210 proto_tree_add_uint64_format()
2211 proto_tree_add_int_format()
2212 proto_tree_add_int64_format()
2213 proto_tree_add_guid_format()
2214 proto_tree_add_oid_format()
2215 ----------------------------
2216 These routines are used to add items to the protocol tree when the
2217 dissector routine wants complete control over how the field and value
2218 will be represented on the GUI tree. The argument giving the value is
2219 the same as the corresponding proto_tree_add_XXX() function; the rest of
2220 the arguments are a "printf"-style format and any arguments for that
2221 format. The caller must include the name of the field in the format; it
2222 is not added automatically as in the proto_tree_add_XXX() functions.
2224 proto_tree_add_bytes_format_value()
2225 proto_tree_add_time_format_value()
2226 proto_tree_add_ipxnet_format_value()
2227 proto_tree_add_ipv4_format_value()
2228 proto_tree_add_ipv6_format_value()
2229 proto_tree_add_ether_format_value()
2230 proto_tree_add_string_format_value()
2231 proto_tree_add_boolean_format_value()
2232 proto_tree_add_float_format_value()
2233 proto_tree_add_double_format_value()
2234 proto_tree_add_uint_format_value()
2235 proto_tree_add_uint64_format_value()
2236 proto_tree_add_int_format_value()
2237 proto_tree_add_int64_format_value()
2238 proto_tree_add_guid_format_value()
2239 proto_tree_add_oid_format_value()
2240 ------------------------------------
2242 These routines are used to add items to the protocol tree when the
2243 dissector routine wants complete control over how the value will be
2244 represented on the GUI tree. The argument giving the value is the same
2245 as the corresponding proto_tree_add_XXX() function; the rest of the
2246 arguments are a "printf"-style format and any arguments for that format.
2247 With these routines, unlike the proto_tree_add_XXX_format() routines,
2248 the name of the field is added automatically as in the
2249 proto_tree_add_XXX() functions; only the value is added with the format.
2251 proto_tree_add_text()
2252 ---------------------
2253 proto_tree_add_text() is used to add a label to the GUI tree. It will
2254 contain no value, so it is not searchable in the display filter process.
2255 This function was needed in the transition from the old-style proto_tree
2256 to this new-style proto_tree so that Wireshark would still decode all
2257 protocols w/o being able to filter on all protocols and fields.
2258 Otherwise we would have had to cripple Wireshark's functionality while we
2259 converted all the old-style proto_tree calls to the new-style proto_tree
2260 calls. In other words, you should not use this in new code unless you've got
2261 a specific reason (see below).
2263 This can also be used for items with subtrees, which may not have values
2264 themselves - the items in the subtree are the ones with values.
2266 For a subtree, the label on the subtree might reflect some of the items
2267 in the subtree. This means the label can't be set until at least some
2268 of the items in the subtree have been dissected. To do this, use
2269 'proto_item_set_text()' or 'proto_item_append_text()':
2272 proto_item_set_text(proto_item *ti, ...);
2275 proto_item_append_text(proto_item *ti, ...);
2277 'proto_item_set_text()' takes as an argument the value returned by
2278 'proto_tree_add_text()', a 'printf'-style format string, and a set of
2279 arguments corresponding to '%' format items in that string, and replaces
2280 the text for the item created by 'proto_tree_add_text()' with the result
2281 of applying the arguments to the format string.
2283 'proto_item_append_text()' is similar, but it appends to the text for
2284 the item the result of applying the arguments to the format string.
2286 For example, early in the dissection, one might do:
2288 ti = proto_tree_add_text(tree, tvb, offset, length, <label>);
2292 proto_item_set_text(ti, "%s: %s", type, value);
2294 after the "type" and "value" fields have been extracted and dissected.
2295 <label> would be a label giving what information about the subtree is
2296 available without dissecting any of the data in the subtree.
2298 Note that an exception might be thrown when trying to extract the values of
2299 the items used to set the label, if not all the bytes of the item are
2300 available. Thus, one should create the item with text that is as
2301 meaningful as possible, and set it or append additional information to
2302 it as the values needed to supply that information are extracted.
2304 proto_tree_add_text_valist()
2305 ----------------------------
2306 This is like proto_tree_add_text(), but takes, as the last argument, a
2307 'va_list'; it is used to allow routines that take a printf-like
2308 variable-length list of arguments to add a text item to the protocol
2311 proto_tree_add_bitmask()
2312 ------------------------
2313 This function provides an easy to use and convenient helper function
2314 to manage many types of common bitmasks that occur in protocols.
2316 This function will dissect a 1/2/3/4 byte large bitmask into its individual
2318 header is an integer type and must be of type FT_[U]INT{8|16|24|32} and
2319 represents the entire width of the bitmask.
2321 'header' and 'ett' are the hf fields and ett field respectively to create an
2322 expansion that covers the 1-4 bytes of the bitmask.
2324 '**fields' is a NULL terminated a array of pointers to hf fields representing
2325 the individual subfields of the bitmask. These fields must either be integers
2326 of the same byte width as 'header' or of the type FT_BOOLEAN.
2327 Each of the entries in '**fields' will be dissected as an item under the
2328 'header' expansion and also IF the field is a boolean and IF it is set to 1,
2329 then the name of that boolean field will be printed on the 'header' expansion
2330 line. For integer type subfields that have a value_string defined, the
2331 matched string from that value_string will be printed on the expansion line as well.
2333 Example: (from the scsi dissector)
2334 static int hf_scsi_inq_peripheral = -1;
2335 static int hf_scsi_inq_qualifier = -1;
2336 static gint ett_scsi_inq_peripheral = -1;
2338 static const int *peripheal_fields[] = {
2339 &hf_scsi_inq_qualifier,
2340 &hf_scsi_inq_devtype,
2344 /* Qualifier and DeviceType */
2345 proto_tree_add_bitmask(tree, tvb, offset, hf_scsi_inq_peripheral, ett_scsi_inq_peripheral, peripheal_fields, FALSE);
2348 { &hf_scsi_inq_peripheral,
2349 {"Peripheral", "scsi.inquiry.preipheral", FT_UINT8, BASE_HEX,
2350 NULL, 0, NULL, HFILL}},
2351 { &hf_scsi_inq_qualifier,
2352 {"Qualifier", "scsi.inquiry.qualifier", FT_UINT8, BASE_HEX,
2353 VALS (scsi_qualifier_val), 0xE0, NULL, HFILL}},
2356 Which provides very pretty dissection of this one byte bitmask.
2358 PROTO_ITEM_SET_HIDDEN()
2359 -----------------------
2360 PROTO_ITEM_SET_HIDDEN is used to hide fields, which have already been added
2361 to the tree, from being visible in the displayed tree.
2363 NOTE that creating hidden fields is actually quite a bad idea from a UI design
2364 perspective because the user (someone who did not write nor has ever seen the
2365 code) has no way of knowing that hidden fields are there to be filtered on
2366 thus defeating the whole purpose of putting them there. A Better Way might
2367 be to add the fields (that might otherwise be hidden) to a subtree where they
2368 won't be seen unless the user opens the subtree--but they can be found if the
2371 One use for hidden fields (which would be better implemented using visible
2372 fields in a subtree) follows: The caller may want a value to be
2373 included in a tree so that the packet can be filtered on this field, but
2374 the representation of that field in the tree is not appropriate. An
2375 example is the token-ring routing information field (RIF). The best way
2376 to show the RIF in a GUI is by a sequence of ring and bridge numbers.
2377 Rings are 3-digit hex numbers, and bridges are single hex digits:
2379 RIF: 001-A-013-9-C0F-B-555
2381 In the case of RIF, the programmer should use a field with no value and
2382 use proto_tree_add_none_format() to build the above representation. The
2383 programmer can then add the ring and bridge values, one-by-one, with
2384 proto_tree_add_item() and hide them with PROTO_ITEM_SET_HIDDEN() so that the
2385 user can then filter on or search for a particular ring or bridge. Here's a
2386 skeleton of how the programmer might code this.
2389 rif = create_rif_string(...);
2391 proto_tree_add_none_format(tree, hf_tr_rif_label, ..., "RIF: %s", rif);
2393 for(i = 0; i < num_rings; i++) {
2396 pi = proto_tree_add_item(tree, hf_tr_rif_ring, ..., FALSE);
2397 PROTO_ITEM_SET_HIDDEN(pi);
2399 for(i = 0; i < num_rings - 1; i++) {
2402 pi = proto_tree_add_item(tree, hf_tr_rif_bridge, ..., FALSE);
2403 PROTO_ITEM_SET_HIDDEN(pi);
2406 The logical tree has these items:
2408 hf_tr_rif_label, text="RIF: 001-A-013-9-C0F-B-555", value = NONE
2409 hf_tr_rif_ring, hidden, value=0x001
2410 hf_tr_rif_bridge, hidden, value=0xA
2411 hf_tr_rif_ring, hidden, value=0x013
2412 hf_tr_rif_bridge, hidden, value=0x9
2413 hf_tr_rif_ring, hidden, value=0xC0F
2414 hf_tr_rif_bridge, hidden, value=0xB
2415 hf_tr_rif_ring, hidden, value=0x555
2417 GUI or print code will not display the hidden fields, but a display
2418 filter or "packet grep" routine will still see the values. The possible
2419 filter is then possible:
2421 tr.rif_ring eq 0x013
2424 1.7 Utility routines.
2426 1.7.1 match_strval and val_to_str.
2428 A dissector may need to convert a value to a string, using a
2429 'value_string' structure, by hand, rather than by declaring a field with
2430 an associated 'value_string' structure; this might be used, for example,
2431 to generate a COL_INFO line for a frame.
2433 'match_strval()' will do that:
2436 match_strval(guint32 val, const value_string *vs)
2438 It will look up the value 'val' in the 'value_string' table pointed to
2439 by 'vs', and return either the corresponding string, or NULL if the
2440 value could not be found in the table. Note that, unless 'val' is
2441 guaranteed to be a value in the 'value_string' table ("guaranteed" as in
2442 "the code has already checked that it's one of those values" or "the
2443 table handles all possible values of the size of 'val'", not "the
2444 protocol spec says it has to be" - protocol specs do not prevent invalid
2445 packets from being put onto a network or into a purported packet capture
2446 file), you must check whether 'match_strval()' returns NULL, and arrange
2447 that its return value not be dereferenced if it's NULL. 'val_to_str()'
2448 can be used to generate a string for values not found in the table:
2451 val_to_str(guint32 val, const value_string *vs, const char *fmt)
2453 If the value 'val' is found in the 'value_string' table pointed to by
2454 'vs', 'val_to_str' will return the corresponding string; otherwise, it
2455 will use 'fmt' as an 'sprintf'-style format, with 'val' as an argument,
2456 to generate a string, and will return a pointer to that string.
2457 You can use it in a call to generate a COL_INFO line for a frame such as
2459 col_add_fstr(COL_INFO, ", %s", val_to_str(val, table, "Unknown %d"));
2461 1.7.2 match_strrval and rval_to_str.
2463 A dissector may need to convert a range of values to a string, using a
2464 'range_string' structure.
2466 'match_strrval()' will do that:
2469 match_strrval(guint32 val, const range_string *rs)
2471 It will look up the value 'val' in the 'range_string' table pointed to
2472 by 'rs', and return either the corresponding string, or NULL if the
2473 value could not be found in the table. Please note that its base
2474 behavior is inherited from match_strval().
2476 'rval_to_str()' can be used to generate a string for values not found in
2480 rval_to_str(guint32 val, const range_string *rs, const char *fmt)
2482 If the value 'val' is found in the 'range_string' table pointed to by
2483 'rs', 'rval_to_str' will return the corresponding string; otherwise, it
2484 will use 'fmt' as an 'sprintf'-style format, with 'val' as an argument,
2485 to generate a string, and will return a pointer to that string. Please
2486 note that its base behavior is inherited from match_strval().
2488 1.8 Calling Other Dissectors.
2490 As each dissector completes its portion of the protocol analysis, it
2491 is expected to create a new tvbuff of type TVBUFF_SUBSET which
2492 contains the payload portion of the protocol (that is, the bytes
2493 that are relevant to the next dissector).
2495 The syntax for creating a new TVBUFF_SUBSET is:
2497 next_tvb = tvb_new_subset(tvb, offset, length, reported_length)
2500 tvb is the tvbuff that the dissector has been working on. It
2501 can be a tvbuff of any type.
2503 next_tvb is the new TVBUFF_SUBSET.
2505 offset is the byte offset of 'tvb' at which the new tvbuff
2506 should start. The first byte is the 0th byte.
2508 length is the number of bytes in the new TVBUFF_SUBSET. A length
2509 argument of -1 says to use as many bytes as are available in
2512 reported_length is the number of bytes that the current protocol
2513 says should be in the payload. A reported_length of -1 says that
2514 the protocol doesn't say anything about the size of its payload.
2517 An example from packet-ipx.c -
2520 dissect_ipx(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree)
2523 int reported_length, available_length;
2526 /* Make the next tvbuff */
2528 /* IPX does have a length value in the header, so calculate report_length */
2529 Set this to -1 if there isn't any length information in the protocol
2531 reported_length = ipx_length - IPX_HEADER_LEN;
2533 /* Calculate the available data in the packet,
2534 set this to -1 to use all the data in the tv_buffer
2536 available_length = tvb_length(tvb) - IPX_HEADER_LEN;
2538 /* Create the tvbuffer for the next dissector */
2539 next_tvb = tvb_new_subset(tvb, IPX_HEADER_LEN,
2540 MIN(available_length, reported_length),
2543 /* call the next dissector */
2544 dissector_next( next_tvb, pinfo, tree);
2547 1.9 Editing Makefile.common to add your dissector.
2549 To arrange that your dissector will be built as part of Wireshark, you
2550 must add the name of the source file for your dissector to the
2551 'DISSECTOR_SRC' macro in the 'Makefile.common' file in the 'epan/dissectors'
2552 directory. (Note that this is for modern versions of UNIX, so there
2553 is no 14-character limitation on file names, and for modern versions of
2554 Windows, so there is no 8.3-character limitation on file names.)
2556 If your dissector also has its own header file or files, you must add
2557 them to the 'DISSECTOR_INCLUDES' macro in the 'Makefile.common' file in
2558 the 'epan/dissectors' directory, so that it's included when release source
2559 tarballs are built (otherwise, the source in the release tarballs won't
2562 1.10 Using the SVN source code tree.
2564 See <http://www.wireshark.org/develop.html>
2566 1.11 Submitting code for your new dissector.
2568 - VERIFY that your dissector code does not use prohibited or deprecated APIs
2570 perl <wireshark_root>/tools/checkAPIs.pl <source-filename(s)>
2572 - TEST YOUR DISSECTOR BEFORE SUBMITTING IT.
2573 Use fuzz-test.sh and/or randpkt against your dissector. These are
2574 described at <http://wiki.wireshark.org/FuzzTesting>.
2576 - Subscribe to <mailto:wireshark-dev[AT]wireshark.org> by sending an email to
2577 <mailto:wireshark-dev-request[AT]wireshark.org?body="help"> or visiting
2578 <http://www.wireshark.org/lists/>.
2580 - 'svn add' all the files of your new dissector.
2582 - 'svn diff' the workspace and save the result to a file.
2584 - Edit the diff file - remove any changes unrelated to your new dissector,
2585 e.g. changes in config.nmake
2587 - Submit a bug report to the Wireshark bug database, found at
2588 <http://bugs.wireshark.org>, qualified as an enhancement and attach your
2589 diff file there. Set the review request flag to '?' so it will pop up in
2590 the patch review list.
2592 - Create a Wiki page on the protocol at <http://wiki.wireshark.org>.
2593 A template is provided so it is easy to setup in a consistent style.
2595 - If possible, add sample capture files to the sample captures page at
2596 <http://wiki.wireshark.org/SampleCaptures>. These files are used by
2597 the automated build system for fuzz testing.
2599 - If you find that you are contributing a lot to wireshark on an ongoing
2600 basis you can request to become a committer which will allow you to
2601 commit files to subversion directly.
2603 2. Advanced dissector topics.
2607 Some of the advanced features are being worked on constantly. When using them
2608 it is wise to check the relevant header and source files for additional details.
2610 2.2 Following "conversations".
2612 In wireshark a conversation is defined as a series of data packets between two
2613 address:port combinations. A conversation is not sensitive to the direction of
2614 the packet. The same conversation will be returned for a packet bound from
2615 ServerA:1000 to ClientA:2000 and the packet from ClientA:2000 to ServerA:1000.
2617 There are five routines that you will use to work with a conversation:
2618 conversation_new, find_conversation, conversation_add_proto_data,
2619 conversation_get_proto_data, and conversation_delete_proto_data.
2622 2.2.1 The conversation_init function.
2624 This is an internal routine for the conversation code. As such you
2625 will not have to call this routine. Just be aware that this routine is
2626 called at the start of each capture and before the packets are filtered
2627 with a display filter. The routine will destroy all stored
2628 conversations. This routine does NOT clean up any data pointers that are
2629 passed in the conversation_new 'data' variable. You are responsible for
2630 this clean up if you pass a malloc'ed pointer in this variable.
2632 See item 2.2.8 for more information about the 'data' pointer.
2635 2.2.2 The conversation_new function.
2637 This routine will create a new conversation based upon two address/port
2638 pairs. If you want to associate with the conversation a pointer to a
2639 private data structure you must use the conversation_add_proto_data
2640 function. The ptype variable is used to differentiate between
2641 conversations over different protocols, i.e. TCP and UDP. The options
2642 variable is used to define a conversation that will accept any destination
2643 address and/or port. Set options = 0 if the destination port and address
2644 are know when conversation_new is called. See section 2.4 for more
2645 information on usage of the options parameter.
2647 The conversation_new prototype:
2648 conversation_t *conversation_new(guint32 setup_frame, address *addr1,
2649 address *addr2, port_type ptype, guint32 port1, guint32 port2,
2653 guint32 setup_frame = The lowest numbered frame for this conversation
2654 address* addr1 = first data packet address
2655 address* addr2 = second data packet address
2656 port_type ptype = port type, this is defined in packet.h
2657 guint32 port1 = first data packet port
2658 guint32 port2 = second data packet port
2659 guint options = conversation options, NO_ADDR2 and/or NO_PORT2
2661 setup_frame indicates the first frame for this conversation, and is used to
2662 distinguish multiple conversations with the same addr1/port1 and addr2/port2
2663 pair that occur within the same capture session.
2665 "addr1" and "port1" are the first address/port pair; "addr2" and "port2"
2666 are the second address/port pair. A conversation doesn't have source
2667 and destination address/port pairs - packets in a conversation go in
2668 both directions - so "addr1"/"port1" may be the source or destination
2669 address/port pair; "addr2"/"port2" would be the other pair.
2671 If NO_ADDR2 is specified, the conversation is set up so that a
2672 conversation lookup will match only the "addr1" address; if NO_PORT2 is
2673 specified, the conversation is set up so that a conversation lookup will
2674 match only the "port1" port; if both are specified, i.e.
2675 NO_ADDR2|NO_PORT2, the conversation is set up so that the lookup will
2676 match only the "addr1"/"port1" address/port pair. This can be used if a
2677 packet indicates that, later in the capture, a conversation will be
2678 created using certain addresses and ports, in the case where the packet
2679 doesn't specify the addresses and ports of both sides.
2681 2.2.3 The find_conversation function.
2683 Call this routine to look up a conversation. If no conversation is found,
2684 the routine will return a NULL value.
2686 The find_conversation prototype:
2688 conversation_t *find_conversation(guint32 frame_num, address *addr_a,
2689 address *addr_b, port_type ptype, guint32 port_a, guint32 port_b,
2693 guint32 frame_num = a frame number to match
2694 address* addr_a = first address
2695 address* addr_b = second address
2696 port_type ptype = port type
2697 guint32 port_a = first data packet port
2698 guint32 port_b = second data packet port
2699 guint options = conversation options, NO_ADDR_B and/or NO_PORT_B
2701 frame_num is a frame number to match. The conversation returned is where
2702 (frame_num >= conversation->setup_frame
2703 && frame_num < conversation->next->setup_frame)
2704 Suppose there are a total of 3 conversations (A, B, and C) that match
2705 addr_a/port_a and addr_b/port_b, where the setup_frame used in
2706 conversation_new() for A, B and C are 10, 50, and 100 respectively. The
2707 frame_num passed in find_conversation is compared to the setup_frame of each
2708 conversation. So if (frame_num >= 10 && frame_num < 50), conversation A is
2709 returned. If (frame_num >= 50 && frame_num < 100), conversation B is returned.
2710 If (frame_num >= 100) conversation C is returned.
2712 "addr_a" and "port_a" are the first address/port pair; "addr_b" and
2713 "port_b" are the second address/port pair. Again, as a conversation
2714 doesn't have source and destination address/port pairs, so
2715 "addr_a"/"port_a" may be the source or destination address/port pair;
2716 "addr_b"/"port_b" would be the other pair. The search will match the
2717 "a" address/port pair against both the "1" and "2" address/port pairs,
2718 and match the "b" address/port pair against both the "2" and "1"
2719 address/port pairs; you don't have to worry about which side the "a" or
2720 "b" pairs correspond to.
2722 If the NO_ADDR_B flag was specified to "find_conversation()", the
2723 "addr_b" address will be treated as matching any "wildcarded" address;
2724 if the NO_PORT_B flag was specified, the "port_b" port will be treated
2725 as matching any "wildcarded" port. If both flags are specified, i.e.
2726 NO_ADDR_B|NO_PORT_B, the "addr_b" address will be treated as matching
2727 any "wildcarded" address and the "port_b" port will be treated as
2728 matching any "wildcarded" port.
2731 2.2.4 The conversation_add_proto_data function.
2733 Once you have created a conversation with conversation_new, you can
2734 associate data with it using this function.
2736 The conversation_add_proto_data prototype:
2738 void conversation_add_proto_data(conversation_t *conv, int proto,
2742 conversation_t *conv = the conversation in question
2743 int proto = registered protocol number
2744 void *data = dissector data structure
2746 "conversation" is the value returned by conversation_new. "proto" is a
2747 unique protocol number created with proto_register_protocol. Protocols
2748 are typically registered in the proto_register_XXXX section of your
2749 dissector. "data" is a pointer to the data you wish to associate with the
2750 conversation. Using the protocol number allows several dissectors to
2751 associate data with a given conversation.
2754 2.2.5 The conversation_get_proto_data function.
2756 After you have located a conversation with find_conversation, you can use
2757 this function to retrieve any data associated with it.
2759 The conversation_get_proto_data prototype:
2761 void *conversation_get_proto_data(conversation_t *conv, int proto);
2764 conversation_t *conv = the conversation in question
2765 int proto = registered protocol number
2767 "conversation" is the conversation created with conversation_new. "proto"
2768 is a unique protocol number created with proto_register_protocol,
2769 typically in the proto_register_XXXX portion of a dissector. The function
2770 returns a pointer to the data requested, or NULL if no data was found.
2773 2.2.6 The conversation_delete_proto_data function.
2775 After you are finished with a conversation, you can remove your association
2776 with this function. Please note that ONLY the conversation entry is
2777 removed. If you have allocated any memory for your data, you must free it
2780 The conversation_delete_proto_data prototype:
2782 void conversation_delete_proto_data(conversation_t *conv, int proto);
2785 conversation_t *conv = the conversation in question
2786 int proto = registered protocol number
2788 "conversation" is the conversation created with conversation_new. "proto"
2789 is a unique protocol number created with proto_register_protocol,
2790 typically in the proto_register_XXXX portion of a dissector.
2793 2.2.7 Using timestamps relative to the conversation
2795 There is a framework to calculate timestamps relative to the start of the
2796 conversation. First of all the timestamp of the first packet that has been
2797 seen in the conversation must be kept in the protocol data to be able
2798 to calculate the timestamp of the current packet relative to the start
2799 of the conversation. The timestamp of the last packet that was seen in the
2800 conversation should also be kept in the protocol data. This way the
2801 delta time between the current packet and the previous packet in the
2802 conversation can be calculated.
2804 So add the following items to the struct that is used for the protocol data:
2809 The ts_prev value should only be set during the first run through the
2810 packets (ie pinfo->fd->flags.visited is false).
2812 Next step is to use the per-packet information (described in section 2.5)
2813 to keep the calculated delta timestamp, as it can only be calculated
2814 on the first run through the packets. This is because a packet can be
2815 selected in random order once the whole file has been read.
2817 After calculating the conversation timestamps, it is time to put them in
2818 the appropriate columns with the function 'col_set_time' (described in
2819 section 1.5.9). There are two columns for conversation timestamps:
2821 COL_REL_CONV_TIME, /* Relative time to beginning of conversation */
2822 COL_DELTA_CONV_TIME,/* Delta time to last frame in conversation */
2824 Last but not least, there MUST be a preference in each dissector that
2825 uses conversation timestamps that makes it possible to enable and
2826 disable the calculation of conversation timestamps. The main argument
2827 for this is that a higher level conversation is able to overwrite
2828 the values of lowel level conversations in these two columns. Being
2829 able to actively select which protocols may overwrite the conversation
2830 timestamp columns gives the user the power to control these columns.
2831 (A second reason is that conversation timestamps use the per-packet
2832 data structure which uses additional memory, which should be avoided
2833 if these timestamps are not needed)
2835 Have a look at the differences to packet-tcp.[ch] in SVN 22966 and
2836 SVN 23058 to see the implementation of conversation timestamps for
2840 2.2.8 The example conversation code with GMemChunk's.
2842 For a conversation between two IP addresses and ports you can use this as an
2843 example. This example uses the GMemChunk to allocate memory and stores the data
2844 pointer in the conversation 'data' variable.
2846 NOTE: Remember to register the init routine (my_dissector_init) in the
2847 protocol_register routine.
2850 /************************ Global values ************************/
2852 /* the number of entries in the memory chunk array */
2853 #define my_init_count 10
2855 /* define your structure here */
2860 /* the GMemChunk base structure */
2861 static GMemChunk *my_vals = NULL;
2863 /* Registered protocol number */
2864 static int my_proto = -1;
2867 /********************* in the dissector routine *********************/
2869 /* the local variables in the dissector */
2871 conversation_t *conversation;
2872 my_entry_t *data_ptr;
2875 /* look up the conversation */
2877 conversation = find_conversation(pinfo->fd->num, &pinfo->src, &pinfo->dst,
2878 pinfo->ptype, pinfo->srcport, pinfo->destport, 0);
2880 /* if conversation found get the data pointer that you stored */
2882 data_ptr = (my_entry_t*)conversation_get_proto_data(conversation, my_proto);
2885 /* new conversation create local data structure */
2887 data_ptr = g_mem_chunk_alloc(my_vals);
2889 /*** add your code here to setup the new data structure ***/
2891 /* create the conversation with your data pointer */
2893 conversation = conversation_new(pinfo->fd->num, &pinfo->src, &pinfo->dst, pinfo->ptype,
2894 pinfo->srcport, pinfo->destport, 0);
2895 conversation_add_proto_data(conversation, my_proto, (void *)data_ptr);
2898 /* at this point the conversation data is ready */
2901 /******************* in the dissector init routine *******************/
2903 #define my_init_count 20
2906 my_dissector_init(void)
2909 /* destroy memory chunks if needed */
2912 g_mem_chunk_destroy(my_vals);
2914 /* now create memory chunks */
2916 my_vals = g_mem_chunk_new("my_proto_vals",
2918 my_init_count * sizeof(my_entry_t),
2922 /***************** in the protocol register routine *****************/
2924 /* register re-init routine */
2926 register_init_routine(&my_dissector_init);
2928 my_proto = proto_register_protocol("My Protocol", "My Protocol", "my_proto");
2931 2.2.9 An example conversation code that starts at a specific frame number.
2933 Sometimes a dissector has determined that a new conversation is needed that
2934 starts at a specific frame number, when a capture session encompasses multiple
2935 conversation that reuse the same src/dest ip/port pairs. You can use the
2936 conversation->setup_frame returned by find_conversation with
2937 pinfo->fd->num to determine whether or not there already exists a conversation
2938 that starts at the specific frame number.
2940 /* in the dissector routine */
2942 conversation = find_conversation(pinfo->fd->num, &pinfo->src, &pinfo->dst,
2943 pinfo->ptype, pinfo->srcport, pinfo->destport, 0);
2944 if (conversation == NULL || (conversation->setup_frame != pinfo->fd->num)) {
2945 /* It's not part of any conversation or the returned
2946 * conversation->setup_frame doesn't match the current frame
2949 conversation = conversation_new(pinfo->fd->num, &pinfo->src,
2950 &pinfo->dst, pinfo->ptype, pinfo->srcport, pinfo->destport,
2955 2.2.10 The example conversation code using conversation index field.
2957 Sometimes the conversation isn't enough to define a unique data storage
2958 value for the network traffic. For example if you are storing information
2959 about requests carried in a conversation, the request may have an
2960 identifier that is used to define the request. In this case the
2961 conversation and the identifier are required to find the data storage
2962 pointer. You can use the conversation data structure index value to
2963 uniquely define the conversation.
2965 See packet-afs.c for an example of how to use the conversation index. In
2966 this dissector multiple requests are sent in the same conversation. To store
2967 information for each request the dissector has an internal hash table based
2968 upon the conversation index and values inside the request packets.
2971 /* in the dissector routine */
2973 /* to find a request value, first lookup conversation to get index */
2974 /* then used the conversation index, and request data to find data */
2975 /* in the local hash table */
2977 conversation = find_conversation(pinfo->fd->num, &pinfo->src, &pinfo->dst,
2978 pinfo->ptype, pinfo->srcport, pinfo->destport, 0);
2979 if (conversation == NULL) {
2980 /* It's not part of any conversation - create a new one. */
2981 conversation = conversation_new(pinfo->fd->num, &pinfo->src,
2982 &pinfo->dst, pinfo->ptype, pinfo->srcport, pinfo->destport,
2986 request_key.conversation = conversation->index;
2987 request_key.service = pntohs(&rxh->serviceId);
2988 request_key.callnumber = pntohl(&rxh->callNumber);
2990 request_val = (struct afs_request_val *)g_hash_table_lookup(
2991 afs_request_hash, &request_key);
2993 /* only allocate a new hash element when it's a request */
2995 if (!request_val && !reply)
2997 new_request_key = g_mem_chunk_alloc(afs_request_keys);
2998 *new_request_key = request_key;
3000 request_val = g_mem_chunk_alloc(afs_request_vals);
3001 request_val -> opcode = pntohl(&afsh->opcode);
3002 opcode = request_val->opcode;
3004 g_hash_table_insert(afs_request_hash, new_request_key,
3010 2.3 Dynamic conversation dissector registration.
3013 NOTE: This sections assumes that all information is available to
3014 create a complete conversation, source port/address and
3015 destination port/address. If either the destination port or
3016 address is know, see section 2.4 Dynamic server port dissector
3019 For protocols that negotiate a secondary port connection, for example
3020 packet-msproxy.c, a conversation can install a dissector to handle
3021 the secondary protocol dissection. After the conversation is created
3022 for the negotiated ports use the conversation_set_dissector to define
3023 the dissection routine.
3024 Before we create these conversations or assign a dissector to them we should
3025 first check that the conversation does not already exist and if it exists
3026 whether it is registered to our protocol or not.
3027 We should do this because it is uncommon but it does happen that multiple
3028 different protocols can use the same socketpair during different stages of
3029 an application cycle. By keeping track of the frame number a conversation
3030 was started in wireshark can still tell these different protocols apart.
3032 The second argument to conversation_set_dissector is a dissector handle,
3033 which is created with a call to create_dissector_handle or
3036 create_dissector_handle takes as arguments a pointer to the dissector
3037 function and a protocol ID as returned by proto_register_protocol;
3038 register_dissector takes as arguments a string giving a name for the
3039 dissector, a pointer to the dissector function, and a protocol ID.
3041 The protocol ID is the ID for the protocol dissected by the function.
3042 The function will not be called if the protocol has been disabled by the
3043 user; instead, the data for the protocol will be dissected as raw data.
3047 /* the handle for the dynamic dissector *
3048 static dissector_handle_t sub_dissector_handle;
3050 /* prototype for the dynamic dissector */
3051 static void sub_dissector(tvbuff_t *tvb, packet_info *pinfo,
3054 /* in the main protocol dissector, where the next dissector is setup */
3056 /* if conversation has a data field, create it and load structure */
3058 /* First check if a conversation already exists for this
3061 conversation = find_conversation(pinfo->fd->num,
3062 &pinfo->src, &pinfo->dst, protocol,
3063 src_port, dst_port, new_conv_info, 0);
3065 /* If there is no such conversation, or if there is one but for
3066 someone else's protocol then we just create a new conversation
3067 and assign our protocol to it.
3069 if ( (conversation == NULL) ||
3070 (conversation->dissector_handle != sub_dissector_handle) ) {
3071 new_conv_info = g_mem_chunk_alloc(new_conv_vals);
3072 new_conv_info->data1 = value1;
3074 /* create the conversation for the dynamic port */
3075 conversation = conversation_new(pinfo->fd->num,
3076 &pinfo->src, &pinfo->dst, protocol,
3077 src_port, dst_port, new_conv_info, 0);
3079 /* set the dissector for the new conversation */
3080 conversation_set_dissector(conversation, sub_dissector_handle);
3085 proto_register_PROTOABBREV(void)
3089 sub_dissector_handle = create_dissector_handle(sub_dissector,
3095 2.4 Dynamic server port dissector registration.
3097 NOTE: While this example used both NO_ADDR2 and NO_PORT2 to create a
3098 conversation with only one port and address set, this isn't a
3099 requirement. Either the second port or the second address can be set
3100 when the conversation is created.
3102 For protocols that define a server address and port for a secondary
3103 protocol, a conversation can be used to link a protocol dissector to
3104 the server port and address. The key is to create the new
3105 conversation with the second address and port set to the "accept
3108 Some server applications can use the same port for different protocols during
3109 different stages of a transaction. For example it might initially use SNMP
3110 to perform some discovery and later switch to use TFTP using the same port.
3111 In order to handle this properly we must first check whether such a
3112 conversation already exists or not and if it exists we also check whether the
3113 registered dissector_handle for that conversation is "our" dissector or not.
3114 If not we create a new conversation on top of the previous one and set this new
3115 conversation to use our protocol.
3116 Since wireshark keeps track of the frame number where a conversation started
3117 wireshark will still be able to keep the packets apart even though they do use
3118 the same socketpair.
3119 (See packet-tftp.c and packet-snmp.c for examples of this)
3121 There are two support routines that will allow the second port and/or
3122 address to be set later.
3124 conversation_set_port2( conversation_t *conv, guint32 port);
3125 conversation_set_addr2( conversation_t *conv, address addr);
3127 These routines will change the second address or port for the
3128 conversation. So, the server port conversation will be converted into a
3129 more complete conversation definition. Don't use these routines if you
3130 want to create a conversation between the server and client and retain the
3131 server port definition, you must create a new conversation.
3136 /* the handle for the dynamic dissector *
3137 static dissector_handle_t sub_dissector_handle;
3141 /* in the main protocol dissector, where the next dissector is setup */
3143 /* if conversation has a data field, create it and load structure */
3145 new_conv_info = g_mem_chunk_alloc(new_conv_vals);
3146 new_conv_info->data1 = value1;
3148 /* create the conversation for the dynamic server address and port */
3149 /* NOTE: The second address and port values don't matter because the */
3150 /* NO_ADDR2 and NO_PORT2 options are set. */
3152 /* First check if a conversation already exists for this
3155 conversation = find_conversation(pinfo->fd->num,
3156 &server_src_addr, 0, protocol,
3157 server_src_port, 0, NO_ADDR2 | NO_PORT_B);
3158 /* If there is no such conversation, or if there is one but for
3159 someone else's protocol then we just create a new conversation
3160 and assign our protocol to it.
3162 if ( (conversation == NULL) ||
3163 (conversation->dissector_handle != sub_dissector_handle) ) {
3164 conversation = conversation_new(pinfo->fd->num,
3165 &server_src_addr, 0, protocol,
3166 server_src_port, 0, new_conv_info, NO_ADDR2 | NO_PORT2);
3168 /* set the dissector for the new conversation */
3169 conversation_set_dissector(conversation, sub_dissector_handle);
3172 2.5 Per-packet information.
3174 Information can be stored for each data packet that is processed by the
3175 dissector. The information is added with the p_add_proto_data function and
3176 retrieved with the p_get_proto_data function. The data pointers passed into
3177 the p_add_proto_data are not managed by the proto_data routines. If you use
3178 malloc or any other dynamic memory allocation scheme, you must release the
3179 data when it isn't required.
3182 p_add_proto_data(frame_data *fd, int proto, void *proto_data)
3184 p_get_proto_data(frame_data *fd, int proto)
3187 fd - The fd pointer in the pinfo structure, pinfo->fd
3188 proto - Protocol id returned by the proto_register_protocol call
3189 during initialization
3190 proto_data - pointer to the dissector data.
3193 2.6 User Preferences.
3195 If the dissector has user options, there is support for adding these preferences
3196 to a configuration dialog.
3198 You must register the module with the preferences routine with -
3200 module_t *prefs_register_protocol(proto_id, void (*apply_cb)(void))
3202 Where: proto_id - the value returned by "proto_register_protocol()" when
3203 the protocol was registered
3204 apply_cb - Callback routine that is call when preferences are applied
3207 Then you can register the fields that can be configured by the user with these
3210 /* Register a preference with an unsigned integral value. */
3211 void prefs_register_uint_preference(module_t *module, const char *name,
3212 const char *title, const char *description, guint base, guint *var);
3214 /* Register a preference with an Boolean value. */
3215 void prefs_register_bool_preference(module_t *module, const char *name,
3216 const char *title, const char *description, gboolean *var);
3218 /* Register a preference with an enumerated value. */
3219 void prefs_register_enum_preference(module_t *module, const char *name,
3220 const char *title, const char *description, gint *var,
3221 const enum_val_t *enumvals, gboolean radio_buttons)
3223 /* Register a preference with a character-string value. */
3224 void prefs_register_string_preference(module_t *module, const char *name,
3225 const char *title, const char *description, char **var)
3227 /* Register a preference with a range of unsigned integers (e.g.,
3230 void prefs_register_range_preference(module_t *module, const char *name,
3231 const char *title, const char *description, range_t *var,
3234 Where: module - Returned by the prefs_register_protocol routine
3235 name - This is appended to the name of the protocol, with a
3236 "." between them, to construct a name that identifies
3237 the field in the preference file; the name itself
3238 should not include the protocol name, as the name in
3239 the preference file will already have it
3240 title - Field title in the preferences dialog
3241 description - Comments added to the preference file above the
3243 var - pointer to the storage location that is updated when the
3244 field is changed in the preference dialog box
3245 base - Base that the unsigned integer is expected to be in,
3247 enumvals - an array of enum_val_t structures. This must be
3248 NULL-terminated; the members of that structure are:
3250 a short name, to be used with the "-o" flag - it
3251 should not contain spaces or upper-case letters,
3252 so that it's easier to put in a command line;
3254 a description, which is used in the GUI (and
3255 which, for compatibility reasons, is currently
3256 what's written to the preferences file) - it can
3257 contain spaces, capital letters, punctuation,
3260 the numerical value corresponding to that name
3262 radio_buttons - TRUE if the field is to be displayed in the
3263 preferences dialog as a set of radio buttons,
3264 FALSE if it is to be displayed as an option
3266 max_value - The maximum allowed value for a range (0 is the minimum).
3268 An example from packet-beep.c -
3270 proto_beep = proto_register_protocol("Blocks Extensible Exchange Protocol",
3275 /* Register our configuration options for BEEP, particularly our port */
3277 beep_module = prefs_register_protocol(proto_beep, proto_reg_handoff_beep);
3279 prefs_register_uint_preference(beep_module, "tcp.port", "BEEP TCP Port",
3280 "Set the port for BEEP messages (if other"
3281 " than the default of 10288)",
3282 10, &global_beep_tcp_port);
3284 prefs_register_bool_preference(beep_module, "strict_header_terminator",
3285 "BEEP Header Requires CRLF",
3286 "Specifies that BEEP requires CRLF as a "
3287 "terminator, and not just CR or LF",
3288 &global_beep_strict_term);
3290 This will create preferences "beep.tcp.port" and
3291 "beep.strict_header_terminator", the first of which is an unsigned
3292 integer and the second of which is a Boolean.
3294 Note that a warning will pop up if you've saved such preference to the
3295 preference file and you subsequently take the code out. The way to make
3296 a preference obsolete is to register it as such:
3298 /* Register a preference that used to be supported but no longer is. */
3299 void prefs_register_obsolete_preference(module_t *module,
3302 2.7 Reassembly/desegmentation for protocols running atop TCP.
3304 There are two main ways of reassembling a Protocol Data Unit (PDU) which
3305 spans across multiple TCP segments. The first approach is simpler, but
3306 assumes you are running atop of TCP when this occurs (but your dissector
3307 might run atop of UDP, too, for example), and that your PDUs consist of a
3308 fixed amount of data that includes enough information to determine the PDU
3309 length, possibly followed by additional data. The second method is more
3310 generic but requires more code and is less efficient.
3312 2.7.1 Using tcp_dissect_pdus().
3314 For the first method, you register two different dissection methods, one
3315 for the TCP case, and one for the other cases. It is a good idea to
3316 also have a dissect_PROTO_common function which will parse the generic
3317 content that you can find in all PDUs which is called from
3318 dissect_PROTO_tcp when the reassembly is complete and from
3319 dissect_PROTO_udp (or dissect_PROTO_other).
3321 To register the distinct dissector functions, consider the following
3322 example, stolen from packet-dns.c:
3324 dissector_handle_t dns_udp_handle;
3325 dissector_handle_t dns_tcp_handle;
3326 dissector_handle_t mdns_udp_handle;
3328 dns_udp_handle = create_dissector_handle(dissect_dns_udp,
3330 dns_tcp_handle = create_dissector_handle(dissect_dns_tcp,
3332 mdns_udp_handle = create_dissector_handle(dissect_mdns_udp,
3335 dissector_add("udp.port", UDP_PORT_DNS, dns_udp_handle);
3336 dissector_add("tcp.port", TCP_PORT_DNS, dns_tcp_handle);
3337 dissector_add("udp.port", UDP_PORT_MDNS, mdns_udp_handle);
3338 dissector_add("tcp.port", TCP_PORT_MDNS, dns_tcp_handle);
3340 The dissect_dns_udp function does very little work and calls
3341 dissect_dns_common, while dissect_dns_tcp calls tcp_dissect_pdus with a
3342 reference to a callback which will be called with reassembled data:
3345 dissect_dns_tcp(tvbuff_t *tvb, packet_info *pinfo, proto_tree *tree)
3347 tcp_dissect_pdus(tvb, pinfo, tree, dns_desegment, 2,
3348 get_dns_pdu_len, dissect_dns_tcp_pdu);
3351 (The dissect_dns_tcp_pdu function acts similarly to dissect_dns_udp.)
3352 The arguments to tcp_dissect_pdus are:
3354 the tvbuff pointer, packet_info pointer, and proto_tree pointer
3355 passed to the dissector;
3357 a gboolean flag indicating whether desegmentation is enabled for
3360 the number of bytes of PDU data required to determine the length
3363 a routine that takes as arguments a packet_info pointer, a tvbuff
3364 pointer and an offset value representing the offset into the tvbuff
3365 at which a PDU begins and should return - *without* throwing an
3366 exception (it is guaranteed that the number of bytes specified by the
3367 previous argument to tcp_dissect_pdus is available, but more data
3368 might not be available, so don't refer to any data past that) - the
3369 total length of the PDU, in bytes;
3371 a routine that's passed a tvbuff pointer, packet_info pointer,
3372 and proto_tree pointer, with the tvbuff containing a
3373 possibly-reassembled PDU, and that should dissect that PDU.
3375 2.7.2 Modifying the pinfo struct.
3377 The second reassembly mode is preferred when the dissector cannot determine
3378 how many bytes it will need to read in order to determine the size of a PDU.
3379 It may also be useful if your dissector needs to support reassembly from
3380 protocols other than TCP.
3382 Your dissect_PROTO will initially be passed a tvbuff containing the payload of
3383 the first packet. It should dissect as much data as it can, noting that it may
3384 contain more than one complete PDU. If the end of the provided tvbuff coincides
3385 with the end of a PDU then all is well and your dissector can just return as
3386 normal. (If it is a new-style dissector, it should return the number of bytes
3387 successfully processed.)
3389 If the dissector discovers that the end of the tvbuff does /not/ coincide with
3390 the end of a PDU, (ie, there is half of a PDU at the end of the tvbuff), it can
3391 indicate this to the parent dissector, by updating the pinfo struct. The
3392 desegment_offset field is the offset in the tvbuff at which the dissector will
3393 continue processing when next called. The desegment_len field should contain
3394 the estimated number of additional bytes required for completing the PDU. Next
3395 time your dissect_PROTO is called, it will be passed a tvbuff composed of the
3396 end of the data from the previous tvbuff together with desegment_len more bytes.
3398 If the dissector cannot tell how many more bytes it will need, it should set
3399 desegment_len=DESEGMENT_ONE_MORE_SEGMENT; it will then be called again as soon
3400 as any more data becomes available. Dissectors should set the desegment_len to a
3401 reasonable value when possible rather than always setting
3402 DESEGMENT_ONE_MORE_SEGMENT as it will generally be more efficient. Also, you
3403 *must not* set desegment_len=1 in this case, in the hope that you can change
3404 your mind later: once you return a positive value from desegment_len, your PDU
3405 boundary is set in stone.
3407 static hf_register_info hf[] = {
3409 {"C String", "c.string", FT_STRING, BASE_NONE, NULL, 0x0,
3415 * Dissect a buffer containing C strings.
3417 * @param tvb The buffer to dissect.
3418 * @param pinfo Packet Info.
3419 * @param tree The protocol tree.
3421 static void dissect_cstr(tvbuff_t * tvb, packet_info * pinfo, proto_tree * tree)
3424 while(offset < tvb_reported_length(tvb)) {
3425 gint available = tvb_reported_length_remaining(tvb, offset);
3426 gint len = tvb_strnlen(tvb, offset, available);
3429 /* we ran out of data: ask for more */
3430 pinfo->desegment_offset = offset;
3431 pinfo->desegment_len = DESEGMENT_ONE_MORE_SEGMENT;
3435 if (check_col(pinfo->cinfo, COL_INFO)) {
3436 col_set_str(pinfo->cinfo, COL_INFO, "C String");
3439 len += 1; /* Add one for the '\0' */
3442 proto_tree_add_item(tree, hf_cstring, tvb, offset, len, FALSE);
3444 offset += (guint)len;
3447 /* if we get here, then the end of the tvb coincided with the end of a
3448 string. Happy days. */
3451 This simple dissector will repeatedly return DESEGMENT_ONE_MORE_SEGMENT
3452 requesting more data until the tvbuff contains a complete C string. The C string
3453 will then be added to the protocol tree. Note that there may be more
3454 than one complete C string in the tvbuff, so the dissection is done in a
3459 The ptvcursor API allows a simpler approach to writing dissectors for
3460 simple protocols. The ptvcursor API works best for protocols whose fields
3461 are static and whose format does not depend on the value of other fields.
3462 However, even if only a portion of your protocol is statically defined,
3463 then that portion could make use of ptvcursors.
3465 The ptvcursor API lets you extract data from a tvbuff, and add it to a
3466 protocol tree in one step. It also keeps track of the position in the
3467 tvbuff so that you can extract data again without having to compute any
3468 offsets --- hence the "cursor" name of the API.
3470 The three steps for a simple protocol are:
3471 1. Create a new ptvcursor with ptvcursor_new()
3472 2. Add fields with multiple calls of ptvcursor_add()
3473 3. Delete the ptvcursor with ptvcursor_free()
3475 ptvcursor offers the possibility to add subtrees in the tree as well. It can be
3476 done in very simple steps :
3477 1. Create a new subtree with ptvcursor_push_subtree(). The old subtree is
3478 pushed in a stack and the new subtree will be used by ptvcursor.
3479 2. Add fields with multiple calls of ptvcursor_add(). The fields will be
3480 added in the new subtree created at the previous step.
3481 3. Pop the previous subtree with ptvcursor_pop_subtree(). The previous
3482 subtree is again used by ptvcursor.
3483 Note that at the end of the parsing of a packet you must have popped each
3484 subtree you pushed. If it's not the case, the dissector will generate an error.
3486 To use the ptvcursor API, include the "ptvcursor.h" file. The PGM dissector
3487 is an example of how to use it. You don't need to look at it as a guide;
3488 instead, the API description here should be good enough.
3490 2.8.1 ptvcursor API.
3493 ptvcursor_new(proto_tree* tree, tvbuff_t* tvb, gint offset)
3494 This creates a new ptvcursor_t object for iterating over a tvbuff.
3495 You must call this and use this ptvcursor_t object so you can use the
3499 ptvcursor_add(ptvcursor_t* ptvc, int hf, gint length, gboolean endianness)
3500 This will extract 'length' bytes from the tvbuff and place it in
3501 the proto_tree as field 'hf', which is a registered header_field. The
3502 pointer to the proto_item that is created is passed back to you. Internally,
3503 the ptvcursor advances its cursor so the next call to ptvcursor_add
3504 starts where this call finished. The 'endianness' parameter matters for
3505 FT_UINT* and FT_INT* fields.
3508 ptvcursor_add_no_advance(ptvcursor_t* ptvc, int hf, gint length, gboolean endianness)
3509 Like ptvcursor_add, but does not advance the internal cursor.
3512 ptvcursor_advance(ptvcursor_t* ptvc, gint length)
3513 Advances the internal cursor without adding anything to the proto_tree.
3516 ptvcursor_free(ptvcursor_t* ptvc)
3517 Frees the memory associated with the ptvcursor. You must call this
3518 after your dissection with the ptvcursor API is completed.
3522 ptvcursor_push_subtree(ptvcursor_t* ptvc, proto_item* it, gint ett_subtree)
3523 Pushes the current subtree in the tree stack of the cursor, creates a new
3524 one and sets this one as the working tree.
3527 ptvcursor_pop_subtree(ptvcursor_t* ptvc);
3528 Pops a subtree in the tree stack of the cursor
3531 ptvcursor_add_with_subtree(ptvcursor_t* ptvc, int hfindex, gint length,
3532 gboolean little_endian, gint ett_subtree);
3533 Adds an item to the tree and creates a subtree.
3534 If the length is unknown, length may be defined as SUBTREE_UNDEFINED_LENGTH.
3535 In this case, at the next pop, the item length will be equal to the advancement
3536 of the cursor since the creation of the subtree.
3539 ptvcursor_add_text_with_subtree(ptvcursor_t* ptvc, gint length,
3540 gint ett_subtree, const char* format, ...);
3541 Add a text node to the tree and create a subtree.
3542 If the length is unknown, length may be defined as SUBTREE_UNDEFINED_LENGTH.
3543 In this case, at the next pop, the item length will be equal to the advancement
3544 of the cursor since the creation of the subtree.
3546 2.8.2 Miscellaneous functions.
3549 ptvcursor_tvbuff(ptvcursor_t* ptvc)
3550 Returns the tvbuff associated with the ptvcursor.
3553 ptvcursor_current_offset(ptvcursor_t* ptvc)
3554 Returns the current offset.
3557 ptvcursor_tree(ptvcursor_t* ptvc)
3558 Returns the proto_tree associated with the ptvcursor.
3561 ptvcursor_set_tree(ptvcursor_t* ptvc, proto_tree *tree)
3562 Sets a new proto_tree for the ptvcursor.
3565 ptvcursor_set_subtree(ptvcursor_t* ptvc, proto_item* it, gint ett_subtree);
3566 Creates a subtree and adds it to the cursor as the working tree but does
3567 not save the old working tree.