The proto tree is needed in several cases when using Lua field extractors,
because they fetch values from the tree. Without a valid field extractor
a Lua plugin may misbehave and display wrong column info.
This fixes column issues when:
- Calling resetColumns() in Qt. This involves adding a display filter,
change time display format, change name resolution and other changes
in UI which requires column updates.
- Print summary lines.
- Export as CSV and PSML.
Change-Id: Ieed6f8578cdf2759f1f836cd8413a4529b7bbd80
Reviewed-on: https://code.wireshark.org/review/13708
Petri-Dish: Stig Bjørlykke <stig@bjorlykke.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
GNU coding standards recommend against it and automake is designed
around it.
This allows overriding the global build flags using AM_CFLAGS, etc.,
or per object flags, something that is difficult or impossible currently
because of automake precedence rules.
Change-Id: I3f1ea12e560af5a46b2f04b342b1882bbf123f12
Reviewed-on: https://code.wireshark.org/review/13455
Petri-Dish: João Valverde <j@v6e.pt>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: João Valverde <j@v6e.pt>
Delay allocation of a GByteArray to avoid a memleak on errors.
Change-Id: I2b2ae6d33407500e81a4f45b86ef82720d3443a4
Reviewed-on: https://code.wireshark.org/review/13553
Petri-Dish: Peter Wu <peter@lekensteyn.nl>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
That removes most of the uses of the frame number field in the
frame_data structure.
Change-Id: Ie22e4533e87f8360d7c0a61ca6ffb796cc233f22
Reviewed-on: https://code.wireshark.org/review/13509
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Add fields for the absolute time stamp (and another field for a presence
flag for the absolute time stamp) and the packet encapsulation for the
packet.
This lets us remove the field for the packet encapsulation in the
frame_data structure; do so.
Change-Id: Ifb910a9a192414e2a53086f3f7b97f39ed36aa39
Reviewed-on: https://code.wireshark.org/review/13499
Reviewed-by: Guy Harris <guy@alum.mit.edu>
This makes it possible to trigger reloading Lua plugins from
within a Lua plugin. This can be used when having a plugin to
update local plugins from a external source.
Renamed reload() to reload_packets() to clarify what's reloaded,
and added a alias (marked as obsoleted) from reload().
Change-Id: I4e529992af5f651613950329e73718dbda317d2e
Reviewed-on: https://code.wireshark.org/review/13024
Petri-Dish: Stig Bjørlykke <stig@bjorlykke.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
proto_item_get_len() is *not* guaranteed to return a correct value.
Even if there's a non-null tree item, it might be pointing to a "faked"
item; it really shouldn't be used.
So add proto_tree_add_item_ret_length() and
proto_tree_add_item_new_ret_length(), which calculate the real length
themselves and return it through a pointer.
Fix as many places as we straightforwardly can to use them rather than
to use proto_item_get_len(). (There's a Lua API for
proto_item_get_len(), so we keep it around, but we should add Lua APIs
for the new routines, and deprecate the old API.)
Fix ptvcursor_add() to do the same thing that
proto_tree_add_item_ret_length() and
proto_tree_add_item_new_ret_length() do.
Split the TRY_TO_FAKE_THIS_ITEM macros into a macro to check for the
tree being null and to try to fake the item. We don't always use the
former macro, as we might need to do more than just return NULL if the
incoming tree is null (for example, calculating the item's real length
and using it...).
new_field_info() never returns NULL; remove checks for it. The check
for a null tree is done before the calls to new_field_info().
Change-Id: I002a218d1f810c73e0de837e0ac6ebcde21bacec
Reviewed-on: https://code.wireshark.org/review/13139
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Free the Pref enum values in Pref__gc().
Change-Id: I4d66dbe7ee4879f3b14094135887d78cba876ea3
Reviewed-on: https://code.wireshark.org/review/12628
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Store the Pref default string value and ensure this is freed both
when registering the pref and when not.
Use g_malloc0 to allocate Pref and avoid several init's.
Change-Id: I5f97a15d06068d7805f02f7c7feea61f9b2030f5
Reviewed-on: https://code.wireshark.org/review/12626
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Reverted some unneeded changes from 7f074364 because it's currently
only used for FT_ABSOLUTE_TIME.
Also fixed some small indent issues.
Change-Id: I923e35bf8ad4e991518bee973b87d4be06137463
Reviewed-on: https://code.wireshark.org/review/12539
Reviewed-by: Anders Broman <a.broman58@gmail.com>
The preference has already been deregistered at that point, so the value
of the preference has been freed and the pointer to it has been set to
null, so it's already been freed and its pointer no longer points to it
and the free from Lua will do nothing.
Change-Id: I11bf74932303151483cd3699659f67d64b466759
Reviewed-on: https://code.wireshark.org/review/12606
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Capture dissectors could be architected like dissection dissectors, with tables and subtables and possibly using tvbs to pass there data instead of raw byte arrays. This is a first step towards that by refactoring capture_info_packet() to work off of a "capture dissector table"
Registering the capture dissection functions instead of calling them directly also clears up a bunch of dissector header files who sole purpose was providing the capture dissection function definition.
Change-Id: I10e9b79e061f32d2572f009823601d4f048d37aa
Reviewed-on: https://code.wireshark.org/review/12581
Petri-Dish: Michael Mann <mmann78@netscape.net>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Michael Mann <mmann78@netscape.net>
The check in tmp_fld_check_assert() does terminate with g_error
if given an empty string as name, so we have to check for this.
Change-Id: I084e3e715bd319484a52f60ef90c1a2aea30df1b
Reviewed-on: https://code.wireshark.org/review/12534
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
Support reloading Lua plugins with expert infos.
Use the same delayed deregister logic as for fields.
Change-Id: I36efa0820050b3a7afed4de7a8b0fa16805e8dfa
Reviewed-on: https://code.wireshark.org/review/12498
Petri-Dish: Stig Bjørlykke <stig@bjorlykke.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Mostly plugins, but also LUA and generated skinny dissector.
Change-Id: Ifeb4205442f9a60875266b4e82841ff38b4fdb63
Reviewed-on: https://code.wireshark.org/review/12515
Reviewed-by: Michael Mann <mmann78@netscape.net>
URL to documentation with https as suggested by Alexis.
(s. commit ecc9c74326)
Change-Id: I64d82b0c07cce5658eb8f4ae758f1a42946b837a
Reviewed-on: https://code.wireshark.org/review/12181
Reviewed-by: Michael Mann <mmann78@netscape.net>
The Lua reference has been moved from User's Guide to Developer's Guide.
Change-Id: I3489d774e54310ce49997e33d5318adf5e0bb2bc
Reviewed-on: https://code.wireshark.org/review/12128
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
Check if description (protocol name) and short_name are used before
registering the protocol. This because proto_register_protocol() makes
sure there's not already a protocol with any of the names registered
and duplicates will be reported with a g_error() which terminates the
Wireshark unexpectedly.
Also check if short_name contains valid characters.
Give appropriate error messages.
Bug: 11739
Change-Id: Ib9776a2a3406ae5278ce744defd61864ebed0282
Reviewed-on: https://code.wireshark.org/review/11995
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
Listeners should not be deregistered in __gc because they will go out
of scope while in use. Instead free allocated data when deregistering
the Listener (Listener.remove() and Reload Lua Plugins).
Bug: 11722
Change-Id: Iadf6506757df06e476ac3cac38c05f1d1d497dc4
Reviewed-on: https://code.wireshark.org/review/11924
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
If we ever change the way file writers work, in a fashion incompatible
with the existing way they work, we'll also rename this member - and get
rid of checks for earlier versions of the Lua interface.
Change-Id: I64065944fa31371f5249cafd930c18f180ad7299
Reviewed-on: https://code.wireshark.org/review/11879
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Added a check for nil valuestring for all ProtoField integer types
to avoid lookup when argument is not set.
Change-Id: Ib4c016b69ee77dbea4bb83ac93c0d9ae9f48f236
Reviewed-on: https://code.wireshark.org/review/11845
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
A ProtField type FRAMENUM cannot fetch value from a Tvb.
Change-Id: Iff0f6df8b00445855c9030dcfa753daa62262171
Reviewed-on: https://code.wireshark.org/review/11832
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
It doesn't actually *close* any handle, so it's best called a "finish"
routine rather than a "close" routine.
In libwiretap modules, don't bother setting the finish routine pointer
to null - it's already initialized to null (it's probably best not to
require modules to set it).
Change-Id: I19554f3fb826db495f17b36600ae36222cbc21b0
Reviewed-on: https://code.wireshark.org/review/11659
Reviewed-by: Guy Harris <guy@alum.mit.edu>
The target here is the Decode As dialog where protocols have multiple registrations into a dissector table and that shows up as multiple entries in the Decode As dialog list with the same name so users are unsure which "dissector" they are choosing.
The "default" behavior (done in this commit) is to not allow duplicates for a dissector table, whether its part of Decode As or not. It's just ENFORCED for Decode As.
Bug: 3949
Change-Id: Ibe14fa61aaeca0881f9cc39b78799e314b5e8127
Reviewed-on: https://code.wireshark.org/review/11405
Petri-Dish: Michael Mann <mmann78@netscape.net>
Reviewed-by: Michael Mann <mmann78@netscape.net>
Move the definitions of hashipv4_t and hashipv6_t to wiretap/wtap.h, as
that's the main place they're used. Change them a bit not to depend on
other stuff from libwireshark, and change the code as required by those
changes.
This should fix the Solaris build; apparently, the Sun^WOracle compiler
is generating code for static inline functions even if they're never
called, so that libwiretap ends up including code that calls tvbuff and
wmem functions.
There's probably further cleanup that could be done here, but this
should at least fix the build, as well as getting rid of a dependency
between two libraries that are at least somewhat independent (libwiretap
should *not* depend on libwireshark, as some programs use libwiretap but
not libwireshark, and, ultimately, we probably want it to be possible to
use libwireshark without libwiretap but that'd be more work).
Change-Id: I91c745282f17d7c8bff7809aa277eab2b3cf47c1
Reviewed-on: https://code.wireshark.org/review/11537
Petri-Dish: Guy Harris <guy@alum.mit.edu>
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Replace CMP_ADDRESS, COPY_ADDRESS, et al with their lower-case
equivalents in the asn1 and epan directories.
Change-Id: I4043b0931d4353d60cffbd829e30269eb8d08cf4
Reviewed-on: https://code.wireshark.org/review/11200
Petri-Dish: Michal Labedzki <michal.labedzki@tieto.com>
Petri-Dish: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Michael Mann <mmann78@netscape.net>
Would leak some bytes after startup.
Caught by LeakSanitizer.
Change-Id: I4644f204343ce5a803a7dfdedac6a1960882807f
Reviewed-on: https://code.wireshark.org/review/10859
Petri-Dish: Peter Wu <peter@lekensteyn.nl>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
The dissect_tcp_pdus function in LUA is passed two LUA functions that
get the PDU length and the dissect a PDU. When one of these functions
fail, a longjmp is made to the the caller of lua_pcall.
This is no problem for the PDU length function, but the PDU dissect
function is wrapped in a TRY/CATCH/ENDTRY block which also uses longjmp
and need to be fully executed. Without doing so, LUA exceptions will
crash on a weird location (except_pop).
Fix the crash by not using luaL_error, but throw dissector errors which
properly breaks out of the tcp_dissect_pdus C function and then convert
it to a LUA error such that the dissector can handle it.
Test with `tshark -X lua_script:crash.lua -r ssl.pcap`:
trivial_proto = Proto("trivial", "Trivial Protocol")
function dissect_foo(tvb, pinfo, tree)
error("triggering a LUA error");
end
function get_pdu_len(tvb, pinfo, tree) return 5; end
function trivial_proto.dissector(tvb, pinfo, tree)
dissect_tcp_pdus(tvb, tree, 5, get_pdu_len, dissect_foo)
end
tcp_table = DissectorTable.get("tcp.port")
tcp_table:add(443, trivial_proto)
It should not crash and will print this:
Lua Error: dissect_tcp_pdus dissect_func: [string "crash.lua"]:3: triggering a LUA error
Change-Id: Ibd079cc5eb3a2e4d2e62ea49a512fa2cc8e561ea
Reviewed-on: https://code.wireshark.org/review/10685
Petri-Dish: Peter Wu <peter@lekensteyn.nl>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: Evan Huus <eapache@gmail.com>
Fixes this failure:
tshark: lapi.c:175: lua_settop: Assertion `(-(idx+1) <= (L->top - (func + 1))) && "invalid new top"' failed.
Found using LUA 5.2 compiled with -DLUA_USE_APICHECK. Note that the
value returned by luaopen_bit does not affect the size of the stack
after using lua_call; if there are excess return values, then these will
be dropped after calling lua_call. See http://www.lua.org/pil/26.2.html
I also verified that the stack is still empty after calling all
wslua_reg_module functions.
Change-Id: Ia72635277ca4d8077e47a5c472391a78fe6d311e
Reviewed-on: https://code.wireshark.org/review/10664
Petri-Dish: Peter Wu <peter@lekensteyn.nl>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Add the "Internals" menu under the View menu instead of at the top level
for now at least. Add the Dissector Tables dialog there.
Change-Id: Ieb23b0015591bac196e4ef94e3443832288333f9
Reviewed-on: https://code.wireshark.org/review/10654
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Get rid of false-positive coverity issues by using a macro instead
of source code if-statement.
Bug: 11501
Change-Id: I07f478ed334931f05bdfb87cb8f614b16bbf8fa8
Reviewed-on: https://code.wireshark.org/review/10376
Petri-Dish: Hadriel Kaplan <hadrielk@yahoo.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Hadriel Kaplan <hadrielk@yahoo.com>
Wslua's Int64.fromhex() and UInt64.fromhex() need to check the sscanf return
value. Found by coverity (CID 1191368 &1191369).
Change-Id: I67fba027e18341d429787515f94c794573dc41c2
Reviewed-on: https://code.wireshark.org/review/10183
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Instead of splitting the stats into two lists as with the GTK+ UI, add
everything to an expandable tree. This allows viewing nodes on more than
one network.
Rename the top-level Bluetooth menu item to Wireless and put the WLAN
stats dialog there.
The Qt UI matches SSIDs (WlanNetworkTreeWidgetItem::isMatch) a bit
differently than the GTK+ UI. Try to make the logic as plain as possible
since we'll likely have to update it in the future.
The addition of a custom BSSID address types means that we can't assume
that everything is AT_ETHER. Add routines for checking for broadcast
BSSIDs and comparing only the data portions of addresses.
Move PercentBarDelegate into its own module. Use it in
WlanStatisticsDialog.
Change-Id: Ie4214eb00671a890871380c4a07213ebfb7585c6
Reviewed-on: https://code.wireshark.org/review/10171
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Stig Bjørlykke