Fix Dereference of null pointer found by Clang analyzer

Also add modeline info and use _U_ only when it is needed (for pinfo)

Change-Id: I71349d6451e1d8304fad04781e5ee226a1a4c729
Reviewed-on: https://code.wireshark.org/review/1466
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Add support for Logical Link Down TLV.
TLV is 1 octet long, and only MSB holds info with 7 unused bits

Change-Id: Ic1f5ce1b774bb7439cd2a23bc9bfb44f7dc25888
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1450
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

MBIM: dissect MBIM_SET_STK_PAC and MBIM_STK_PAC_INFO messages

Change-Id: Ib58dece7ce7278e292d88f68368c238695c32394
Reviewed-on: https://code.wireshark.org/review/1463
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Remove 'include win32.mak' from certain nmake files when not needed.

Change-Id: I05a505d884abb5fd7fca87bb89af29e3e4502b74
Reviewed-on: https://code.wireshark.org/review/1460
Reviewed-by: Bill Meier <wmeier@newsguy.com>

Add a comment re certain variables beibg defined in win32.mak.

Change-Id: I9246a1786a5bf8c48744378f25a66f6881caed81
Reviewed-on: https://code.wireshark.org/review/1459
Reviewed-by: Bill Meier <wmeier@newsguy.com>

Add a comment re VS2013EE requiring that win32.mak needs to be separately obtained.

Change-Id: I53413f7bdde50149fd923bbe35c995145d5139f2
Reviewed-on: https://code.wireshark.org/review/1458
Reviewed-by: Bill Meier <wmeier@newsguy.com>

Cleanup SPDY dissector.

Address Evan's missed comments from https://code.wireshark.org/review/767/

From Alexis La Goutte
Fix unused variable (pinfo) and fix crash (wrong variable in while)

Change-Id: I1b134585f4d522d3e808069f606beb4aae0488fa
Reviewed-on: https://code.wireshark.org/review/1164
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Evan Huus <eapache@gmail.com>

Fix -Wunreachable-code found by Clang

Change-Id: I895194655ff674fff5e84fa93b3f9a00a0f70e27
Reviewed-on: https://code.wireshark.org/review/1457
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Add OSPF Authentication Trailer (RFC6506)

Change-Id: Ie5c536a6977e35acc12222af3c48d29740cd57f2
Closed-bug:9941
Reviewed-on: https://code.wireshark.org/review/1440
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Michael Mann <mmann78@netscape.net>

Unreachable code in ui/gtk/main.c

Unreachable code due to conflicting defines: checking for GTKOSXAPPLICATION inside _WIN32

Found by Edwin

Closed-bug: 10065
Change-Id: I074fa14069e859b3fc4b7f50c443a67e7f0e954b
Reviewed-on: https://code.wireshark.org/review/1456
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix Dereference of null pointer found by Clang analyzer

packet-teamspeak2.c:456:21: warning: Access to field 'fragmented' results in a dereference of a null pointer (loaded from variable 'frag')

Change-Id: Ie253991f8d664c4594254bcd4569883614aafe2f
Reviewed-on: https://code.wireshark.org/review/1455
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix Dead Store (Dead assignement/Dead increment) warning found by Clang

Change-Id: I8ab25b1a3326ade46408ebd2e14cf841191c57d9
Reviewed-on: https://code.wireshark.org/review/1454
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix Dead Store (Dead assignement/Dead increment) warning found by Clang

Change-Id: Id38f366c04edcf3326b323fc851a6644a942d79d
Reviewed-on: https://code.wireshark.org/review/1453
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix procedure of finding packages

Before this change the code using make package was just showing
a string ${PACKAGE_VAR}_FOUND instead of showing if the
${PACKAGE_VAR}_FOUND variable was set true by find_package function.
Now the user will be informed about which packages were not found
and an error will be raised, as packages sought are the ones that
are selected for the build.

This change also ads fixes to some improper PACKAGE_VAR
variables that prevented find_package from finding
proper packages.

Change-Id: Ic20bc50ace65bab385059a7430909f95367520a3
Reviewed-on: https://code.wireshark.org/review/1410
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

gluster: add support for GlusterD v3 procedures

The GlusterD Management protocol has been updated to version 3. All procedures are implemented with this change.

This change also takes care of converting values passed in a dictionary
to real and displayable GUIDs.

Change-Id: I93808eda393745aa9ab3b986a6b7997f6f80df50
Reported-by: Vikhyat Umrao <vumrao@redhat.com>
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: https://code.wireshark.org/review/1411
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

AT: do not always prepend "/" to protocol name

Change-Id: Ice043e0355ea4d9112b377398079c94e66e35cc8
Reviewed-on: https://code.wireshark.org/review/1452
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Try to fix warning C4244: '-=' : conversion from '__int64' to 'int',
possible loss of data

Change-Id: I8cf74090e507f21eebfb40d72a3630d9f9bb0390
Reviewed-on: https://code.wireshark.org/review/1451
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fixed typo in docbook (usuall to usually)

Change-Id: I1aad1d6487cb6acd52bb851c08d9c4965bbb9f55
Reviewed-on: https://code.wireshark.org/review/1449
Reviewed-by: Evan Huus <eapache@gmail.com>

Document to_str-int.h functions.

Change argument name of hex_to_str_back pad->len it seems to sounds better.
Make uint_to_str_back_len() value uint32, to be sure about required buffer size.

Change-Id: I48fd560683c5c0845cbb60813887a18328ec01e6
Reviewed-on: https://code.wireshark.org/review/1448
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Optimize display_signed_time() - remove g_snprintf

When tshark is printing columns (default) display_signed_time() is called for every frame.
Current implemention is using g_snprintf() which makes this function costly when there is lot frames.

Change-Id: I109c8699d38bfbd05475d457ae4173e937c6812d
Reviewed-on: https://code.wireshark.org/review/1447
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix uninitialized struct in SMB2

Bug:10056
Change-Id: I6c7d3900208ee058ca9505e8d9f673ce4f4801cb
Reviewed-on: https://code.wireshark.org/review/1446
Reviewed-by: Evan Huus <eapache@gmail.com>

Add '#include "packet-ipsec.h'; Do some cleanup.

Cleanup:
- #include <stdio.h> & <ctype.h> only when needed;
- Add editor modelines;
- Convert tabs to spaces;
- Use a consistent indentation (2 spaces).

Change-Id: I5ef67b5c6aff99bf5dc1655b20760f862cd14d7f
Reviewed-on: https://code.wireshark.org/review/1445
Reviewed-by: Bill Meier <wmeier@newsguy.com>

make-sminmpec.pl fixups.

Enforce a minimum number of entries and do our work in the epan
directory.

Change-Id: I69cc6ae3255b23706a2e67db890a9718e10568b2
Reviewed-on: https://code.wireshark.org/review/1398
Reviewed-by: Gerald Combs <gerald@wireshark.org>

IPsec: fix compilation when building without libgcrypt

Change-Id: If525fb27ec607201efceb3ab9207305c3a4953dc
Reviewed-on: https://code.wireshark.org/review/1439
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Cleanup libz streams in SPDY dissector

Fixes major memory leak.

Bug:10062
Change-Id: Ife70f12b8697a7873860c702ee0c740f98d98821
Reviewed-on: https://code.wireshark.org/review/1435
Reviewed-by: Evan Huus <eapache@gmail.com>

Fix indent (Use Tab)

Change-Id: I811fb3db41a31d2334eb80f1821beea109bfa0d9
Reviewed-on: https://code.wireshark.org/review/1437
Reviewed-by: Anders Broman <a.broman58@gmail.com>

MBIM: use hash map instead of red/black tree for UUID extension table

Change-Id: Iaab9ff66e94bfe38fc875ad312b8e795b80814f4
Reviewed-on: https://code.wireshark.org/review/1438
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Try to make Mac OS X Buildbot Happy

packet-ipsec.c:1905: warning: implicit conversion shortens 64-bit value into a 32-bit value

Change-Id: Ic293e6b297f8f52aef7a102ab5ca0e0e98b9d7a7
Reviewed-on: https://code.wireshark.org/review/1436
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Add an API for programmatically adding ESP SAs (most likely from a private
dissector).

Also, speed up ESP decryption in several ways:
- store gcrypt_cipher_hd in the SA struct, rather than continually
open, setkey and close for each PDU
- don't convert the key string from ascii to hex each time - do it upon
during update callback and keep
- do the decryption in-place, avoiding the need to allocate, memcpy and
free a separate buffer for encrypted data
- when matching addresses, avoid doing a strlen until after we check
whether or not we're matching against "*"

Change-Id: Ibb66663ec21326351abc761be78a0d57f7797fa5
Reviewed-on: https://code.wireshark.org/review/1421
Reviewed-by: Anders Broman <a.broman58@gmail.com>

SVN --> GIT in a few places. Fixes Bug #10060.

Change-Id: I5dcf7699e295dcdbd1e870af5ec68b010a63bf6f
Reviewed-on: https://code.wireshark.org/review/1434
Reviewed-by: Bill Meier <wmeier@newsguy.com>

Fix cast for older versions of GLib.

The version of GLib on the 32-bit OS X builder defines

Unfortunately this results in a cast warning when u is a guint16. Newer
versions of GLib cast to gulong so do that.

Change-Id: I2d7e56ef9e5164789ee3a59115cb0f4a76155880
Reviewed-on: https://code.wireshark.org/review/1433
Reviewed-by: Gerald Combs <gerald@wireshark.org>

Fix crash when switching profiles

rescan_packets() starts a new dissection session by free'ing the
previous state and creating a new one. A reference to the old session
was not updated though.

This is the only place where epan is overwritten. After init, col_init
will at some point initialize cf->cinfo.epan with a cf->epan.

Bug: 9463
Change-Id: I8cf4e0823328f44a0dccfeeec1594f9929829d2d
Reviewed-on: https://code.wireshark.org/review/1430
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Fix an extraneous parentheses warning.

"if (G_UNLIKELY(...))" triggers an extraneous parentheses warning when
compiling with XCode's clang-500.2.75. From looking at the macro
definition we *should* be able to get rid of the outer parentheses
everywhere.

Change-Id: I710e1cc391e1167c1243c4ddb032f2831f0a9498
Reviewed-on: https://code.wireshark.org/review/1432
Reviewed-by: Gerald Combs <gerald@wireshark.org>

Add casts to satisfy buildbot

I'm really not sure why it's complaining about a cast *up* from a guint16 to a
guint...

Change-Id: Ie24ae0ea2361b54ed255773a0fc86250a240b55a
Reviewed-on: https://code.wireshark.org/review/1431
Reviewed-by: Evan Huus <eapache@gmail.com>

Make id a guint to try to please OSX buildbot.

Change-Id: Ia97aa45ff0fc52fb21d2e864d1ca328646de547b
Reviewed-on: https://code.wireshark.org/review/1429
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Convert a few dissectors with simple request/response tracking from red/black tree to hash map
Update the readme file accordingly

Change-Id: I056d1ab1f77df641b83fa9b3618b6c25d66e1a83
Reviewed-on: https://code.wireshark.org/review/1420
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Add support for Congestion Notification TLVs (802.1Qau)

I don't have sample captures for this, but the TLV layout is pretty simple and only contains two 1 octet bitmaps

Change-Id: Ibd7db7f40187904bbbf348c2cb89a5494cd46aab
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1427
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Set CMake defaults for Windows build to:
  Use GTK2 as per nmake.
  Not to build the dumpabi executables.

Change-Id: Ia28cf1a80b81e10595f4af5bbd46da3b3675faf5
Reviewed-on: https://code.wireshark.org/review/1376
Reviewed-by: Graham Bloice <graham.bloice@trihedral.com>
Tested-by: Graham Bloice <graham.bloice@trihedral.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Updated items in dissect_ospf_v2_lsa to use filterable item fields

Change-Id: I226c6fd3d5aef9d3881416fba9c6259e6170357c
Reviewed-on: https://code.wireshark.org/review/1426
Reviewed-by: Anders Broman <a.broman58@gmail.com>

prefs: prevent double-free on changing prefs

Prevent double-free when changing profiles. After switching from the
default profile to a profile without a preferences file,
saved_at_version will be freed, but a new version will not be inserted
(because the prefs file is empty). On a subsequent switch to another
profile, the saved_at_version field is freed again leading to a crash.

Bug: 9463
Change-Id: Ia9e0936e32603b4676a37dc428bda40c264a6eaa
Reviewed-on: https://code.wireshark.org/review/1425
Reviewed-by: Anders Broman <a.broman58@gmail.com>

prefs: spelling fix, simplify code

Found spelling error "a another". All those dynamic allocations and
magic numbers are horrible and unnecessary. Simplify the gtk code
rewording the message and merging strings.

Simplify the version code by using fscanf to take care of matching
the first line of a preference file.

Change-Id: I1e75803aacaa494ba5005791bcbd023e0807aaaa
Reviewed-on: https://code.wireshark.org/review/1424
Reviewed-by: Anders Broman <a.broman58@gmail.com>

gtk: fix glib assert for dialogs without cancel

Dialogs with only "Save" and "Don't save" buttons do not have a default
cancel action. Check for the existence of an OK button before
registering one.

This fixes three warnings when trying to trying to overwrite preferences
with unknown keys.

Change-Id: Ie583c40d3db8405d94b8d77710a7fb6556eac89d
Reviewed-on: https://code.wireshark.org/review/1423
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Make a dissector table out of IPv6 Next header.

This cleans up the implementation of the extension header, but also allows Decode As functionality for Next Header fields. This allows "old" SHIM6 IP protocol number (not IANA registered) to still be dissected.

bug:9995
Change-Id: I6f29a1f7db144478c5b6416d27da2c9c5a7a1de1
Reviewed-on: https://code.wireshark.org/review/1378
Reviewed-by: Michael Mann <mmann78@netscape.net>

Fix checkAPIs warnings in packet-lldp.c. repoted_length() used instead of captured_length() as is seems more logical

Change-Id: I02860d5c45e3fd794d9f99e2d382bc02ae886b36
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1422
Reviewed-by: Evan Huus <eapache@gmail.com>

This block of defines was always changed when I was retabbing my additions, this change fixes the tabs

Change-Id: I10e260a68cadf18d1aca10c7f037d45d311b6477
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1418
Reviewed-by: Evan Huus <eapache@gmail.com>

HTTP2: with HTTP2 draft-1x, now upgrade use h2-XX or h2c-XX for upgrade HTTP1 Protocol (with XX number of draft)

Change-Id: I7b6394e1a84db34e34b4a05245414b81a62292af
Partial-Bug:10054
Reviewed-on: https://code.wireshark.org/review/1409
Reviewed-by: Evan Huus <eapache@gmail.com>

Update SR and DRX tracking when see UL CRC error'd frame, and add an
option to add generated items corresponding to the middle of
BSR ranges (suitable for plotting/stats).

Change-Id: I73b4e82ea5fb869d93d9666b9f8c9a10562db365
Reviewed-on: https://code.wireshark.org/review/1419
Reviewed-by: Martin Mathieson <martin.r.mathieson@googlemail.com>

Fix potential invalid-reads in catapult files

As caught by CppCheck, the conditions checking that n is in bounds should
occur *before* the conditions using n as an index so that if n is out of bounds
we don't try and use it anyways.

Change-Id: I107c983153aa12203f8c88b14e1addd3807d6b6e
Reviewed-on: https://code.wireshark.org/review/1415
Reviewed-by: Michael Mann <mmann78@netscape.net>

With the use of tvb_reported_length_remaining(), there is no need to calculate consumedLen anymore

Change-Id: I8f4c32012061d3bff5e4efb80dfce8d82f7400df
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1417
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Fix Dereference of null pointer found by Clang analyzer

Change-Id: Icc25714cd923fcb514541f28ca19c7728dc1be55
Reviewed-on: https://code.wireshark.org/review/1412
Reviewed-by: Evan Huus <eapache@gmail.com>

Make filterable OSPF Header...

* Remove if(tree)
* use proto_tree_add_item
* Enhance checksum display
...

Change-Id: I66062ecbe7f0c8170fbe9c4a2d7654d62a87c18b
Reviewed-on: https://code.wireshark.org/review/1414
Reviewed-by: Michael Mann <mmann78@netscape.net>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Fix up indentation.

Change-Id: I9c217cfb0246c763ec7084bb928440d8aab88f99
Reviewed-on: https://code.wireshark.org/review/1416
Reviewed-by: Anders Broman <a.broman58@gmail.com>

DPoE OAM message decoding as per CableLabs spec.

Priority given to messages tested in DPoE 1.0 ATP.

Change-Id: I6ba3f1a8ca018f2231ad60f2f347ac57f1f93a00
Reviewed-on: https://code.wireshark.org/review/1076
Reviewed-by: Evan Huus <eapache@gmail.com>

Fix Restoration-Priority's AVP code.

Fixes https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=10057

Bug: 10057
Change-Id: I88bb6825006b038065f8edff61066557445f1dc6
Reviewed-on: https://code.wireshark.org/review/1413
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Add dissector for LINKTYPE_EPON

See IEEE Standard 802.3-2012 Section 5, Clause 65 and CableLabs DPoE
Security and Certificate Specification 1.0, Section 6.

Currently dissects 1G mode. 10G mode will be added when hardware is
available.

Change-Id: I6232af9bf6807644ef66a120d97e5fa5927988fe
Reviewed-on: https://code.wireshark.org/review/1284
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Evan Huus <eapache@gmail.com>

MBIM: make the backward compatibility function static

Change-Id: I51bbdf9c0dbb80a8350ec7762e588ca3182dce63
Reviewed-on: https://code.wireshark.org/review/1408
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Fix Dead Store (Dead assignement/Dead increment) warning found by Clang

Change-Id: I349a5aa7248979aa32eda9f853fd660dc01bdd08
Reviewed-on: https://code.wireshark.org/review/1407
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Add Support for IEEE DCBx (8021az)

Sample captures attached at https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=10055

Closed-bug:10055
Change-Id: Ib10e8d2a8aecdcf488eb174ed76a259433528f4a
Signed-off-by: Anish Bhatt <anish@chelsio.com>
Reviewed-on: https://code.wireshark.org/review/1403
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Fix issues reported by clang-analyzer

Change-Id: Iefb52c674be99af167eef7b2eba249607c1dfefd
Signed-off-by: Lorand Jakab <ljakab@ac.upc.edu>
Reviewed-on: https://code.wireshark.org/review/1406
Reviewed-by: Anders Broman <a.broman58@gmail.com>

Updated ospf link state router messages to use proto_tree_add_item instead of proto_tree_add_text.

Change-Id: Icb04aaa68e7e8b79116a6ac48d1897eb387aab3e
Reviewed-on: https://code.wireshark.org/review/1404
Reviewed-by: Anders Broman <a.broman58@gmail.com>

tvb_reported_length_remaining() returns gint

not size_t, which was giving 64/32 conversion errors on some platforms

Change-Id: Idf81dc98f8921a92840731d742d6e46a40e1387f
Reviewed-on: https://code.wireshark.org/review/1405
Reviewed-by: Evan Huus <eapache@gmail.com>

Add support for DCBX cee/cin protocol

CEE Spec here : http://www.ieee802.org/1/files/public/docs2008/az-wadekar-dcbx-capability-exchange-discovery-protocol-1108-v1.01.pdf

From a wireshark POV, CIN is the same as CEE with a different OUI subtype, and one less supported TLV.

bug:10039
Change-Id: I5aa4f1be2ecc43ecad6e86633231e6e39695af6e
Reviewed-on: https://code.wireshark.org/review/1330
Reviewed-by: Michael Mann <mmann78@netscape.net>

Change preference default to use RRC-signalled value for SN field length of DRBs.

Change-Id: Ibc56c901938f490f1fb0fede7e2fe0fbfc659250
Reviewed-on: https://code.wireshark.org/review/1402
Reviewed-by: Martin Mathieson <martin.r.mathieson@googlemail.com>

Fix rare case of wmem map being O(n)

For 'x' equal to 0, HASH() macro also returns 0 which makes wmem map O(n).
When random generator will return 0 just use 1.

Change-Id: If484091352a719aea27135a705d37ff4c184a13b
Reviewed-on: https://code.wireshark.org/review/1387
Reviewed-by: Michael Mann <mmann78@netscape.net>

Revert part of g757db64e484b009c3

It was intended to change the DTLS decryption test, but changed the SSL test
file instead, which led to the SSL test mysteriously failing. The SSL capture
really is http, so that's the right protocol, and the port is the standard 443,
not 4433 (which was perhaps a typo?).

Change-Id: I84448c2326d2a4301a4bba9607f8ba90a495531d
Reviewed-on: https://code.wireshark.org/review/1401
Reviewed-by: Evan Huus <eapache@gmail.com>

MBIM: fix compilation with GLib versions < 2.22.0

Change-Id: I27e794347dc0988f0f34798cb62f08a11a217176
Reviewed-on: https://code.wireshark.org/review/1400
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Fix Dereference of null pointer found by Clang analyzer

packet-gvcp.c:2101:7: warning: Access to field 'req_frame' results in a dereference of a null pointer (loaded from variable 'gvcp_trans')

Change-Id: If39453f9f2ade551fd8c7e369fd60325c16df24b
Reviewed-on: https://code.wireshark.org/review/1393
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Michael Mann <mmann78@netscape.net>

MBIM: use wmem hash map functions for request/response tracking

Change-Id: Ie550b0ef2b57827f610c4c6a83ff14868e853125
Reviewed-on: https://code.wireshark.org/review/1399
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

swIPe IP Security Protocol

bug:9996
Change-Id: I7c7ab7dfa81e44d6bf9affe7bfb0cd9f0a221adf
Reviewed-on: https://code.wireshark.org/review/1379
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

packet-http2: updated to http2 draft-12

From Alexis La Goutte
Fix warning found by pre-commit

Partial-Bug: 10054
Change-Id: I976884a240a55bb2287a802d72668a2c845179c0
Reviewed-on: https://code.wireshark.org/review/1295
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Tested-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Fix Dead Store (Dead assignement/Dead increment) warning found by Clang

Change-Id: I75aa99b020c92e7caf7efebe34cd1af3763cf592
Reviewed-on: https://code.wireshark.org/review/1397
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Add Modelines info

Change-Id: Ib6e1c68e47f2c66d444ac26d1d48f5ef3e14b949
Reviewed-on: https://code.wireshark.org/review/1396
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Add Modelines info

Change-Id: I8ce92fcd103a18fad2ea4c0b873914af38d48155
Reviewed-on: https://code.wireshark.org/review/1395
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Add Modelines info

Change-Id: I978198266523a0b1f0ba9d4ed6848664a6fa3637
Reviewed-on: https://code.wireshark.org/review/1394
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Fix warning found by -Wunused-const-variable (Clang 3.5)

packet-gvcp.c:493:27: error: unused variable 'linkconfiguration_class' [-Werror,-Wunused-const-variable]

Change-Id: Id31ced029af9700f47230df4afa5c0a4aaecf6d3
Reviewed-on: https://code.wireshark.org/review/1392
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Reorder fields in http_conv_t to save some memory.

On amd64 before patch: sizeof(http_conv_t) == 88, after: 72 bytes.

Change-Id: I2863f1c7223a904dadfbe68dc41bc61fc2cc378a
Reviewed-on: https://code.wireshark.org/review/1388
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

Try to fix Mac OS X buildbot
packet-http.c:2629: warning: implicit conversion shortens 64-bit value into a 32-bit value

Change-Id: I6a423639a53c24431fcfd79e0a235f2885ea86c2
Reviewed-on: https://code.wireshark.org/review/1389
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

prefs: fix ASAN error due to invalid indirection

For pref_current, indirection of pref->varp.string will cause a read of
size 8. This will cause a global buffer overflow error for all smaller
types, for example lbmc_use_heuristic_subdissectors (size 4).

Reproduce: compile Wireshark with -fsanitize=address, open Preferences
and select OK or Apply. Result: ASAN crash.

To fix this, only indirect a pointer if the storage size is known, a
void pointer stores the address of the constant value (pref_default,
pref_stashed) or the address to the value (pref_current). Note that
pointers of different types are of equal size, I could take
valp.pref_(anything).

While at it, remove superfluous 'break' keywords where a 'return'
keyword is present.

Change-Id: I05a69e8f14a1ecb4e5d2a0c0f0b71ed3f0a41d70
Reviewed-on: https://code.wireshark.org/review/1286
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Michael Mann <mmann78@netscape.net>

http: Fix src/dest mapping for proxied connections

Using value_is_in_range is making quite some assumptions, namely (1) the
proxy server is always run on a registered HTTP port, and (2) the
source (client) port is always not HTTP. The former is quite a strong
assertion which fails to hold when using a custom port (8008) that got
detected through heuristics.

Fix this by recording the source address and port pair for the server
and then check this against the current packet.

This fixes detection of a SSL conversation where two conversations got
detected instead of one. Example: 8008 is proxy, 443 is target server.
Now the proxied conversation got detected as 443 --> "client port"
(server to client, ok) and 443 --> 8008 (client to server, not ok,
should be "client port" --> 443).

bug:7717
Change-Id: I05113ec2aca6c9296184759a8a62eb32cbfcbb4f
Reviewed-on: https://code.wireshark.org/review/1380
Reviewed-by: Michael Mann <mmann78@netscape.net>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>

ssl/dtls: add keyfile support to dtls

This moves the keyfile and psk options from the ssl code into ssl-utils
and then uses them also for dtls.

This is the last missing part for bug 9499 from my side.

Change-Id: Ie2fe5bc565eabe1e6ce62498c985b8a36e913b0f
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Reviewed-on: https://code.wireshark.org/review/1369
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: Michael Mann <mmann78@netscape.net>

http: Display a list of cookies

For long cookies, the label "[truncated] Cookie: foo=v..." is not really
helpful. Add a new subtree to display individual cookies, this makes
copying values much easier.

A new "http.cookie_pair" field was added instead of re-using
"http.cookie". This has the advantage that `tshark -Tfields -e
http.cookie` does not end up with duplicates. At the same time, one can
match against individual cookie values.

I also considered to limit the number of cookies to be split, but as
there is no limit on the number of headers, I decided not to be
restrictive for cookies either.

Change-Id: I98d9522867811278ade3e04aab02e517f997928b
Reviewed-on: https://code.wireshark.org/review/1186
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Michael Mann <mmann78@netscape.net>

[Automatic manuf, services and enterprise-numbers update for 2014-04-27]

Change-Id: I0745df4a0a0447abf7ab4c7555d9d195ac40eb78
Reviewed-on: https://code.wireshark.org/review/1384
Reviewed-by: Gerald Combs <gerald@wireshark.org>

SAP: fix dissection of Originating Source address

Regression introduced in commit c131f9a0745f6307a5c910d8486eb9a37a05935c

Bug: 10029
Change-Id: Ia99ca36216dd26e41f926cc67e52827e665c40eb
Reviewed-on: https://code.wireshark.org/review/1383
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Limit the start of HTTP request/response to a printable character.

From Martin Mathieson.

In a profiled run with FTP traffic, the HTTP dissector looking for the end of a line of data (which was binary) was taking around 3% of runtime.

bug:8822
Change-Id: I2617d1e49030bd5ad85b0e818c48c01dc6fae075
Reviewed-on: https://code.wireshark.org/review/1373
Reviewed-by: Michael Mann <mmann78@netscape.net>

T38: fix access to uninitialised variables

Bug: 10050
Change-Id: I7bc2d1b99bcb1241705a2561d9360a006fb9cb47
Reviewed-on: https://code.wireshark.org/review/1382
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

SSL: fix compilation without GnuTLS and Gcrypt

Change-Id: Ic783f9ac2b7abcbace67ce48239b0506ac237267
Reviewed-on: https://code.wireshark.org/review/1381
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>

Change preferences - always want to try to decrypt signalling PDUs

Change-Id: Ib34f12b5f8dd276612aed2fe0192c94e847858f2
Reviewed-on: https://code.wireshark.org/review/1377
Reviewed-by: Martin Mathieson <martin.r.mathieson@googlemail.com>

Have DTLS just dissect the test capture as data

Follow-up to g757db64e484b009c33b67b5fa38e109d7b8f5e78 which changed the filter
being tested but didn't change the target protocol, so the test was still
failing because it was still trying to use HTTP.

Change-Id: I6675cfad3bba63f7a536eb7ae82e4b25132d108e
Reviewed-on: https://code.wireshark.org/review/1375
Reviewed-by: Evan Huus <eapache@gmail.com>

Fix FSF address.

Change-Id: Ic79fcf7e1bebb9821d9d9db43c2a4f4291beb706
Reviewed-on: https://code.wireshark.org/review/1374
Reviewed-by: Guy Harris <guy@alum.mit.edu>

dtls: run dissector for decryption also on server hello extensions

The dissector only ran through the server hello extensions for the tree
and not in the ssl decryption pass. This resulted in
ssl_dissect_hnd_hello_ext() being always called with ssl == NULL. For
SSL this was also called with ssl != NULL.

Change-Id: I22f7b1089731124b3ca1a2b8515f307c4a021b7f
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Reviewed-on: https://code.wireshark.org/review/1370
Reviewed-by: Evan Huus <eapache@gmail.com>

ssl: support any address/port combi for priv key

Both "any port" and "any address" are supported separatedly, but not the
combination of both. This also has the effect that the combination of
any address with the special keyword "start_tls" did not work.

Fix this by checking for a private key with the combination of any
address and port.

Change-Id: Icb49d6728f032a05007dcb7ac73ec0528778441a
Reviewed-on: https://code.wireshark.org/review/1368
Reviewed-by: Evan Huus <eapache@gmail.com>

ssl: return early if no private keys are found

There is no need to check for private keys if there are none. In
addition, print the number of keys for debugging purposes.

Change-Id: Idc9d650e0bf087c0f647dba4e5bd4920b4f6e228
Reviewed-on: https://code.wireshark.org/review/1367
Reviewed-by: Evan Huus <eapache@gmail.com>

ssl: Reduce chance of hash-collisions SSL private keys

The wildcard address contains all zeroes, resulting in the same hash
for 0.0.0.0 and ::. Not really problematic, but it does not sound
great either.

Change-Id: I099128973a1bd8bb5c88d0abcab3ea4ecc3a96c9
Reviewed-on: https://code.wireshark.org/review/1366
Reviewed-by: Evan Huus <eapache@gmail.com>

Change ssl_find_private_key to void

No caller checks its return value (which is always 0).

Change-Id: I18461ee6e5d369722c8c2b2ea1e409423aa5d631
Reviewed-on: https://code.wireshark.org/review/1365
Reviewed-by: Michael Mann <mmann78@netscape.net>
Reviewed-by: Evan Huus <eapache@gmail.com>

Added fields used by HP MSM access points.

Change-Id: I367c99bb351993f05161d683eb54f08e5852145f
Reviewed-on: https://code.wireshark.org/review/1347
Reviewed-by: Michael Mann <mmann78@netscape.net>

Dissect all details in NAT-T LCAF type...

... and remove last remaining proto_tree_add_text() calls!

Change-Id: I22e5446a06c22ba1f30f342b21f7676641a7f2e7
Signed-off-by: Lorand Jakab <ljakab@ac.upc.edu>
Reviewed-on: https://code.wireshark.org/review/1352
Reviewed-by: Michael Mann <mmann78@netscape.net>

Updated gvcp dissector

Change-Id: I49f6acecdbcdf171ba28af171f8067322cc5ecf1
Reviewed-on: https://code.wireshark.org/review/1220
Reviewed-by: Michael Mann <mmann78@netscape.net>

wtap_read() and wtap_seek_read() handle caplen > len; don't do it here.

Change-Id: Iac083abdbaaa4666124e5b296e6f8306ee90f32d
Reviewed-on: https://code.wireshark.org/review/1363
Reviewed-by: Guy Harris <guy@alum.mit.edu>

This doesn't have separate captured and actual length, either.

Change-Id: I4605bde8a2951ea78d507bab6fcf004ebef6f38f
Reviewed-on: https://code.wireshark.org/review/1362
Reviewed-by: Guy Harris <guy@alum.mit.edu>

Put the common code for reading Logcat packets into a routine.

Then have the read and seek-read routines both use that routine.

Change-Id: I3d11df82644207d0ae59486231c91e1f044090ab
Reviewed-on: https://code.wireshark.org/review/1361
Reviewed-by: Guy Harris <guy@alum.mit.edu>