Instead of grabbing the set of IDBs found at open time, have a loop
using wtap_get_next_interface_description() to read all unread IDBs run
after opening the input file, after reading a packet from the input
file, and after getting an EOF on the input file.
Add a routine wtap_uses_interface_ids() to check whether the file type
and subtype for a dump file uses interface IDs and requires IDBs. If
so, in the aforementioned loop, add the IDBs to the dump stream.
Add a routine wtap_dump_add_idb() to add IDBs to a dump stream. Have it
call a file-format-specific routine to add the IDBs; the only file type
that supports it is pcapng, and it 1) writes out the IDB and 2) adds it
to the set of IDBs for the stream.
Add a wtap_dump_params_init_no_idbs() routine that prevents the IDBs
from the input file from being used to initialize the output file; use
it in cases where we're using the aforementioned loop to copy over IDBs.
Don't require any IDBs to be present when opening a pcapng file for
writing; 1) the simplest pcapng file has just an SHB in it, 2) that
requirement causes dumps that don't provide IDBs at open time to fail,
and 3) the real issue is that we don't want packets with an interface ID
not corresponding to a known IDB, and we already have a check for that.
(There are some hacks here; eventually, when everything processes the
IDBs in such a loop, we may be able to get rid of the "two favors of
dump parameter initialization" hack.)
Fixes#15844.
Addresses the same issue in #15502, but there are other issues there
that also need to be addressed.
In addition, the merge code also needs to be changed to handle this.
Fix Qt 5.15 deprecation warnings in QCustomPlot, similar to 76d92ba7e7.
Use default flags constructors instead of 0.
Use QWheelEvent::angleDelta() instead of QWheelEvent::angle().
Use QWheelEvent::position() instead of QWheelEvent::pos().
Use date::startOfDay() instead of QDateTime(date).
Use QMultiMap instead of QMap where needed.
According to [MS-FSCC] if the file has the REPARSE_TAG attribute, the
EaSize field must be interpreted as a reparse tag for the following
info levels:
* FileFullDirectoryInfo
* FileBothDirectoryInfo
* FileIdFullDirectoryInfo
* FileIdBothDirectoryInfo
From tools/check_typed_item_calls.py output:
epan/dissectors/packet-bssgp.c:655 proto_tree_add_item called for hf_bssgp_bss_area_ind - item type is FT_UINT8 but call has len 2
epan/dissectors/packet-bssgp.c:1468 proto_tree_add_item called for hf_bssgp_unit_val - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:1469 proto_tree_add_item called for hf_bssgp_gprs_timer - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:2606 proto_tree_add_item called for hf_bssgp_unit_val - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:2607 proto_tree_add_item called for hf_bssgp_gprs_timer - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:2635 proto_tree_add_item called for hf_bssgp_unit_val - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:2636 proto_tree_add_item called for hf_bssgp_gprs_timer - item type is FT_UINT8 but call has len 3
epan/dissectors/packet-bssgp.c:3276 proto_tree_add_item called for hf_bssgp_cell_acc_mode - item type is FT_UINT8 but call has len 4
It currently wraps wtap_block_create() and wtap_block_copy(); if there
are no remaining use cases for wtap_block_copy() at some point, it can
just *replace* wtap_block_copy().
In a wtap, keep track of the first interface description not yet fetched
with wtap_get_next_interface_description() and, when
wtap_get_next_interface_description() is called, have it return that
description, as a wtap_block_t for its IDB. If there are no
as-yet-unfetched interface descriptions, return NULL; there may, in the
future, be more interface descriptions for the file, so this should be
called:
* after the file is opened;
* after wtap_read() returns TRUE, indicating that it's returned a
record (and *before* you process the record that wtap_read()
returns, as it might be the interface description for the
interface on which the packet in that record arrived);
* after wtap_read() returns FALSE, indicating an EOF or an error
return (as there might have been interfaces at the end of the
file or before the error point).
At each of those points, the caller should loop until
wtap_get_next_interface_description() returns NULL.
Not used yet (but tested with capinfos, which found a reason why you
have to wait until the end of the file before processing the interface
information - there's now a comment in the code giving that reason).
This will probably be used in the future.
Add support internally to using iconv (always present with glib) to convert
strings from various encodings to UTF-8 (using REPLACEMENT CHARACTER as
recommended), and use that to support GB 18030 and EUC-KR. Replace call
directly to iconv in ANSI 637 for EUC-KR to new API. Update comments
and documentation around character encodings. It is possible to replace
the calls to iconv with an internal decoder later. Tested on Linux and
on Windows (including with illegal characters). Closes#16630.
For WPA security association (SA) entries are created on sucessful
PTK derivation from 4-way handshake frames. WEP though don't use
4-way handshake frames for key derivation and therefore no SA entry
is created. Still WEP decryption implementaton expects to find
an SA otherwise the decryption is skipped.
Fix broken WEP decryption by removing the check for an existing SA
entry and instead form the SA on first successful decryption.
Add also a test for WEP decryption.
Fixes: v3.3.0rc0-1263-g099d241046 ("dot11decrypt: Avoid allocating SA on packet decryption")
The SOCKS dissector temporarily changes the pinfo values for destport
or srcport, so it should get the tcp_conversation_data after doing so
before recursively calling the TCP dissector again. Otherwise the TCP
dissector will be confused about whether a TCP multisegment PDU is in
progress or not, causing failure to lookup and store fragments correctly,
including both failed desegmentation and failed asserts (when it expects
an entry in the table which isn't there, as it was stored under a different
port number.) Fixes#16646.
In captures where a lot of packets are missing, requests and ACKs are
sometimes incorrectly paired. With this improvement, ACKs must arrive in
a reasonable time to be paired with a request.
What we care about here is the packet type, not the file type, as we
care what's at the beginning of the packet. This should have no effect
in practice, but it makes it clearer what we're testing for.
Currently, the only file types that use them are pcapng and IBM's
iptrace; we don't support writing the latter, so this is mainly of
interest for pcapng.
This makes it a bit more obvious what some "is this pcapng?" tests are
really trying to determine, and allows them to automatically support any
new file types that use them.
(With regard to interface descriptions, tere are three types of file:
1) files that contain no interface information;
2) files that contain "just FYI" interface information but that don't
tie packets or other records to particular interfaces;
3) files that contain interface information and tie all packets (and
possibly other records) to an interface.
This tests for files of type 3.)
A previous change initialized the k_connection_handle, so we don't
compare random data with remote_bdaddr->chandle, but perhaps we
shouldn't compare it at all if we didn't find a handle pair.
MSVC doesn't, by default, define __STDC_VERSION__, which means that the
code generated by newer versions of winflexbison3's Bison end up
defining YYPTRDIFF_T as long, which is wrong on 64-bit Windows, as
that's an LLP64 platform, not an LP64 platform, and causes warnings to
be generated. Those warnings turn into errors.
With MSVC, if __STDC_VERSION__ isn't defined, Forcibly include
<stdint.h> here to work around that.
Fixes#16924.
Bison 3.4 and later generate deprecation warnings for the "%pure-parser"
directive. As https://git.savannah.gnu.org/cgit/bison.git/tree/NEWS says,
----
** Deprecated features
The %pure-parser directive is deprecated in favor of '%define api.pure'
since Bison 2.3b (2008-05-27), but no warning was issued; there is one
now. Note that since Bison 2.7 you are strongly encouraged to use
'%define api.pure full' instead of '%define api.pure'.
----
Rename our .y files to .y.in, and modify FindYACC.cmake to detect newer
versions of Bison and configure our .y files with "%pure-parser" or
"%define api.pure" as needed. Squelches warnings from Bison in #16924.
The google.protobuf.Timestamp is a standard protobuf message type and
consists of seconds and nanos fields. We dissect protobuf field in
google.protobuf.Timestamp type as wireshark FT_ABSOLUTE_TIME field.
And add tvb_get_protobuf_field_uint() to make it easy to get a
Protobuf field of varint type from the tvb.
close#16927
The function to dissect CSuite Sel returns offset not number of
dissected bytes so calling function must assign new offset rather
than incrementing. For consistency also update the CSuite List
function to return offset.
Fix issues found by running ./tools/check_typed_item_calls.py
epan/dissectors/packet-eap.c:1475 proto_tree_add_item called for hf_eap_gpsk_failure_code - item type is FT_UINT16 but call has len 4
epan/dissectors/packet-eap.c:1479 proto_tree_add_item called for hf_eap_gpsk_failure_code - item type is FT_UINT16 but call has len 4
Speed functions to print hex bytes, escape XML strings and
print out indents by avoiding specifier calls, and building
larger strings before calling fputs().
Someone mentioned this in the sharkfest chat yesterday.
Also, Ostinato relies upon this when importing from pcap.
An example capture I have has gone from 18 to 11 seconds.
As noted in bug #16386, glib's `g_base64_decode_inplace()` aborts
decoding of base64 strings that aren't padded. This addresses that by
adding padding "=" characters if needed to the buffer which will be
decoded.
I added the test case from the bug report to the test suite, though the
location therein may not be ideal.
Closes#16386
Found by running ./tools/check_typed_item_calls.py
epan/dissectors/packet-ieee80211.c:14209 proto_tree_add_item called for hf_ieee80211_osen_akm_count - item type is FT_UINT8 but call has len 2
epan/dissectors/packet-ieee80211.c:20025 proto_tree_add_item called for hf_ieee80211_tclas_ether_type - item type is FT_UINT8 but call has len 2
Fix issues found by running ./tools/check_typed_item_calls.py
epan/dissectors/packet-gtp.c:4414 proto_tree_add_item called for hf_gtp_sel_mode - item type is FT_UINT8 but call has len 2
epan/dissectors/packet-gtp.c:6807 proto_tree_add_item called for hf_gtp_rai_rac - item type is FT_UINT8 but call has len 2
epan/dissectors/packet-gtp.c:7600 proto_tree_add_item called for hf_gtp_bssgp_cause - item type is FT_UINT8 but call has len 2
epan/dissectors/packet-gtpv2.c:3607 proto_tree_add_item called for hf_gtpv2_trace_id - item type is FT_UINT16 but call has len 3
epan/dissectors/packet-gtpv2.c:5049 proto_tree_add_item called for hf_gtpv2_trace_id - item type is FT_UINT16 but call has len 3
Make display_signed_time() take a 64-bit signed number of seconds, and,
in calls to it, cast the argument to gint64, not gint32.
Addresses issue #16909.
This unfortunately includes the name of the filter element but "sack" in TCP
should not mean "a packet with syn+ack set" to most networking people nowadays.
Jaap Keuter