In commit 4afd70d ("Use g_hash_table_new_full to free some values"), the
hashtable gained a destroy handler which frees memory. This
inadvertently destroyed a conversation during key updates.
Fix this by not calling _remove (and thereby calling the destroy
handler), but use _steal instead. (Suggestion by Evan Huus).
Bug: 10263
Change-Id: I9fa7f5a697599f42894d38718b00b9c0c1b57004
Reviewed-on: https://code.wireshark.org/review/2924
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: Evan Huus <eapache@gmail.com>
Encountered with the following attachments from bugs:
6575-test-dlmapc-wmx
6577-test-dlmapc-wmx
6579-test-dlmapc-wmx
Also added modelines.
Change-Id: I3859bed618830ad359bd0b2b1fdfc3c5fc13269e
Reviewed-on: https://code.wireshark.org/review/2920
Reviewed-by: Evan Huus <eapache@gmail.com>
target type
Change-Id: If7efc629247480fbfb231b965a358803b67e1504
Reviewed-on: https://code.wireshark.org/review/2919
Reviewed-by: Anders Broman <a.broman58@gmail.com>
* use grep -c instead of a grep/wc combo;
* use cmp instead of diff, since we're only interested in whether the files
are identical;
* remove the temporary file if it's the same as the existing one.
Change-Id: If61c5675535c483b9dc2aee179d7be58c7bd6e4f
Reviewed-on: https://code.wireshark.org/review/2894
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
For each displayed packet list row, save a copy of or a pointer to
column strings similar to ui/gtk/packet_list_store.c. This lets us call
epan_dissect_run only once per row.
Bug: 9511
Change-Id: I17e8ebeb5ed70518c9047413c3b2a46f01e904ef
Reviewed-on: https://code.wireshark.org/review/2752
Reviewed-by: Anders Broman <a.broman58@gmail.com>
New dissectors return an integer, not a boolean, fix that.
Change-Id: I79dea0cc4274d503d61ce8823dc783c542764f6b
Reviewed-on: https://code.wireshark.org/review/2884
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: Evan Huus <eapache@gmail.com>
Change-Id: Ie10ac4f5f04a23344d183e095bbafb23c6409144
Reviewed-on: https://code.wireshark.org/review/2904
Reviewed-by: Michael Mann <mmann78@netscape.net>
The WRETH dissector showed up some garbage in the column display. Upon
further inspection, it turns out that the format string had a trailing
percent sign which caused (unsigned)-1 to be returned by
g_printf_string_upper_bound (in emem_strdup_vprintf). Then ep_alloc is
called with (unsigned)-1 + 1 = 0 memory, no wonder that garbage shows
up. ASAN could not even catch this error because EP is in charge of
this.
So, start adding G_GNUC_PRINTF annotations in each header that uses
the "fmt" or "format" paramters (grepped + awk). This revealed some
other errors. The NCP2222 dissector was missing a format string (not
a security vuln though).
Many dissectors used val_to_str with a constant (but empty) string,
these have been replaced by val_to_str_const. ASN.1 dissectors
were regenerated for this.
Minor: the mate plugin used "%X" instead of "%p" for a pointer type.
The ncp2222 dissector and wimax plugin gained modelines.
Change-Id: I7f3f6a3136116f9b251719830a39a7b21646f622
Reviewed-on: https://code.wireshark.org/review/2881
Reviewed-by: Evan Huus <eapache@gmail.com>
When ASAN is enabled, a false buffer overflow warning is raised. It is
a false positive since everything starting at '\0' will get ignored by
the PCMPISTRI instruction (see Intel(r) SSE4 Programming Reference,
5.3.1.5 "Valid/Invalid Override of Comparisons", and 5.3.1.2
"Aggregrate Operation", case "Equal any").
Concerns about reading past the end of the page turns out to be false,
there always seem to be a valid page after the current one (for static
and heap memory at least). It is an non-issue since strlen also does
not have issues with this.
Rather than fully disabling SSE 4.2 and using the fallback
implementation for ASAN-enabled builds, read the set of characters from
'a' into the mask without a 128-bit read and then still use SSE 4.2
for the actual query.
Bug: 10214
Change-Id: Ie4a526e60b43bfc08dd1d821556766f14a49be4d
Reviewed-on: https://code.wireshark.org/review/2618
Reviewed-by: Peter Wu <peter@lekensteyn.nl>
Reviewed-by: Evan Huus <eapache@gmail.com>
With autotools, CMake, and nmake, if we have a function, #define
HAVE_{function_name_in_all_caps}, otherwise don't #define it.
If we provide our own version of a function in libwsutil, make sure we
have a header that declares it, and *ONLY* include that header if
HAVE_{function_name_in_all_caps} is *NOT* defined, so that we don't have
the system declaration and our declaration colliding.
Check for inet_aton, strncasecmp, and strptime with CMake, just as we do
with autotools.
Simplify the addition of {function_name_in_all_caps}_LO to libwsutil in
autotools.
Change-Id: Id5be5c73f79f81919a3a865324e400eca7b88889
Reviewed-on: https://code.wireshark.org/review/2903
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Add support for Prefix, STAP, MTAP, NI-MTAP, and PACSI packet types.
Add support for Microsoft SEI messages [MS-H264PF]
Add support for dissecting scalable profiles
SDP: Add profile-level-id decode for payload type H264-SVC
MS-H264PF: http://msdn.microsoft.com/en-us/library/hh659565.aspx
Update #1 - Fix Tabs -> Spaces, Reinsert accidentally removed entry 19
from h264_type_values
Update #2 - Changed to using expert info for exceptions and Microsoft
errata.
Update #3:
- Correct handling of truncated packets
- Use guid functions and compare techniques
- Correct ranges for expert info messages
- Change to using reported_length from captured_length
Change-Id: I520a3c9a6d85c78a976b520cf5a6a405064a48f1
Reviewed-on: https://code.wireshark.org/review/2580
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>
Change-Id: Id1997c3c8f6d0460da7f0faa58798a72737dd667
Reviewed-on: https://code.wireshark.org/review/2898
Reviewed-by: Michael Mann <mmann78@netscape.net>
Change-Id: I3efa2a81ab2685cde6eae0a00b24520478a545ce
Reviewed-on: https://code.wireshark.org/review/2900
Reviewed-by: Michael Mann <mmann78@netscape.net>
As suggested by Alexis reviewing change Ie76c5a810af927b, this issue was
scattered throughout the v4/v5 openflow dissectors. Hopefully I got all of them.
Original issue found in bug 10259.
Change-Id: Idf87deea34132aec5cd4e8f83264906a29b15568
Reviewed-on: https://code.wireshark.org/review/2878
Reviewed-by: Evan Huus <eapache@gmail.com>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
The return value of new dissectors is the number of bytes that were
successfully dissected, not a boolean. Fix that and get rid of an
unnecessary iSCSIPdusDissected variable.
Change-Id: Ie31df393a1eb44f185d320a4c2d35f5e8b7d7bd9
Reviewed-on: https://code.wireshark.org/review/2889
Reviewed-by: Evan Huus <eapache@gmail.com>
new dissectors return an int, not a boolean...
Change-Id: I88e19f7c0dc14da3649d1522ffe936538a867753
Reviewed-on: https://code.wireshark.org/review/2888
Reviewed-by: Evan Huus <eapache@gmail.com>
Return value "-1" means that more data is requested. The dissector
actually doesn't have any idea what to do with the data, so just return
the data dissected so far.
tcp stream 3182 from c05-http-reply-r1.pcap (SampleCaptures) was
incorrectly detected as MIKEY, it was actually HTTP...
Change-Id: Idca3d3e2f85e821df70436a675699b5834236f89
Reviewed-on: https://code.wireshark.org/review/2887
Reviewed-by: Evan Huus <eapache@gmail.com>
uint16 & 0xFF00 will always equal 0. Use tvb_get_guint8 instead. Convert
from tvb_length to tvb_captured_length.
Fix the return value of dissectors, 0 means reject but that cannot be
combined with reassembly requests.
Change-Id: I5fca66e2e809699392237aff5813eecdfb15857f
Reviewed-on: https://code.wireshark.org/review/2885
Reviewed-by: Evan Huus <eapache@gmail.com>
Fixes a good 80-90KB of leaks in certain cases.
Bug: 10261
Change-Id: I81d57ac67219e730b03649b9fdfc2306807bdb97
Reviewed-on: https://code.wireshark.org/review/2879
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Change-Id: Ia2567695ffed30c990eda3740b08bfab101cea96
Reviewed-on: https://code.wireshark.org/review/2883
Reviewed-by: Michael Mann <mmann78@netscape.net>
The pStr argument to dissect_mq_charv() isn't modified (and always
points to a character string), so make it a "const char *", and
eliminate the casts to "guint8 *" in calls to it.
Change-Id: I21dad38c41324528be297a8ddc1854beff2276db
Reviewed-on: https://code.wireshark.org/review/2877
Reviewed-by: Guy Harris <guy@alum.mit.edu>
BYTE2WORD() doesn't modify what its argument points to, so make that
argument a const unsigned char *.
This lets us get rid of casts that cast away constness.
Change-Id: I44a58bd3d75fc77a022b7e8f7fa9b43990bcf81c
Reviewed-on: https://code.wireshark.org/review/2876
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Instead of calling the grep/sed pipelines for each file, build the
list of files in the beginning and call each pipeline only once,
passing the list to the first grep.
This results in a massive speedup in Cygwin; in my test, the time
it takes to run make-dissector-reg . dissectors packet-*.c in dissectors/epan
is reduced from ~116 to ~3 seconds. I also tried it on NetBSD, where
the time do to the same goes from ~6 to ~0.5 seconds.
Amend makefile comments to elide mentions of invoking multiple processes
per file.
Change-Id: Iad441e7d2b6cc3669dada57646e2f8f6b987fd34
Reviewed-on: https://code.wireshark.org/review/2826
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Pull the "cast this to a pointer to an __m128i" idiom into a macro, and
use the macro; have that macro use "const" in the casts.
Change-Id: Ife90b7daef2f09368790f3b2ffbb227d6d794dea
Reviewed-on: https://code.wireshark.org/review/2871
Reviewed-by: Guy Harris <guy@alum.mit.edu>
That combines more common code from the read and seek-read code
paths.
Also, separate out the individual metadata record types, with a comment
for each, to simplify the process of supporting some or all of them in
the future.
Change-Id: Ic8ded397d9550ec6013c1f5f138333b1ef5c37e5
Reviewed-on: https://code.wireshark.org/review/2869
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Change-Id: I66f0bffb987568c3d4c14a06bdc90465c877b27f
Reviewed-on: https://code.wireshark.org/review/2867
Reviewed-by: Michael Mann <mmann78@netscape.net>
Peter Wu