Adds a pre-commit hook for detecting and replacing
occurrences of `g_malloc()` and `wmem_alloc()` with
`g_new()` and `wmem_new()`, to improve the
readability of Wireshark's code, and
occurrences of
`g_malloc(sizeof(struct myobj) * foo)`
with
`g_new(struct myobj, foo)`
to prevent integer overflows
Also fixes all existing occurrences across
the codebase.
That makes them work as input to a mergecap that writes pcapng files.
File types that don't have a single per-file encapsulation type need
more work, with multiple fake IDBs, one for each packet encapsulation
type seen in the file, unless we can generate real IDBs.
Change-Id: I2859e4f7fb15ec0c0f31a4044dc15638e5db7826
Reviewed-on: https://code.wireshark.org/review/37983
Petri-Dish: Guy Harris <gharris@sonic.net>
Tested-by: Petri Dish Buildbot
Reviewed-by: Guy Harris <gharris@sonic.net>
The call must ensure enough bytes are in the buffer for subsequent
casts. Next cast is for nspr_pktracefull_v20_t.
Change-Id: I8b77aa243f528f82786af1047e8d26100f306a07
Reviewed-on: https://code.wireshark.org/review/35837
Reviewed-by: Dario Lombardo <lomato@gmail.com>
Petri-Dish: Dario Lombardo <lomato@gmail.com>
Tested-by: Petri Dish Buildbot
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Change all wireshark.org URLs to use https.
Fix some broken links while we're at it.
Change-Id: I161bf8eeca43b8027605acea666032da86f5ea1c
Reviewed-on: https://code.wireshark.org/review/34089
Reviewed-by: Guy Harris <guy@alum.mit.edu>
That makes it - and the routines that implement it - work more like the
seek-read routine.
Change-Id: I0cace2d0e4c9ebfc21ac98fd1af1ec70f60a240d
Reviewed-on: https://code.wireshark.org/review/32727
Petri-Dish: Guy Harris <guy@alum.mit.edu>
Tested-by: Petri Dish Buildbot
Reviewed-by: Guy Harris <guy@alum.mit.edu>
That made PACKET_DESCRIBE() do nothing, causing warnings from the Clang
Static Humiliator.
Change-Id: I6f433cd193b6398d89038e95c7bf5deb24aa186d
Reviewed-on: https://code.wireshark.org/review/32437
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Move the read out of nspm_signature_version(), to make it a bit clearer
what file I/O we do in the open process; have nspm_signature_version()
just look for a signature in a single page.
In the loop in nspm_signature_version(), make sure we have enough of the
record header to look at the type and length fields in that header
before looking at them and, when we can look at them, make sure the
length of the record 1) fits in what remains of the page we're looking
at and 2) is big enough to be the length of a signature record.
Change-Id: I7d625859136e6f39c40b166067fc7efea806d9b0
Reviewed-on: https://code.wireshark.org/review/32426
Petri-Dish: Guy Harris <guy@alum.mit.edu>
Tested-by: Petri Dish Buildbot
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Fix more crashes found in the provided bug report.
Bug: 15497
Change-Id: If84498fa879ad56c8677f8c1442a8dc0e5906003
Reviewed-on: https://code.wireshark.org/review/32333
Petri-Dish: Dario Lombardo <lomato@gmail.com>
Tested-by: Petri Dish Buildbot
Reviewed-by: Anders Broman <a.broman58@gmail.com>
When testing the signature field against a given version's signature:
we require that the signature field's size (size, not C
null-terminated string length) be at least the size of the
signature string (otherwise, it can't possibly match);
we check to make sure that the first N bytes of the signature
field, where N is the size of the version's signature string
(not including any terminating '\0' in that string), match the
version's signature string.
I.e., we require that the version's signature string is a prefix of the
signature string in the file.
This does not require that the signature string in the file be
null-terminated.
It also doesn't allow the file's signature string to be a substring of
the version's signature string, as that's *NOT* sufficient to identify
the file as a NetScaler trace file, especially if we forcibly
null-terminate the file's signature string and we trucate it to be
zero-length, as, in that case, it's *always* a prefix of the version's
signature string, and the file is incorrectly identified as a NetScaler
trace file.
(While we're at it, we make the nspm_signature_isvXXX() routines return
true if it *is* and false if it *isn't*, rather than the reverse; having
a routine with a name containing "is", and not "isnt", return true if it
*isn't* is confusing.)
Change-Id: I3694773a71b8b63d280e42f146698c82a0f0c332
Ping-Bug: 15601
Reviewed-on: https://code.wireshark.org/review/32403
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Multiple bugs have been found in the provided bug. Some of them have
been fixed in gefe920a, others here. The main problem is when malformed
files give wrong lenghts to the code, that casts and dereference it
without checking, causing oob reads. The fix introduces a check function
that prevents to go beyond the limits, early returning with a malformed
file message.
Other bugs have been fixed by forcing the string terminator that allows
the use of strlen() and MIN() that prevent wrong reads.
Bug: 15497
Change-Id: I8411208b5ea0f1a0720a17b882f704d03296d1c4
Reviewed-on: https://code.wireshark.org/review/32194
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot
Reviewed-by: Gerald Combs <gerald@wireshark.org>
When reading a malformed packet, it can occur that we go close to
the end of the buffer. We need to check if we have 2 bytes before
reading a uint16.
Bug: 15497
Change-Id: I2b00f44933ca11b925ffbf05b9855684feebcda5
Reviewed-on: https://code.wireshark.org/review/32028
Petri-Dish: Dario Lombardo <lomato@gmail.com>
Tested-by: Petri Dish Buildbot
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Separate the stuff that any record could have from the stuff that only
particular record types have; put the latter into a union, and put all
that into a wtap_rec structure.
Add some record-type checks as necessary.
Change-Id: Id6b3486858f826fce4b096c59231f463e44bfaa2
Reviewed-on: https://code.wireshark.org/review/25696
Reviewed-by: Guy Harris <guy@alum.mit.edu>
The first is deprecated, as per https://spdx.org/licenses/.
Change-Id: I8e21e1d32d09b8b94b93a2dc9fbdde5ffeba6bed
Reviewed-on: https://code.wireshark.org/review/25661
Petri-Dish: Anders Broman <a.broman58@gmail.com>
Petri-Dish: Dario Lombardo <lomato@gmail.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Use WTAP_MAX_PACKET_SIZE_STANDARD, set to 256KB, for everything except
for D-Bus captures. Use WTAP_MAX_PACKET_SIZE_DBUS, set to 128MB, for
them, because that's the largest possible D-Bus message size. See
https://bugs.freedesktop.org/show_bug.cgi?id=100220
for an example of the problems caused by limiting the snapshot length to
256KB for D-Bus.
Have a snapshot length of 0 in a capture_file structure mean "there is
no snapshot length for the file"; we don't need the has_snap field in
that case, a value of 0 mean "no, we don't have a snapshot length".
In dumpcap, start out with a pipe buffer size of 2KB, and grow it as
necessary. When checking for a too-big packet from a pipe, check
against the appropriate maximum - 128MB for DLT_DBUS, 256KB for
everything else.
Change-Id: Ib2ce7a0cf37b971fbc0318024fd011e18add8b20
Reviewed-on: https://code.wireshark.org/review/21952
Petri-Dish: Guy Harris <guy@alum.mit.edu>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Guy Harris <guy@alum.mit.edu>
The loop was using bytes_read, but wasn't setting it. Go back to
something similar to the previous loop condition, but don't lose the
error tests.
Fixes Coverity CID 1403388.
Change-Id: I557cbfa6e9ad81491af4fc90e85ce87c71fec8aa
Reviewed-on: https://code.wireshark.org/review/20776
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Makes Windows vscodeanalysis a little happier.
Change-Id: Ie744e91ab3f2a9744ae21c932ab6ea25467ad2fa
Reviewed-on: https://code.wireshark.org/review/20724
Petri-Dish: Michael Mann <mmann78@netscape.net>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Michael Mann <mmann78@netscape.net>
Also, if we return WTAP_OPEN_ERROR from an open routine after we've set
our close routine, that routine is called, which frees up our private
data structures; don't free them ourselves before returning
WTAP_OPEN_ERROR.
Change-Id: I03eebe1a1677e2161fdacec8de14668093cf03a3
Reviewed-on: https://code.wireshark.org/review/20522
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Bug: 13478
Change-Id: I6be2972979ff7cabf27e70d236c581d539d6ddac
Reviewed-on: https://code.wireshark.org/review/20515
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 maximum record length is 255*128 + 127 = 32767; that fits in a
guint32, which is large enough to support the biggest packet we'd ever
support without stretching several size values to 64 bits.
It's not a size of an object in memory, so it doesn't have to be a
size_t, and a size_t could be too large to fit in the record sizes we're
using.
Just cast to guint32.
Change-Id: Ie664fda3ce9945893fd992bbb9a81a5d632a3fcb
Reviewed-on: https://code.wireshark.org/review/20479
Reviewed-by: Guy Harris <guy@alum.mit.edu>
When vmnames are included in the header of a netscaler packet trace,
number of bytes equal to the size of vmnames is omitted from the packet,
by the dissector.
Bug: 13459
Change-Id: I0f907e9c2e08c1cbebd47f7e50d8284a6aaade59
Reviewed-on: https://code.wireshark.org/review/20446
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>
The record size fields are guint8, but NSPR_V20RECORDSIZE_2BYTES was
0x80, which has type int, promoting the result to int. Make it 0x80U,
which means everything is unsigned.
This squelches a compiler warning.
Change-Id: I1c63e485352a90c7f675ab0dacaaeba794235b35
Reviewed-on: https://code.wireshark.org/review/20344
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Do the check early in the process of processing the record, and do it
for all record types.
Bug: 13429
Change-Id: Id7f4d12415c6740241850d8f873cff52909e7110
Reviewed-on: https://code.wireshark.org/review/20330
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Records in a properly formatted NetScaler file shouldn't go past the end
of a page, but nothing guarantees that a NetScaler file will be properly
formatted.
NetScaler 3.x files allow record bodies to go past the end of a page,
but 1.x and 2.x files don't, so treat record headers that go past the
end of a page, and record bodies in 1.x and 2.x files that go past the
end of a page, as errors.
Clean up some stuff while we're at it.
Bug: 13430
Change-Id: I3b1d56086e3bb14b246406f306e3d730df337561
Reviewed-on: https://code.wireshark.org/review/20326
Reviewed-by: Guy Harris <guy@alum.mit.edu>
These file formats have 16-bit packet lengths, so, even with some extra
metadata added, the packet data length will never be bigger than
WTAP_MAX_PACKET_SIZE. (No, we won't ever reduce WTAP_MAX_PACKET_SIZE.)
Change-Id: I9e1b1d90971f91cc6e5d66d0aa93841445b2bc22
Reviewed-on: https://code.wireshark.org/review/15186
Reviewed-by: Guy Harris <guy@alum.mit.edu>
in size.
Wireshark/Tshark hangs when netscaler trace file smaller than 16KB is opened.
It also hangs when a gzipped trace file is opened. With this fix,
Files with sizes that are not multiple of 16KB and gzipped files can be opened.
Bug: 12083
Change-Id: I26b2fc406edafcb2f1f6161d69064ba5662ddf29
Reviewed-on: https://code.wireshark.org/review/13721
Reviewed-by: Michael Mann <mmann78@netscape.net>
The offsetof() macro is an ANSI C library feature.
Change-Id: I2ac91b0b4c94c6f6baf14133b076fdc5ed2e182b
Reviewed-on: https://code.wireshark.org/review/12707
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>
Only this issue with a netscaler trace file
Bug: 11248
Change-Id: I2bc2cae5c988eeff7bdd08471bf421faafcd4e73
Reviewed-on: https://code.wireshark.org/review/8672
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Change-Id: Ica74f3f9239a96486967cf248feb4313bc390734
Reviewed-on: https://code.wireshark.org/review/7751
Petri-Dish: Michael Mann <mmann78@netscape.net>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Michael Mann <mmann78@netscape.net>
Create TIMEDEFV10() and, for it and for V20, V21, and V22, undefine the
ones used for the read routine and define dummy ones for the read-seek
routines (as those record formats have relative, not absolute, times).
Have all of them set presence_flags.
That way:
1) if we *can* set the time stamps in the seek-read routine, we do;
2) we always set presence_flags in read and seek-read routines.
Change-Id: I837507245e8a0cbc68c7c5fd8365f474d085488e
Reviewed-on: https://code.wireshark.org/review/7445
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Provide {FULL,PART}SIZEDEFV10 macros, similar to what's provided for the
other versions, containing the code to set the packet length and
captured data length, and use the TRACE_V10_REC_LEN_OFF() macro directly
after that to set the various length and offset values in the
pseudo-header.
Change-Id: I3513371057601b44821d89ebaa7565ab370a67f5
Reviewed-on: https://code.wireshark.org/review/7393
Reviewed-by: Guy Harris <guy@alum.mit.edu>
No need to export them outside this file, and making them constants
might convince the compiler not to bother fetching their values from
memory when referring to them.
Change-Id: Ib8605bf0bb9091721a51827c45fe75d19a15ba26
Reviewed-on: https://code.wireshark.org/review/7378
Reviewed-by: Guy Harris <guy@alum.mit.edu>
That way it'll set the captured length and reported length the same way
it's done in the read routine.
Change-Id: I8b1f2d0d8fa688f44b6f61f16dc2e21b5648fc12
Reviewed-on: https://code.wireshark.org/review/7376
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Rename the V10 PACKET_DESCRIBE()'s "fpp" argument to "type", and have it
declare the pointer variable.
Change-Id: I3ac52ebdef0aec7bc95052277537185132886b57
Reviewed-on: https://code.wireshark.org/review/7374
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Move it in the argument list, and use it instead of a hard-coded 10
(even though it's always passed as 10), to match the other
PACKET_DESCRIBE() macros.
Change-Id: Idd0a23a58cc6bb0d2de799b039db776d279cc03e
Reviewed-on: https://code.wireshark.org/review/7372
Reviewed-by: Guy Harris <guy@alum.mit.edu>
The name "type" is thrown around to mean both a version number and a
part of a structure member, and some other macros use "ver" for the
version number; try to make things a little less inconsistent.
Change-Id: I61405cf41cca43fe607154af7498944c5ec0ef11
Reviewed-on: https://code.wireshark.org/review/7370
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Makes the various flavors of that macro more similar.
Change-Id: Ie95ee8db0b226458146b055b54b7d61835f1e508
Reviewed-on: https://code.wireshark.org/review/7368
Reviewed-by: Guy Harris <guy@alum.mit.edu>
A bit of regularization.
Change-Id: I60e0bd50891e1ba3e9c40f8d1d14d63534e08138
Reviewed-on: https://code.wireshark.org/review/7364
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Use that rather than TYPE or acttype (to indicate that it's a header
version number to use as the "xxx" in NSPR_HEADER_VERSIONxxx, and to use
the same name throughout).
Change-Id: I14cfc819e44ce4388c27d521a1256dec8d80df2e
Reviewed-on: https://code.wireshark.org/review/7361
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Just like the read routine.
Change-Id: If6bd98bea15f1c8dc7454a5dac0ea57920bddc8c
Reviewed-on: https://code.wireshark.org/review/7355
Reviewed-by: Guy Harris <guy@alum.mit.edu>
That makes it even more like V2x and V3x and slightly less unclear.
Change-Id: I798ead123ba6418be0252067773a951390e26ce8
Reviewed-on: https://code.wireshark.org/review/7353
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Have it set the record type and initialize the presence bits.
That makes it a bit more like V2x and V3x, and makes the code slightly
less unclear.
Change-Id: Ibfbe2143b24a68c3fa3f576616fde81918b01feb
Reviewed-on: https://code.wireshark.org/review/7351
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Moshe Kaplan