If the *first* read for a packet gets an EOF, it means that there is an
EOF right at the point where you're reading, which means "no more
packets".
If you get an EOF on any *subsequent* reads for the packet, it means the
file was cut off in the middle of the packet's record, which is an error.
ws_debug() inserts the file name, line number, and function name into
the ws_debug() message (assuming the function name can be obtained from
a macro), so there's no need to include it in the text of the message
(we don't do so elsewhere).
This has a few effects on the behavior of wtap_get_compression_type()
and wtap_get_all_compression_type_extensions():
Make capinfos correctly report the compression type (instead of
saying gzip compressed for zstd and lz4 compressed files).
Makes files with the .zstd and .lz4 extension show up in the file
chooser when "Files of type" is set to something other than "All Files",
such as "All Capture Files" or "Wireshark/... pcapng"
Makes the UI not default to gzip compression when saving a file
compressed as zstd or lz4 (write support for zstd and lz4 doesn't
exist yet, and the GUI doesn't have hooks for it anyway, though
this can help as a prerequisite for later support for writing.)
Also replace a couple of assert() with ws_assert().
Update the PURPOSE in CMakeLists for zstd and lz4 to note that they
can be used to read compressed capture files.
The Ubuntu build commented on some spelling errors in executable code
files. Fix the errors that don't come from external files containing
the spelling errors (USB product and vendor IDs, PCI IDs, ASN.1
specifications), and fix some errors that don't show up in the
executable code files (e.g., in comments and variable names).
Use the same style of message for too-short block errors ("pcapng: total
block length XXX of {a,an} XXX is too small...").
Add an additional check for the "skip" Netflix cutom block, to make sure
it has enough room for the 4-byte "skipped" value.
wblock->internal is not initialized on pcapng_read_custom_block function
pcapng.c:3747:9: warning: Branch condition evaluates to a garbage value [core.uninitialized.Branch]
nettrace_3gpp_32_423.c:256:2: warning: Value stored to 'prev_pos' is never read [deadcode.DeadStores]
nettrace_3gpp_32_423.c:295:2: warning: Value stored to 'next_msg_pos' is never read [deadcode.DeadStores]
nettrace_3gpp_32_423.c:487:4: warning: Value stored to 'port_type_defined' is never read [deadcode.DeadStores]
Use compute_options_size() to get the total size of all the options, and
use write_options() to write out the options for those blocks, as we do
for other blocks.
Get rid of wtap_block_option_get_value_size() and
wtap_block_get_options_size_padded(); they're no longer needed, and
their notion of an option's "size" is "size in a pcapng file", so that
doesn't belong in code that's intended to support all file types.
Have a routine to read the Sniffer record header, and call that in
ngsniffer_read() and ngsniffer_seek_read(). Only call
ngsniffer_process_record() for frame records that we understand, so that
we only allocate a block for those rather than for records we don't
understand or for EOF records, potentially leaking them.
Without that, you could add a comment to a record in a file format the
reading code for which doesn't allocate blocks, but the comment doesn't
get saved, as there's no block in which to save the comment option.
This simplifies some code paths, as we're either using the record's
modified block or we're using the block as read from the file, there's
no third possibility.
If we attempt to read a record, and we get an error, and a block was
allocated for the record, unreference it, so the individual file readers
don't have to worry about it.
Move the extraction of the option value from the option content from the
callers of pcapng_process_uint32_option() to
pcapng_process_uint32_option() itself.
That way, add-on modules to handle block types not handled by the core
pcapng code can use pcap_process_options() to process a block's options
and can use the routines to handle the "standard" option value types to
handle particular options.
Also, allow both everything-is-little-endian and
everything-is-big-endian Custom Block types in pcap_process_options().
Have wsutil/exported_pdu_tlvs.h define the LINKTYPE_WIRESHARK_UPPER_PDU
TLV type and length values, as well as the port type values written to
files in EXP_PDU_TAG_PORT_TYPE TLVs.
Update the comment that describes the LINKTYPE_WIRESHARK_UPPER_PDU TLVs
to more completely and correctly reflect reality (it was moved from
epan/exported_pdu.h to wsutil/exported_pdu_tlvs.h).
Rename those port type values from OLD_PT_ to EXP_PDU_PT_; there is
nothing "old" about them - yes, they originally had the same numerical
values as the PT_ enum values in libwireshark, but that's no longer the
case, and the two are now defined independently. Rename routines that
map between libwireshark PT_ values and EXP_PDU_PT_ values to remove
"old" from the name while we're at it.
Don't include epan/exported_pdu.h if we only need the
LINKTYPE_WIRESHARK_UPPER_PDU definitions - just include
wsutil/exported_pdu_tlvs.h.
In extcap/udpdump.c, include wsutil/exported_pdu_tlvs.h rather than
defining the TLV types ourselves.
Bug 17478 was caused by `wtap_rec.block` being allocated for each
packet, but not freed when it was done being used -- typically at the
end of a loop.
Rather than requiring each caller of `wtap_read()` to know to free a
member of `rec`, I added a new function `wtap_rec_reset()` for a
slightly cleaner API. Added calls to it everywhere that seemed to make
sense.
Fixes#17478
Initialize name_str right before we fill it in if the name is present,
and note what we're doing, to make it a bit clearer what the
initializing code is doing.
Erik Rigtorp