This admittedly is a quite esoteric combination of protocols. But - as
we do correctly support them, it seems nice to have a slightly more
complete testcase that covers this.
* origin/topic/awelzel/data-end-offset-v1:
signatures: Add data_end_offset to signature_match() and custom events
Add pattern_end_offset to signature_state
This change tracks the current offset (number of bytes fed into matchers)
on the top-level RuleEndpointState such that we can compute the relative ending
for matched texts individually.
Additionally, it adds the data_end_offset as a new optional parameter to
signature_match().
This commit fixes three issues with Zeek's Modbus message logging:
1 - Some exception responses (e.g., READ_COILS_EXCEPTION) are logged
twice: once without and once with the exception message.
2 - Some exception responses (e.g., PROGRAM_484_EXCEPTION) are not
logged.
3 - Some known but reserved function codes (e.g., PROGRAM_UNITY) are
logged as unk-xxx (e.g., unk-90), while it would be possible to
log their known name.
To address these inconsistencies, the modbus parser has been updated
to parse all exception responses (i.e., all responses where the MSB
of the function code is set) using the already defined Exception
message.
Also, the Modbus main.zeek script has been updated to consistently
demand logging exception responses to the specialized
modbus_exception event, rather than logging some exception responses
in the modbus_message event and others in the modbus_exception event.
Finally, the main.zeek script has been updated to make sure that
for every known function code, the corresponding exception code was
also present, and the enumeration of known function codes in
consts.zeek has been expanded.
Closes#3984
* origin/topic/vern/CPP-when-capture-naming:
fixed "-O gen-C++" naming of "when" captures to avoid ambiguities due to inlining
Simplified & made more robust maintenance helper script for "-O gen-C++" testing
"-a cpp" baseline updates to reflect recent BTest changes
remove instance of plus sign to account for real plus in sql
account for spaces encoding to plus signs in sqli regex detection
add test cases for sqli space to plus
account for spaces encoding to plus signs in sqli regex detection
forgot semicolon
account for spaces encoding to plus signs in sqli regex detection
* topic/christian/telemetry-make-bifs-primary:
Telemetry framework: move BIFs to the primary-bif stage
Minor comment tweaks for init-frameworks-and-bifs.zeek
Adding a metric for the network time value itself should make it
possible to observe it stopping or growing slowly as compared to
realtime when Zeek isn't able to keep up.
Also, modify the telemetry/log.zeek test to include misc/stats and
log at a higher frequency with a more interesting pcap.
This stops invoking Telemetry::sync() via a scheduled event and instead
only invokes it on-demand. This makes metric collection network time
independent and lazier, too.
With Prometheus scrape requests being processed on Zeek's main thread
now, we can safely invoke the script layer Telemetry::sync() hook.
Closes#3947
This commit prevents most non-Modbus TCP traffic on port 502 to be
reported as Modbus in conn.log as well as in modbus.log.
To do so, we have introduced two &enforce checks in the Modbus
protocol definition that checks that some specific fields of the
(supposedly) Modbus header are compatible with values specified in
the specs.
To ensure non-regression, with this commit we also introduce a
new btest.
Closes#3962
This moves the Telemetry framework's BIF-defined functionalit from the
secondary-BIFs stage to the primary one. That is, this functionality is now
available from the end of init-bare.zeek, not only after the end of
init-frameworks-and-bifs.zeek.
This allows us to use script-layer telemetry in our Zeek's own code that get
pulled in during init-frameworks-and-bifs.
This change splits up the BIF features into functions, constants, and types,
because that's the granularity most workable in Func.cc and NetVar. It also now
defines the Telemetry::MetricsType enum once, not redundantly in BIFs and script
layer.
Due to subtle load ordering issues between the telemetry and cluster frameworks
this pushes the redef stage of Telemetry::metrics_port and address into
base/frameworks/telemetry/options.zeek, which is loaded sufficiently late in
init-frameworks-and-bifs.zeek to sidestep those issues. (When not doing this,
the effect is that the redef in telemetry/main.zeek doesn't yet find the
cluster-provided values, and Zeek does not end up listening on these ports.)
The need to add basic Zeek headers in script_opt/ZAM/ZBody.cc as a side-effect
of this is curious, but looks harmless.
Also includes baseline updates for the usual btests and adds a few doc strings.
This isn't a straightforward fix, unfortunately. The existing GetLine()
implementation didn't deal well with input that's incrementally produced
where individually read chunks wouldn't end with the separator.
The prior implementation increased the buffer each time it failed to find
a separator in the current buffer, but then also ended up not searching the
full new buffer size for the terminator, doing that endlessly.
This change reworks the Raw reader to rely only on bufpos for reading
and searching purposes and skip reallocation if the buffer size if it
wasn't actually exhausted.
Closes#3957
Processing out-of-order commands or finishing commands based on invalid
server responses resulted in inconsistent analyzer state, potentially
triggering null pointer references for crafted traffic.
This commit reworks cf9fe91705 such that
too many pending commands are simply discarded, rather than any attempt
being made to process them. Further, invalid server responses do not
result in command completion anymore.
Test PCAP was crafted based on traffic produced by the OSS-Fuzz reproducer.
Closes#215
That test got flaky probably from #3949 on centosstream9 CI. You can
replicate that behavior by increasing the sleep time when waiting for
the file such that the test will attempt to read the missing file again.
Since the one second wait for file is glacially slow for this, speeding
it up should mean that the file gets created sooner and so the test
won't try to open the file again. But, it's always still technically
possible, since the test will wait for 10 seconds and the heartbeat
seems to be 1 second. At least if that happens, it's probably a bug or
massive slowdown of some kind.
It seems like other similar tests get by because they have more "stuff"
before they call `terminate()` most likely. But, to be safe, just
removing the "received termination signal" line seems like the best
approach.
Invalid lines in a file was the one case that would not suppress future
warnings. Just make it suppress warnings too, but clear that suppression
if there is a field in between that doesn't error.
Fixes#3692
Log flushing is currently triggered based on the threading heartbeat timer
of WriterBackends and the hard-coded WRITE_BUFFER_SIZE 1000.
This change introduces a separate timer that is managed by the logger
manager instead of piggy-backing on the heartbeat timer, as well as a
const &redef for the buffer size.
This allows to modify the log flush frequency and batch size independently
of the threading heartbeat interval. Later, this will allow to re-use the
buffering and flushing logic of writer frontends for non-Broker cluster
backends, too.
One change here is that even frontends that do not have a backend will
be flushed regularly. This is wanted for non-Broker backends and should be
very cheap. Possibly, Broker can piggy back on this timer down the road, too,
rather than using its own script-level timer (see Broker::log_flush()).
The cmds list may grow unbounded due to the POP3 analyzer being in
multiLine mode after seeing `AUTH` in a Redis connection, but never
a `.` terminator. This can easily be provoked by the Redis ping
command.
This adds two heuristics: 1) Forcefully process the oldest commands in
the cmds list and cap it at max_pending_commands. 2) Start raising
analyzer violations if the client has been using more than
max_unknown_client_commands commands (default 10).
Closes#3936
Really, they both should be count. But, they were getting provided as an
integer. Port is easy since it is backed by an unsigned value. Enums
*should* be unsigned, but aren't. This doesn't address that, it just
takes the other name for this operator (absolute value) and makes the
enum value positive if it's negative.
This fixes a case where using the size of operator on enum/port values
in certain contexts (like the default parameter of a struct) would cause
an internal error.
