In the case we exceed the number of simultaneously open
file we can reach a state were we will not close the file
after writing.
Thanks to Steve Grubb <sgrubb@redhat.com> for the analysis.
Fix mpm progress being updated by irrelevant engines. Esp in the
case of file_data engines, signature can contain multiple versions
of the same engine, registered for different 'progress' values.
This would lead to signatures being considered 'can't match' even
in cases where they clearly could still match.
Only consider those progress values that apply to the protocol in
use.
If the data file can't be found in the default location, which
normally is 'default-rule-path', try to see if it can be found
in the path of the rule file that references it.
This makes QA much easier.
If the global detect.prefilter.default setting is not "auto", it is
wasteful to run each prefilter setup routine. This patch tracks which
of the engines have been explicitly enabled in the rules and only
runs those.
When prefilter is not enabled globally, it is still possible to
enable it per signature. This was broken however, as the setup
code would never be called.
This commit always call the setup code and lets that sort out
which signatures (if any) to enable prefiltering for.
If prefilter is used on a content keyword, it acts as a simple
fast_pattern statement. This was broken because the SIG_FLAG_PREFILTER
flag bypasses MPM for a sig. This commits fixes this by not setting
the flag when it should act as fast_pattern.
The HTTP bodies (http_client_body and http_server_body/file_data) use
settings to control how much data we have before doing first inspection:
request-body-minimal-inspect-size
response-body-minimal-inspect-size
These settings default to 32k as quite some existing rules need this.
At the same time, the 'raw stream' inspection uses its own limits. By
default it inspects the data in blocks of about 2.5k. This could lead
to a situation where rules would not match.
For example, with 2 rules like this:
content:"abc"; content:"data="; http_client_body; depth:5; sid:1;
content:"xyz"; sid:2;
Sid 1 would only be inspected when the POST body reached the 32k limit
or when it was complete. Observed case shows the POST body to be 18k.
Sid 2 is inspected as soon as the 2.5k limit is reached, and then again
for each 2.5k increment. This moves the raw stream tracker forward.
So by the time sid 1 is inspected, some 18/19k into the stream, the
raw stream tracker is actually already moved forward for approximately
17.5k, this leads to the stream match of sid 1 possibly not matching.
Since the body match is at the start of the buffer, it makes sense
that the body and stream are inspected together.
The body inspection uses a tracker 'body_inspected', that keeps track
of how far into the body both MPM and per signature inspection has
moved.
This patch updates the logic in 2 ways:
1. it triggers earlier HTTP body inspection, which is matched to the
stream inspection. When the detection engine finds it has stream
data available for inspection, it passes the new 'STREAM_FLUSH'
flag to the HTTP body inspection code. Which will then do an
early inspection, even if still before the min inspect size.
2. to still somewhat adhere to the min inspect size, the body
tracker is not updated until the min inspect size is reached.
This will lead to some re-evaluation of the same body data.
If raw stream reassembly is disabled, this 'STREAM_FLUSH' flag is
never set, and the old behavior is used.
Bug #2522.
Trigger app layer reassembly in both directions as soon as we've set
the TCP state to closed.
In IDS mode, if a toserver packet would close the state, the app layer
would not get updated until the next toclient packet. However, in
detection, the raw stream inspection would already use all available
stream data in detection and move the 'raw stream progress' tracker
forward. When in later (a) packet(s) the app layer was updated and
inspection ran on the app layer, the stream progress was already
moved too far forward. This would lead to signatures that matched
on both stream and app layer to not match.
By triggering the app layer reassembly as soon as the TCP state is
set to closed, the inspection as both the stream and app layer data
available at the same time so these rules can match.
Bug: #2570
Bug: #2554
If a stream is async we see only on side of the traffic. This would
lead to the flow engine not flagging packets as 'established' even
if the flow state was in fact established. The flow was tagged as
such by the TCP engine.
This patch considers the flow state for setting the packet flag.
Bug #2491.
When a file in TOSERVER direction is being stored and
libhtp or stream depth limit is reached,
it will be closed by HTPCallbackRequest without setting
any flags so the file state will be set to CLOSED
instead of TRUNCATED.
When an app-layer parser is enabled, it could set its
own stream_depth value calling the API AppLayerParserSetStreamDepth.
Then, the function AppLayerParserPostStreamSetup will replace
the stream_depth value already set with stream_config.reassembly_depth.
To avoid overwriting, in AppLayerParserSetStreamDepth API a flag
will be set internally to specify that a value is already set.
Add device to tenant mapping support:
mappings:
- device: ens5f0
tenant-id: 1
- device: ens5f1
tenant-id: 23
Implemented by assigning the tenant id to the 'livedev', which means
it's only supported for capture methods that use the livedev API.
It's also currently not supported for IPS. In a case like 'eth0 -> eth1'
it's unclear which tenant should be used for the return traffic in a
flow, where the incoming device is 'eth1'.
Change the decode handler signature to increase the size of its decode
handler, from uint16 to uint32. This is necessary to let suricata use
interfaces with mtu > 65535 (ex: lo interface has default size 65536).
It's necessary to change several primitive for Packet manipulation, to
unify the parameter "packet length" whenever we are before IP decoding.
Add tests before calling DecodeIPVX function to avoid a possible
integer overflow over the len parameter.
fix bug in fileext and filename preventing negated match to work
correctly. Previously, negated fileext (such as !"php") would cause a
match anyway on files that have extension php, as the last if would not
be accessed.
Using the same workflow as detect-filemagic we remove the final
isolated if and set it as a branch of the previous if.
Substitute json_string with SCJsonString custom function.
SCJsonString will ensure string passed is json-encodable (utf-8).
If it's not, the string will be converted in such a way that any
non-printable character will be encoded in its hex form.
The resulting json object will be returned.
rust modification will encode any non-printable character during its
conversion in to_cstring.
4474889667 did break the logic of IPS
mode detection for layer 2 IPS mode like Netmap and AF_PACKET. This
pach fixes the issue by calling the IPS detection mode after
complete registration of the interfaces.
Reported-by: Alexander Gozman <a.gozman@securitycode.ru>
If a signature didn't explicitly specified 'dcerpc' or 'smb' as the
app proto, false positives on other traffic could happen. This was
caused by the sig not having a app_proto set. This isn't set as the
rule is supposed to match against either ALPROTO_DCERPC or ALPROTO_SMB.
To avoid adding runtime costs for checking for both protocols, this
patch adds a new flag for DCERPC in the 'mask' logic. The flag is set
on the sig if dce_* keywords are present and set on the packet if the
flow's app proto is either ALPROTO_DCERPC or ALPROTO_SMB.
Bug #2559
Reported-by: Jason Taylor
In case of a valid RST on a SYN, the state is switched to 'TCP_CLOSED'.
However, the target of the RST may not have received it, or may not
have accepted it. Also, the RST may have been injected, so the supposed
sender may not actually be aware of the RST that was sent in it's name.
In this case the previous behavior was to switch the state to CLOSED and
accept no further TCP updates or stream reassembly.
This patch changes this. It still switches the state to CLOSED, as this
is by far the most likely to be correct. However, it will reconsider
the state if the receiver continues to talk.
To do this on each state change the previous state will be recorded in
TcpSession::pstate. If a non-RST packet is received after a RST, this
TcpSession::pstate is used to try to continue the conversation.
If the (supposed) sender of the RST is also continueing the conversation
as normal, it's highly likely it didn't send the RST. In this case
a stream event is generated.
Ticket: #2501
Reported-By: Kirill Shipulin
https://redmine.openinfosecfoundation.org/issues/2417
Add option to have pcap files deleted after they have been processed.
This option combines well with pcap file continuous and streaming
files to a directory being processed.
Calls with both START and MIDSTREAM mean the record might be cut and the
start of it could be missing. For this case, enable the same logic as is
used when catching up after a GAP. Search for the start of the record
instead of assuming it sits exactly at the start of the input data.
Last multi-detect changes broken delayed-detect by refusing to reload
a 'stub' detect engine. This patch distinguishes between a stub for
multi-tenancy and for delayed detect.
There are 3 types of detect engine objects:
1. normal
The normal detection engine if no multi-tenancy is in use
2. tenant
A per tenant detection engine
3. stub
A stub (or minimal as it was called before) detect engine
that is needed to have something in place when there are
only tenants.
A stub is also used in case of 'delayed detect', where we
need a minimal detect engine to start up which is replaced
by a full (normal type) detect engine after startup.
This patch adds a new field 'type' to the DetectEngineCtx object
to distinguish between the types. This replaces the boolean 'minimal'.
The global keyword registration and per thread init handling used
the lock from the DetectEngineMasterCtx. This lead to a dead lock
situation at multi-tenancy tenant reloads.
The lock was unnecessary however, as the only time the registration
list is updated is at engine initialization. At that time Suricata
is still running in a single thread. After this, the data structure
doesn't change anymore.
Bug #2516.
Enables IPS functionality on Windows using the open-source
(LGPLv3/GPLv2) WinDivert driver and API.
From https://www.reqrypt.org/windivert-doc.html : "WinDivert is a
user-mode capture/sniffing/modification/blocking/re-injection package
for Windows Vista, Windows Server 2008, Windows 7, and Windows 8.
WinDivert can be used to implement user-mode packet filters, packet
sniffers, firewalls, NAT, VPNs, tunneling applications, etc., without
the need to write kernel-mode code."
- adds `--windivert [filter string]` and `--windivert-forward [filter
string]` command-line options to enable WinDivert IPS mode.
`--windivert[-forward] true` will open a filter for all traffic. See
https://www.reqrypt.org/windivert-doc.html#filter_language for more
information.
Limitation: currently limited to `autofp` runmode.
Additionally:
- `tmm_modules` now zeroed during `RegisterAllModules`
- fixed Windows Vista+ `inet_ntop` call in `PrintInet`
- fixed `GetRandom` bug (nonexistent keys) on fresh Windows installs
- fixed `RandomGetClock` building on Windows builds
- Added WMI queries for MTU
Much of encrypted traffic is uninteresting to Suricata. Once encrypted
communication starts, inspecting the packet payloads is generally
not interesting anymore. The default behavior is to disable the parts
of the detection engine and stream reassembly that relate to raw content
inspection.
The tls app-layer parser also had a crude option to affect this behavior:
set 'no-reassemble' to true went much further than the default behavior.
It disabled the TCP reassembly on the flow completely, disabled all
inspection on the flow and enabled bypass if available.
This patch adds a new option: full inspection. This continues to treat
a TLS session as any other, so without any limits to inspection.
The new option is implemented in a new config option 'encrypt-handling',
that replaces 'no-reassemble'. The new option has 3 values:
'default', 'full' and 'bypass'. Default is the current default behavior,
'bypass' is the current 'no-reassemble = true' behavior and 'full'
is the new full inspection mode.
The use of stream_size in combination with raw content matches is an
indication that the rule needs to be evaluated per packet, not just
per reassembled stream chunk.
This is a DHCP decoder and logger written in Rust. Unlike most
parsers, this one is stateless so responses are not matched
up to requests by Suricata. However, the output does contain
enough fields to match them up in post-processing.
Rules are included to alert of malformed or truncated options.
Before setting up a sub eve-logger, check that it is enabled. This
allows us to set "enabled: no" for loggers that are not registered
with the system without generating an error. An example of this
is loggers that are only available with Rust.
getrandom syscall availability is detected at runtime. So it is
possible that the build is done on a box that supports it but
the run is done on a system with no availability. So a workaround
solution is needed to fix this case.
Also we have seen some issue in docker environment where the build
is detecting getrandom but where it does not work at runtime.
For both reasons, the code is updated to have a call to a fallback
function if ever the getrandom call returns that the syscall is
not available.
Fix Coverity issue:
** CID 1435535: Null pointer dereferences (REVERSE_INULL)
/src/output-json-file.c: 212 in JsonBuildFileInfoRecord()
Where we check a variable for being NULL, when all paths to the
code show that it can't be NULL.
XFF configuration is already set in app-layer-htp-xff, and in
output-json-alert. Extending XFF configuration to files and HTTP allow
to get the same behavior as for alerts.
Extend the configuration of filestore json to let filestore metafile
dump be aware of xff. This is available only if write-fileinfo is set
to yes and file-store version is 2.
When a file is transferred over anything other than HTTP, the previously hard-coded HTTP protocol would trigger a non-existent index into htp_list_array_get(), causing a segfault. This patch mimics the logic in detect-lua-extensions.c.
Validate that the content that follows the 'tls_cert_serial' keyword
is on the correct form. If it's longer than two bytes it should be
separated by colons.
Introduces the option 'outputs.pcap-log.compression' which can be set
to 'none' or 'lz4', plus options to set the compression level and to
enable checksums. SCFmemopen is used to make pcap_dump() write to a
buffer which is then compressed using liblz4.
When traffic is becoming null (mainly seen in tests) we reach the
situation where there is timeouts in the poll on the socket and
only that. Existing code is then just looping on the poll and
the result is that the packet iface counters are not updated.
This patch calls the dump counter function to be sure to get
the counter right faster (and not only right at exit).
The stream depth computation was partly done with the stream_config
depth instead of using the value in the TCP session. As a result,
some configuration were resulting in abnormal behavior.
In particular, when stream depth was 0 and the file store depth was
not 0, Suricata was stopping the streaming on the flow as soon as
the filestore was started.
Reported-by: Pascal Delalande <pdl35@free.fr>
There is a difference in the size of the buffer length as passed from
the content buffers (cfr HttpReassembledBody.buffer_len) and the buflen
variable passed to mpm primitives. This can cause a misdetection
whenever the bufferlen is multiple of 65536 (as uint16(X*65536) == 0).
Increasing the buflen variable type to uint32 solves the issue (this
does not cause any issue with primitives, they all accept uint32).
If a file transfer stops on flow timeout, it won't be closed or
truncated. This patch makes sure that in such cases the files
are indeed truncated. This fixes the filestore-v2 output module,
as that requires a sha256 for storing the partial file correctly.
CID 1428797 (#1 of 1): Unchecked return value (CHECKED_RETURN)
check_return: Calling HashTableAdd without checking return value
(as is done elsewhere 5 out of 6 times).
Don't treat 'external' parsers as more experimental. All parsers
depend on crates to some extend, and all have C glue code. So the
distinction doesn't really make sense.
Add a new parser for Internet Key Exchange version (IKEv2), defined in
RFC 7296.
The IKEv2 parser itself is external. The embedded code includes the
parser state and associated variables, the state machine, and the
detection code.
The parser looks the first two messages of a connection, and analyzes
the client and server proposals to check the cryptographic parameters.
https://redmine.openinfosecfoundation.org/issues/2451
When a missing (or empty named) file is passed to source-pcap-file while
using unix socket, the pcap processing thread will incorrectly be stopped,
and no longer available for subsequent files.
As an optimization, reset bidirectional flag for rules with same src and dst.
If one created bidirectional rule like 'alert tcp any any <> any any ...',
the rule was checked twice (for each packet in every direction). This is
suboptimal and may give duplicated alerts. To avoid this, bidirectional
rules are now checked for the same src and dst (addresses and ports) and
if it's the case, the rule is treated as unidirectional and a corresponding
message is logged.
Add Ja3GetString() to return the content of the JA3 string buffer from the
TLS session.
Example:
function init (args)
local needs = {}
needs["protocol"] = "tls"
return needs
end
function setup (args)
filename = SCLogPath() .. "/ja3_string.log"
file = assert(io.open(filename, "a"))
end
function log (args)
ja3_string = Ja3GetString()
if ja3_string == nil then
return
end
file:write(ja3_string .. "\n")
file:flush()
end
function deinit (args)
file:close()
end
Add Ja3GetHash() to return the content of the JA3 hash buffer from the
TLS session.
Example:
function init (args)
local needs = {}
needs["protocol"] = "tls"
return needs
end
function setup (args)
filename = SCLogPath() .. "/ja3_hash.log"
file = assert(io.open(filename, "a"))
end
function log (args)
ja3_hash = Ja3GetHash()
if ja3_hash == nil then
return
end
file:write(ja3_hash .. "\n")
file:flush()
end
function deinit (args)
file:close()
end
In the (useless) example above, each JA3 hash is logged to a log file.
Match on JA3 hash using ja3_hash keyword, e.g:
alert tls any any -> any any (msg:"JA3 hash test";
ja3_hash;
content:"e7eca2baf4458d095b7f45da28c16c34";
sid:1;)
Eric Leblond
Victor Julien
Maurizio Abba
Giuseppe Longo
Danny Browning
Jason Ish
Jacob Masen-Smith
fooinha
Pierre Chifflier
Elazar Broad
Mats Klepsland
Max Fillinger
Alexander Gozman
Antoine LUONG
Antti Tönkyrä