As is done for other targets,
so that all app-layer protocols are enabled,
even the ones disabled by default such as enip
And resets protocol detection every time we try
so that probing_parser_toserver_alproto_masks are fresh.
Implement a special sticky buffer to select frames for inspection.
This keyword takes an argument to specify the per protocol frame type:
alert <app proto name> ... frame:<specific frame name>
Or it can specify both in the keyword:
alert tcp ... frame:<app proto name>.<specific frame name>
The latter is useful in some cases like http, where "http" applies to
both HTTP and HTTP/2.
alert http ... frame:http1.request;
alert http1 ... frame:request;
Examples:
tls.pdu
smb.smb2.hdr
smb.smb3.data
Consider a rule like:
alert tcp ... flow:to_server; content:"|ff|SMB"; content:"some smb 1 issue";
this will scan all toserver TCP traffic, where it will only be limited by a port,
depending on how rules are grouped.
With this work we'll be able to do:
alert smb ... flow:to_server; frame:smb1.data; content:"some smb 1 issue";
This rule will only inspect the data portion of SMB1 frames. It will not affect
any other protocol, and it won't need special patterns to "search" for the
SMB1 frame in the raw stream.
The idea of stream frames is that the applayer parsers can tag PDUs and
other arbitrary frames in the stream while parsing. These frames can then
be inspected from the rule language. This will allow rules that are more
precise and less costly.
The frames are stored per direction in the `AppLayerParserState` and will only
be initialized when actual frames are in use. The per direction storage has a
fixed size static portion and dynamic support for a larger number. This is done
for effeciency.
When the Stream Buffer slides, frames are updated as they use offsets relative
to the stream. A negative offset is used for frames that started before the
current window.
Frames have events to inspect/log parser errors that don't fit the TX model.
Frame id starts at 1. So implementations can keep track of frame ids where 0
is not set.
Frames affect TCP window sliding. The frames keep a "left edge" which
signifies how much data to keep for frames that are still in progress.
Pruning of StreamBufferBlocks could remove blocks that fell entirely
after the target offset due to a logic error. This could lead to data
being evicted that was still meant to be processed in theapp-layer
parsers.
Bug: #4953.
This commit makes a small optimization when comparing IPv4 and IPv6
addresses by making the host order value invariant and calculating the
value once, before entering the loop.
Explicitly truncate file names to UINT16_MAX
Before, they got implicitly truncated, meaning a UINT16_MAX + 1
file name, went to 0 file name (because of modulo 65536)
This commit adds a signal handler for SIGSEGV when configured. The
signal handler emits a one line stack trace using SCLogError. The intent
is to provide diagnostic information in deployments where core files are
not possible.
The diagnostic message is from the offending thread and includes the
stack trace; each frame includes the symbol + offset.
Ticket: 4920
Completes commit c8dbe24fb6
which introduced AppProtoEquals to have a generic
check for http in signature can mean http1 or http2 in
traffic.
This commit missed this case, as I only looked for
git grep "alproto ==" and here we deal with
alproto_tc and alproto_ts, but not alproto by itself.
It appears that DNS servers will still process a DNS request even if the
z-bit is set, our parser will fail the transaction. So create the
transaction, but still set the event.
Ticket #4924
Set RSS hash function according to Intel ICE PMD available hash functions
Set hash functions according to the support by the ICE PMD, so that no warning
regarding RSS setting is issued.
Set RSS hash function according to Intel IXGBE PMD available hash functions.
During configuration, a warning appeared stating that RSS hash function
has been changed from one value to the other. This has meant that
the supported hash functions did not cover all required hash functions
by the configuration. This commit solves the warning.
Due to peculiar behavior of i40e PMD driver, the RSS is required to be set
via rte_flow rules or a hash filter as compared to other NICs where RSS is
configured through port configuration structure.
RTE_FLOW rules are created on 5-tuples (as opposed to 3-tuple configured
on the other NICs). Fragmented traffic have been tested with this setup
and it has been proven that fragmented packets of the same flow are
received on the same queue. At the same time, setting 3-tuple on rte_flow
rules have not yield in the expected results.
Notes from the experiments:
- Configuration of 5-tuple (as is in the commit):
fragmented and nonfragmented packets are received by the same workers
even when I applied seed to alter them via tcpreplay-edit (option --seed)
- Setting only ETH_RSS_FRAG_IPV4 and ETH_RSS_IPV4 (i.e. setting 3-tuple):
when setting ETH_RSS_IPV4, the PMD driver says that pctype is not
supported (generally this means that the "type" of traffic is not
a valid configuration for the i40e)
- Setting only ETH_RSS_FRAG_IPV4 and ETH_RSS_NONFRAG_IPV4_OTHER:
this doesn't work well, packets of the same flow are received on
the different workers (my explanation is that the fragmented packets are
matched with ETH_RSS_FRAG_IPV4 but the other UDP packets are not matched
with ETH_RSS_NONFRAG_IPV4_OTHER rte_flow rule (they would be matched with
ETH_RSS_NONFRAG_IPV4_UDP).
Register a new runmode - DPDK. This enables a new flag on Suricata start
(--dpdk).
With the flag given, DPDK runmode is enabled.
Runmode loads the configuration and then initializes EAL.
If successful, it configures the physical NICs according to the configuration
file. After that, worker threads are initialized and then are in continuous
receive loop.
Ticket: 4857
If a pattern such as GET is seen ine the beginning of the
file transferred over ftp-data, this flow will get recognized
as HTTP, and a HTTP state will be created during parsing.
Thus, we cannot override directly alproto's values
This solves the segfault, but not the logical bug that the flow
should be classified as FTP-DATA instead of HTTP
Instead of storing a name and description as a pointer in DetectBufferType
store them in fixed size arrays. This is in preparation of runtime registration
of buffer types, where a constant name/desc is not available.
In preparation of more dynamic logic in rule loading also doing
some registration, allow for buffers to be registered as fast_patterns
during rule parsing.
Leaves the register time registrations mostly as-is, but copies the
resulting list into the DetectEngineCtx and works with that onwards.
This list can then be extended.
Instead of a map that is constantly realloc'd, use 2 hash tables for
DetectBufferType entries: one by name (+transforms), the other by
id. Use these everywhere.
There is a hack to know the type of an integer
and do an explicit cast in the python script
generating the C file
Also extends some bounds check against negative values
Update the packet payload after decode, instead of during IPS send.
This means the updates happen in the capture thread, and the VLAN header
is available to logging as well.
Ticket: #4805.
Ticket: #4796.
V2 (for IDS and IPS) and V3 (for IDS) are widely supported. V2 was introduced
in 2008, so we can safely assume that all systems can run V2+.
314ec77f88 had the unintended side affect
of enabling the template parser and detection buffer if unittests were
enabled.
Fix this by using the new `Default` method for registering parsers.
However, the buffer still needs an explicit configuration check.
Also convert Notice debug messages to Debug to reduce output when in
unittest mode. If we feel stronly this should still be Notice in the
template, that is a conversion we can make in the generation script when
generating a new parser.
If a HTTP2 transaction gets freed before the end of the range
request, we need to have the files container which is in
the state, to transfer owernship of this file to the files
container.
Ticket: 4811
The SMTP transaction logger was not writing the log if the email
portion of the logger failed, such as in the case of STARTTLS
where this is no email decoded.
Ticket #4817
Every transaction has an existing mandatory field, tx_data. As
DetectEngineState is also mandatory, include it in tx_data.
This allows us to remove the boilerplate every app-layer has
for managing detect engine state.
Rules profiling was returning invalid results when used with sample
rate. The problem was that the sample condition was run twice in the
packet flow. As a result, the second pass was not initializing the
variable storing the initial CPU ticks and the resulting performance
counters were reporting invalid values.
Bug: #4836.
Ticket: #4569
If a FIN+SYN packet is sent, the destination may keep the
connection alive instead of starting to close it.
In this case, a later SYN packet will be ignored by the
destination.
Previously, Suricata considered this a session reuse, and thus
used the sequence number of the last SYN packet, instead of
using the one of the live connection, leading to evasion.
This commit errors on FIN+SYN so that they do not get
processed as regular FIN packets.
Special handling for RST packets if they have an TCP MD5 or AO header option.
The options hash can't be validated. The end host might be able to validate
it, as it can have a key/password that was communicated out of band.
The sender could use this to move the TCP state to 'CLOSED', leading to
a desync of the TCP session.
This patch builds on top of
843d0b7a10 ("stream: support RST getting lost/ignored")
It flags the receiver as having received an RST and moves the TCP state
into the CLOSED state. It then reverts this if the sender continues to
send traffic. In this case it sets the following event:
stream-event:suspected_rst_inject;
Bug: #4710.
Ticket: #4562
As the data which triggered the opposing side
was the same protocol and not another one,
that means the protocol change failed.
Prevents a memory leak in later call of AppLayerParserParse
which would allocate a new state and leak the old one
When InspectionBufferGet gets called with base_id
Later InspectionBufferSetup must also be called with base_id
In case there were transforms, we had base_id != list_id
Not calling InspectionBufferSetup with the right id
resulted in leaving a dangling pointer,
because it was not added to det_ctx->inspect.to_clear_queue
Bug: #4681.
Flows have been shown to linger for a long time w/o giving up their
resources. This would lead to higher memory use and memcaps getting
reached.
Three main causes have been identified:
Slow passes hash passes. By default the flow manager will scan the
flow hash slowly. It is based on the flow timeout settings, and with
the default config it will take 4 minutes for a full scan to be
complete. This leaves a window for flows that are timed out to linger
for minutes longer than expected.
Flow Manager yields under pressure. The per row TryLock causes work
to be delayed more. The Flow manager will use trylock on a hash row
and will yield immediately if the row is busy. This means that it will
take a full pass before the row is revisited again. If the row holds
busy flows, this could happen many times in a row.
Flow Manager favors evicted flows over active flows. The Flow Manager
will only process the evicted flows if they are present. These flows
have been evicted by workers. The active flows on that hash row will
have to wait until the next hash pass. Of course by then there could
be more evicted flows.
Combined these factors could lead to flows not being considered for
freeing and logging for a very long time, potentially even indefinitly.
The patch addresses the latter two flow manager issues by no longer
using TryLock. It will now simply wait for the lock to be released and
then do its work on it. Additionally for each row both the evicted list
and the active flow list will be processed.
Bug: #4650.
PacketSetData() can't fail unless the input pointer is NULL, which is
impossible from the af-packet paths calling it. Remove error check to
avoid possible branching.
The AFPSwitchState function would close the socket and free the
other resources when the interface went down _and_ the ref cnt was
0. However in autofp mode it was common to get to this point while
packets were still processed in the autofp worker threads, meaning
the ref cnt would not be 0. On the interface coming back up the
initialization code would overwrite the socket and rings, leading
to resource leaks.
Socket ref cnt is decremented from the v2 release callback. If the
callback would get to ref cnt 0, the packet would not be released
in the kernel, but it would (possibly) close the socket if the
iface was down, but not free other resources.
This patch changes the logic to first release the packet to the
kernel and then decrement the ref cnt and it makes the main receive
loop the only one responsible for opening and closing sockets. Wait
with closing the socket and rings until the ref count is 0, which can
happen after AFPSwitchState is called due to packets still being
processed by autofp worker threads.
Bug: #4803.
Tpacket v3 only supports workers mode, which means the packet that would
reference a socket won't leave the thread. Therefore keeping a ref count
on the socket is not needed.
This patch removes the per packet reference count increment. The decrement
was missing, so this fixes the ref cnt handling so that after a iface up/
down capture can recover.
It should also lead to a minor performance increase as we avoid a round
of atomic operations per packet.
Bug: #4804.
Bug: #4801.
The Suricata AF_PACKET code opens a socket per thread, then after some minor
setup enters a loop where the socket is poll()'d with a timeout. When the
poll() call returns a non zero positive value, the AF_PACKET ring will be
processed.
The ringbuffer processing logic has a pointer into the ring where we last
checked the ring. From this position we will inspect each frame until we
find a frame with tp_status == TP_STATUS_KERNEL (so essentially 0). This
means the frame is currently owned by the kernel.
There is a special case handling for starting the ring processing but
finding a TP_STATUS_KERNEL immediately. This logic then skip to the next
frame, rerun the check, etc until it either finds an initialized frame or
the last frame of the ringbuffer.
The problem was, however, that the initial uninitialized frame was possibly
(likely?) still being initialized by the kernel. A data race between the
notification through the socket (the poll()) and the updating of the
`tp_status` field in the frame could lead to a valid frame getting skipped.
Of note is that for example libpcap does not do frame scanning. Instead it
simply exits it ring processing loop. Also interesting is that libpcap uses
atomic loads and stores on the tp_status field.
This skipping of frames had 2 bad side effects:
1. in most cases, the buffer would be full enough that the frame would
be processed in the next pass of the ring, but now the frame would
out of order. This might have lead to packets belong to the same
flow getting processed in the wrong order.
2. more severe is the soft lockup case. The skipped frame sits at ring
buffer index 0. The rest of the ring has been cleared, after the
initial frame was skipped. As our pass of the ring stops at the end
of the ring (ptv->frame_offset + 1 == ptv->req.v2.tp_frame_nr) the code
exits the ring processing loop at goes back to poll(). However, poll()
will not indicate that there is more data, as the stale frame in the
ring blocks the kernel from populating more frames beyond it. This
is now a dead lock, as the kernel waits for Suricata and Suricata
never touches the ring until it hears from the kernel.
The scan logic will scan the whole ring at most once, so it won't
reconsider the stale frame either.
This patch addresses the issues in several ways:
1. the startup "discard" logic was fixed to not skip over kernel
frames. Doing so would get us in a bad state at start up.
2. Instead of scanning the ring, we now enter a busy wait loop
when encountering a kernel frame where we didn't expect one. This
means that if we got a > 0 poll() result, we'll busy wait until
we get at least one frame.
3. Error handling is unified and cleaned up. Any frame error now
returns the frame to the kernel and progresses the frame pointer.
4. If we find a frame that is owned by us (TP_STATUS_USER_BUSY) we
yield to poll() immediately, as the next expected status of that
frame is TP_STATUS_KERNEL.
5. the ring is no longer processed until the "end" of the ring (so
highest index), but instead we process at most one full ring size
per run.
6. Work with a copy of `tp_status` instead of accessing original touched
also by the kernel.
Bug: #4785.
Reset PacketRelease callback to make sure its not set to a capture
specific callback.
As an example:
0x000055e00af09d35 in AFPReleaseDataFromRing (p=0x7f1d884cb830) at source-af-packet.c:653
0x000055e00af09dd0 in AFPReleasePacket (p=0x7f1d884cb830) at source-af-packet.c:678
0x000055e00ab53d7e in TmqhOutputPacketpool (t=0x55e00fb79250, p=0x7f1d884cb830) at tmqh-packetpool.c:465
0x000055e00af08dec in TmThreadsSlotProcessPkt (tv=0x55e00fb79250, s=0x55e012134790, p=0x7f1d884cb830) at tm-threads.h:201
0x000055e00af08e70 in TmThreadsCaptureInjectPacket (tv=0x55e00fb79250, p=0x7f1d884cb830) at tm-threads.h:221
0x000055e00af08f2e in TmThreadsCaptureHandleTimeout (tv=0x55e00fb79250, p=0x0) at tm-threads.h:245
0x000055e00af0ba76 in ReceiveAFPLoop (tv=0x55e00fb79250, data=0x7f1d884ccb60, slot=0x55e01198e4b0) at source-af-packet.c:1321
0x000055e00ab55257 in TmThreadsSlotPktAcqLoop (td=0x55e00fb79250) at tm-threads.c:312
0x00007f1dca9d5609 in start_thread (arg=<optimized out>) at pthread_create.c:477
0x00007f1dca7c6293 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95
Here the packet was a pseudo packet to handle a timeout condition. But
the ReleasePacket callback was still set to AFPReleasePacket from a
previous use of the Packet.
Bug: #4807.
The logic flow we want to achieve with unittests, where first we have
all FAIL statements and then just one PASS statement could become more
convoluted with the existence of the PASS_IF macro. Besides, what could
be written as a FAIL_IF might in some cases be written in not so clear
ways with the PASS_IF option available.
Also: fix inverted check values in documentation, update copyright year
Optimization: #4795
Also: change them to comply with the deletion of PASS_IF macro &
condense checks for invalid dsizes in one test, have all checks on same
valid dsize happen in a single test.
Task: #4021
Previously the flow would hold on to the app-layer and segment data
until the end of the flow, even though it would never be accessed again.
This patch clears app-layer and stream data, but not stream ssn as its
used in flow logging.
Bug: #4778.
Run flow eviction and flow inject queues for bypassed packets as well,
to avoid a scenario where these won't get run at all if too much of the
traffic is bypassed.
Bug: #4779.
Locally bypassed flows had unsafe updates to `Flow::use_cnt` leading to a race
issue. For a packet it would do the flow lookup, attach the flow to the packet,
increment the `use_cnt`. Then it would detect that the flow is in the bypass
state, and unlock it while holding a reference (so alos not decrementing the
`use_cnt`). When the packet was then returned to the packet pool, the flow would
be disconnected from the packet, which would decrement `use_cnt` without holding
the flow lock.
This patch addresses this issue by disconnecting the flow from the packet
immediately when the bypassed state is detected. This moves the `use_cnt`
decrement to within the lock.
Bug: #4766.
In the case, we receive a range request with expected
overlap then new bytes, but the response does not get to the
new bytes, we are still skipping, but the HttpRangeContainerBlock
had the ownership of the files, and need to give it back
DNP3, ENIP, HTTP2 and Modbus are supposed to be disabled
by default. That means the default configuration does it,
but that also means that, if they are not in suricata.yaml,
the protocol should stay disabled.
Instead of using local implementations for the queue.h macro,
wrap the system provided queue.h and then adding missing
features as needed.
The idea is that Suricata when integrated with another library
that includes sys/queue.h can look at the same source of truth
for these macros.
But not all operating systems include a queue.h with the same
features, and some don't include it at all, like Windows. So
on Windows this will be a full implementation of all the queue.h
features Suricata needs.
ThresholdHandlePacketRule may take ownership of an allocated
DetectThresholdEntry, and places it in a position of the
array th_entry. But it never got released
DetectEngineReloadThreads does not work for the fuzz targets
as there is no_of_detect_tvs = 0 as we did not register
real threads and slots.
So, we force the flow worker module to use the newly detect engine
conetxt with all it needs
The flows exceeding the spare pools config setting would be freed
per at max 100 flows a second. After a high speed test this would
lead to excessive memory use for a long time.
This patch updates the logic to free 10% of the excess flows per
run, freeing multiple blocks of flows as needed.
Bug: #4731.
In a scenario where there was suddenly no more traffic flowing, flows
in a threads `flow_queue` would not be processed. The easiest way to
see this would be in a traffic replay scenario. After the replay is done
no more packets come in and these evicted flows got stuck.
In workers mode, the capture part handles timeout this was updated to
take the `ThreadVars::flow_queue` into account.
The autofp mode the logic that puts a flow into a threads `flow_queue`
would already wake a thread up, but the `flow_queue` was then ignored.
This has been updated to take the `flow_queue` into account.
In both cases a "capture timeout" packet is pushed through the pipeline
to "flush" the queues.
Bug: #4722.
When testing for fanout support a cluster-id of 1 was always being
used instead of the configured cluster-id. This limited fanout
support to only one Suricata instance.
Instead of hardcoding an ID of 1, use the configured cluster-id.
Also make cluster_id a uint16_t instead of an int in AFPThreadVars.
Redmine issue:
https://redmine.openinfosecfoundation.org/issues/3419
A block is determined out of order on opening.
But on closing, the gap before it may have been filled.
So, we must post-process it, ie iterate over the red and black
tree so see what blocks we can get.
Better structure design to ensure that one flow maximum
is owning and appending into the file, adding fileOwning field.
Adds also a gap field in a range buffer, so that we can
feed the gap on closing, when we are protected from concurrency
by a lock, (lock which got removed in the append path)
Fixes memcap when encountering a duplicate while inserting
in red and black tree
Adds many comments
Simplify locking by using the THashData lock instead of a separate
range lock.
Avoid size_t in function arguments.
Clean up file handling functions.
Implement handling of alloc errors.
Rename yaml entry to byterange
Unify public api naming
adds a container, ie a thread safe hash table whose
key is the filename
keep a tree of unordered ranges, up to a memcap limit
adds HTPFileOpenWithRange to handle like HTPFileOpen
if there is a range : open 2 files, one for the whole reassembled,
and one only for the current range
Moves Lua includes to util-lua.h instead of suricata-common
so plugins (or library users) don't need to be aware of the
Lua that Suricata was built with. Instead only source files
that need to be Lua aware can include util-lua.h.
The internal flag is not really used. This also reverts the behaviour of
the plugin refactor of passing a ConfNode based on the plugin name
instead of the eve ConfNode.
Register known eve output file types during eve registration. This
removes the function to load internal plugins as they are not really
plugins and moves the registration of them into their respective
subsystem.
With internal code using the plugin API to register an Eve
filetype, the name plugin no longer makes sense. This is
part of my idea that internal plugins aren't plugins at all,
and the plugin interface should be an abstraction over
internal APIs.
Along that idea, this is the start of a refactor of the plugin
file types to be internal, where the plugin API is just an external
interface to that internal API.
This commit adds an interface for loading plugins that are internal to
Suricata. These are always loaded and are in a modified format from
external pliguns.
This value is used as a sentinel to determine if the configured
eve.json filetype was located.
First, the built-in and internal output plugins are checked. If the
sentinel value remains set, the external plugins are searched for the
filetype.
The callback for FileInfo was returning the wrong value, resulting
in loss of some tuple values for one calling SCFileInfo in a script.
The documentation comment wasn't mentioning the sha items that are
pushed.
If the default port is set via the Rust registration table but the probe
fns to server and to client are set to None, the port is never used.
Setting port in such a case is useless so error out.
Allow progress values in the range 0-47 so we have 48 bits to track
prefilter engines.
Mark bits 48-62 as reserved explicitly.
Add debug validation checks to make sure the reserved space isn't used.
Fix FNs in case of too many prefilter engines. A transaction was tracking
which engines have run using a u64 bit array. The engines 'local_id' was
used to set and check this bit. However the bit checking code didn't
handle int types correctly, leading to an incorrect left shift result of
a u32 to a u64 bit value.
This commit addresses that by fixing the int handling, but also by
changing how the engines are tracked.
To avoid wasting prefilter engine tracking bit space, track what
ran by the progress they are registered at, instead of the individual
engine id's. While we can have many engines, the protocols use far
fewer unique progress values. So instead of tracking for dozens of
prefilter id's, we track for the handful of progress values.
To allow for this the engine array is sorted by tx_min_progress, then
app_proto and finally local_id. A new field is added to "know" when
the last relevant engine for a progress value is reached, so that we
can set the prefilter bit then.
A consquence is that the progress values have a ceiling now that
needs to fit in a 64 bit bitarray. The values used by parsers currently
does not exceed 5, so that seems to be ok.
Bug: #4685.
SigTestDropFlow04 was incorrectly expecting an alert in the packet
following a "drop" packet. The first drop is applied to the flow, so
it should lead to the 2nd packet being dropped before inspection is
run.
Clean up the test as well.
Unify handling of signature matches between various rule types and
between noalert and regular rules.
"noalert" sigs are added to the alert queue initially, but removed
from it after handling their actions. This way all actions are applied
from a single place.
Make sure flow drop and pass are mutually exclusive.
The above addresses issue with pass and drops not getting applied
correctly in various cases.
Bug: #4663
Bug: #4670
Victor Julien
Philippe Antoine
Lukas Sismis
Steven Ottenhoff
Modupe Falodun
Eric Leblond
Jeff Lucovsky
Jason Ish
Victor Julien
Sam Muhammed
Benjamin Wilkins
Juliana Fajardini
Shivani Bhardwaj