There's two variants for the Ms Radio Access Capabilities.
* The usual encoding with spare bits (usually to fill up to octet boundary)
as defined in TS 24.008 Table 10.5.146
And there's too:
* MS Radio Access Capabilities 2 IE from TS44.060 section 12.30, which is
the same but removing all spare bits, and which is used in messages like
Packet Resource Request and Additional MS RAC messages.
The later is used basically for messages having extra IEs after the MS
Radio Access capabilities IE, since they are encoded immediatelly
afterwards.
So this patch does:
* Adds the expected spare bits (M_PADDING) to MS_Radio_Access_capability_t
* Creates a new MS_Radio_Access_capability2_t without padding
* Updates code to use the new "2" version where needed.
Note RLCMACTest long de/encoding line logs change only because the name
of the struct changes (the "2" is added).
Change-Id: Ibd756f80a03452a651e2771dbc628d701e55ac4b
It seems the assumptions regarding maximum number of RA capabilitites
in one message were wrong. Doing some rough calculations, each RA
capabilitiy value (without extensions) can take around 20ish bits, which
means for a message containing up to 52 bytes that quite a lot of
different values could be theoretically fed in. Let's be safe and
increase the array size to be able to handle all different access
technologies listed in See TS 24.008 table 10.5.146 following
restrictions:
* "The MS Radio Access capability is a type 4 information element, with a maximum length of 52 octets."
* "Among the three Access Type Technologies GSM 900-P, GSM 900-E and GSM 900-R only one shall be present."
* "the mobile station should provide the relevant radio access
capability for either GSM 1800 band OR GSM 1900 band, not both".
Wireshark requires similar fix (it's not important though because it
currently uses another ad-hoc decoder for RAcap).
Related: OS#4463
Change-Id: I5334eaacfbc238fae8bea50c9e9667c2117f81ff
This way if CSN1 encoded bitstream contains more elements than what the
defintion expects it will fail instead of overflowing the decoded
buffer.
RA cap struct placed in unit test is taken from a real android phone
sending the value when attaching to the network. Then SGSN sends it back
and osmo-pcu would crash similar to unit test:
*** stack smashing detected ***: terminated
Process terminating with default action of signal 6 (SIGABRT): dumping core
at 0x4C62CE5: raise (in /usr/lib/libc-2.31.so)
by 0x4C4C856: abort (in /usr/lib/libc-2.31.so)
by 0x4CA62AF: __libc_message (in /usr/lib/libc-2.31.so)
by 0x4D36069: __fortify_fail (in /usr/lib/libc-2.31.so)
by 0x4D36033: __stack_chk_fail (in /usr/lib/libc-2.31.so)
by 0x124706: testRAcap2(void*) (RLCMACTest.cpp:468)
Related: OS#4463
Change-Id: I9fe0e55e0a6a41ae2cc885fba490c1d4a186231e
This way the macros can be used to access the arrays themselves and
calculate its static size to enable validation later.
In the case of Packet_Access_Reject_t, modify the description to use a
M_REC_TARRAY_1 object to get rid of access to 2nd element. The new
description is the correct one, since the first element is mandatory
according to TS 44.060 Table 11.2.1.
Change-Id: I6f10350d4734360c7a15a702c72b59efd84987ee
It was recently discovered that the Racap value used for testRAcap was
actually malformed (it was taken from a TTCN3 test). It had the presence
bit for "EGPRS multislot class" set but no struct was put after it.
So let's move that malformed value to a new test named
testMalformedRAcap(). Since it doesn't make sense trying to re-encode or
do similar things with an initially malformed value, let's drop all
those parts in this new test.
For the old testRAcap() test, use a new bitstream this time with the
"EGPRS multoslot class" struct set inside (class 3).
Change-Id: I1e7f8d8866695732ee24a79d8b54d660fd4f22d5
low-level text decodes of CSN.1 messages certainly are not NOTICEable
events, but rather something used for debugging.
Right now we get various text CSN.1 log output of osmo-pcu in it's
default configuration. Despite all log levels being relatively high
(NOTICE), we still see those messages as they simply are logged
at the wrong level.
Related: OS#2577
Change-Id: I7b42c9e21ad8d8a5b54e7a3b68490934ce3d3198
This commit is basically a revert of
f4bb42459c, which disabled the code. That
commit claimed the SGSN may have providen inacurate or wrong data at the
time, but then it should be fixed in the SGSN.
Related: OS#1525, OS#3499
Change-Id: Ie36ae23203110018d4b5ae47591e0a64989e23a0
We should really be using monotonic clock in all places that
gettimeofday is used right now. Since clock_gettime() uses timespec,
let's move all code to use timespecs instead to avoid having to convert
in several places between timespec and timeval.
Actually use osmo_clock_gettime() shim everywhere to be able to control
the time everywhere from unit tests.
Change-Id: Ie265d70f8ffa7dbf7efbef6030505d9fcb5dc338
Both csnStreamDecoder() and csnStreamEncoder() shall not return 0
prematurely if no more bits left in the input / output bit-vector.
Returning CSN_ERROR_NEED_MORE_BITS_TO_UNPACK might make more sense,
however we don't know in advance (i.e. without entering the loop)
whether it's an error or not. Some CSN.1 definitions have names
like 'M_*_OR_NULL', what basically means that they're optional
and can be ignored or omitted.
Most of the case statements do check whether the number of remaining
bits is enough to unpack / pack a value, so let's leave it up to
the current CSN_* handler (pointed by pDescr) if no bits left.
Return CSN_ERROR_NEED_MORE_BITS_TO_UNPACK only if the number of
remaining bits is negative as this is an error in any case.
Change-Id: Ie3a15e210624599e39b1e70c8d34efc10c552f6c
When running the test under a RaspberryPI4:
+../../../src/llc.cpp:216:29: runtime error: signed integer overflow: 864197544 * 1000 cannot be represented in type 'long int'
864197544 comes from comparing initial insertion time and dequeue time
(test does a big jump in time): 987654321 - 123456777
let's use more realistic time changes, since the current one account for
about 37 years.
Change-Id: I28abc9192e0e7c590bc1c3c88950627cf669ffaf
Port from wireshark.git de028e81c53f9c45ccc5adb3bffd2f16ae2017bf
This commit breaks transcoding of the test vectors containing
the MS RA Capability IE due to the reasons explained in [1].
The more fields we add, the longer gets the output of the CSN.1
encoder. This is not critical, since we never need to encode
messages containing the MS RA Capability IE on practice.
[1] Ibb4cbd3f5865415fd547e95fc24ff31df1aed4c0
Ported-by: Pau Espin Pedrol <pespin@sysmocom.de>
Change-Id: Ibb4cbd3f5865415fd547e95fc24ff31df1aed4c0
This change fixes a bug that was reported by Keith Whyte and
confirmed in [1]. The problem is that a user-defined handler
in case of CSN_SERIALIZE may parse only a part of the given
bit-stream, leaving some bits unhandled. This is expected
because the sender (i.e. the MS) may use more recent RLC/MAC
message definitions containing new fields at the end.
Those bits that were left unhandled by serialize() shall not be
interpreted as continuation of the message, they shall be skipped.
Note that the encoded vector in the RLCMAC unit test still does
not match the original one. That's a known bug explained in [2].
[1] If5873355d52d7ddb06c2716154a88d34100f6ab5
[2] Ic46d6e56768f516203d27d8e7a5adb77afdf32b7
Change-Id: Id4cc042fed68fc54aca0355dcb986cab3f6b49ea
Related: OS#4338
This test vector (from HTC Desire 628) demonstrates a bug in the
CSN.1 decoder. For some reason, OsmoPCU fails to decode it:
DCSN1 ERROR csnStreamDecoder: error NEED_MORE BITS TO UNPACK (-5)
at EGPRS_TimeslotLinkQualityMeasurements (idx 164)
while Wireshark dissects it without any errors.
Change-Id: If5873355d52d7ddb06c2716154a88d34100f6ab5
Reported: https://osmocom.org/issues/4338#note-15
Related: OS#4338
Move the call to send_gsmtap() before the call to decode_gsm_rlcmac_uplink() as if
the latter returns error we return and never get to see the packet on the GSMTAP.
Change-Id: Ia6af9f40590f28fcae3fef50d9c601d8435412cd
The main idea of this change is to demonstrate a weakness of the
CSN.1 codec that most likely causes a unit test breakage in [1].
The problem seems to be that the transitional structures, where
the CSN.1 decoder stores the results, do not contain any details
about presence of the optional fields (such as M_UINT_OR_NULL).
In other words, it's impossible to know whether some optional
field is omitted in the encoded message (NULL), or is it just
set to 0. This means that the encoder will always include all
optional fields, even if they're not present in the original
message.
[1] Ibb4cbd3f5865415fd547e95fc24ff31df1aed4c0
Change-Id: Ic46d6e56768f516203d27d8e7a5adb77afdf32b7
As seen in OS#4420, setting the MetaInfo.recv_time outside of
llc_queue before calling llc_queue::enqueue() and later on using that
value in llc_queue itself at dequeue time is not a good idea, since it
can provoke errors if the recv_time was not set correctly.
For instance, LlcTest was not setting the value for recv_time on some
test, which ended up with a huge millisec value when substracting now()
from it:
"""
llc.cpp:215:29: runtime error: signed integer overflow: 1582738663 * 1000 cannot be represented in type 'long int'
"""
This issue only appeared when started building on a raspberrypi4.
Let's better set/store the MetaInfo.recv_time internally during
llc_queue::enqueue(). Then, enqueue() only needs the
MetaInfo.expire_time, so let's change its arg list to only receive that
to avoid confusions.
Take the chance to move the llc_queue APIs to use osmo_gettimeofday,
since we need to fake the time now that the API itself sets that time.
Also take the chance during this refactor to disallow passing null
pointer by default since no user needs that.
Finally, update the LlcTest accordingly with all API/behavior changes.
Related: OS#4420
Change-Id: Ief6b1464dc779ff22adc2b02da7a006cd772ebce
Long story short: as it turns out the test vector '12a5146200'O
has been generated by TITAN, and it's malformed. The length
indicator it contains must be at least 29 bits, not 21. This
field is calculated by TITAN automatically, so I guess there
is a bug somewhere in its RAW encoder implementation.
It's funny that Wireshark decodes the old malformed vector without
any problems if it's encapsulated into the BSSGP DL-UNITDATA. The
reason for that is because BSSGP dissector does not actually use
the CSN.1 codec and relies on its own hand-written parser [1],
which does not respect the length constraints.
Furthermore, table 10.5.146/3GPP TS 24.008, describing the format
of MS Radio Access Capability IE, has the following comment:
< Multislot capability struct > ::=
{ 0 | 1 < HSCSD multislot class : bit (5) > }
...
-- error: struct too short, assume features do not exist
so ideally our CSN.1 decoder should be more tolerant to the old
malformed vector, but unfortunately error handling is not implemented.
[1] See de_gmm_ms_radio_acc_cap() in epan/dissectors/packet-gsm_a_gm.c.
Change-Id: I5f810397b8d09c18e069168023429f6a4d899c86
Found while doing differential analysis (comparison against the
original implementation from Wireshark).
Change-Id: Ibd0b7400d78f7873c2a8d45267332f511b5c6fbb
Found while doing differential analysis (comparison against the
original implementation from Wireshark).
Change-Id: I9f7fa9c3f2f4ff5213dded930cee7ec509b9d799
Found while doing differential analysis (comparison against the
original implementation from Wireshark).
Change-Id: Id2a4f03035cd8354d3fba0ad37571453d3986d21
The implementation of CSN.1 codec was taken from Wireshark, where
it's implemented in pure C. For some reason it was mixed with C++
specific features, mostly using references in parameter
declaration. Not sure what are the benefits.
Change-Id: I56d8b7fbd2f9f4e0bdd6b09d0366fe7eb7aa327a
This would allow us to catch more bugs. Note that I had to remove
printing of pointer address to make the output deterministic.
Change-Id: I1a77441eb957353c919bc73f8e3a2e38f4a383a9
As was discovered recently (see OS#4388), bitvec_read_field()
would never return a negative value because its return type
is unsigned (uint64_t).
We don't really need to get more than one bit, so let's just
use the bitvec_get_bit_pos() instead.
Change-Id: I763a295cd955cd33f542292c85d97ff82f6b49bc
Related: OS#4388
It contails no valid identity, and thus violates the specs.
Let's keep it 'as-is' to check that decoder actually fails.
Change-Id: I663edfdaac0c065e08ab7b6dc50d2f18e433433c
Related: OS#4392
After the recent changes [1], it was noticed that one of the unit
tests fails. In particular, a decode-encode cycle of Packet
Polling Request produces a different vector:
vector1 = 49 13 e0 08 50 88 40 13 a8 04 8b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b
vector2 = 49 13 01 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b 2b
vector1 == vector2 : FALSE
As it turns out, the original (input) vector itself is malformed
because it contails no valid identity, and thus violates the
specs. The CSN.1 decoder from Pycrate [2] throws an exception
while trying to decode it. I believe we should do the same.
Let's stop decoding the bit stream and return an error in case
if neither of a given list of the choice items matched.
[1] Ia0f8cc224a4c38e80699f834fd83d4c0d99322ea
[2] https://github.com/P1sec/pycrate
Change-Id: I420144773ed5e80372534e0f18db5e74cdb2999d
Fixes: OS#4392
The problem is that bitvec_free() is not NULL-safe. Ideally we
need to fix it in libosmocore [1], but let's also fix it here,
so OsmoPCU can be safely used with older libosmocore versions.
[1] https://gerrit.osmocom.org/c/libosmocore/+/17114
Change-Id: I7647d17b3d03f8e193ef6e793a2d3c1967744eef
Fixes: CID#208181, CID#208179
Output was incorrect before this patch. LOPC was being called without
having any initial LOGP, and trailing newline was usually missing at the
end.
Since csnDecoder/encoder functions are recursive, it's difficult to
handle logging state in a coherent way inside them. Let's better simply
control start/end of logging related topics in the callers of those
functions, and simply use LOGPC everywhere in csn1.cpp.
Change-Id: I50da7560939fac360b7545e2a6bfaf45ed0c4832
It was noticed that OsmoPCU leaks memory when trying to reconnect
to the BTS. It could be easily fixed, but we don't really need to
allocate the PCU socket state on heap as we never have more than
one connection.
Change-Id: Iea8930f443caa16f522f7c5375e0004e4e2315cb
Since I2b32b4fe20732728db6e9cdac7e484d96ab86dc5, go_parent_cb()
is completely optional. It no longer has the task to determine
the correct parent node. The is_config_node() callback is no
longer needed too. Get rid of them.
Since Ic5e69a396df659933fd4d50298b9925e837a6861 we depend on 1.3.0.
Change-Id: Id7ce8c4e1ac43747ad40a06d01433c366da07b42
As was discovered by pespin, changing logging level of DCSN1 makes
the CSN.1 decoder behave differently (see OS#4375). In particular,
this makes RLCMACTest (encode / decode test) fail.
I did a quick investigation and noticed that some of the logging
statements call bitvec_read_field(). By definition this function
moves the internal pointer (current bit position) of a given
vector and increments readIndex by a given amount of bits.
The problem is that LOGPC would not evaluate its format string if
the logging message is not going to be printed, e.g. if a given
logging level is lower than the current one, or in case if
logging is not enabled at all.
The first two conditional calls to bitvec_read_field() are related
to CSN_PADDING_BITS, so that's not critical because padding is
always in the end of messages. The later two are related to
CSN_RECURSIVE_ARRAY and CSN_RECURSIVE_TARRAY respectively.
Let's use bitvec_get_uint() instead to keep readIndex unchanged.
Change-Id: Ia331048db9f790ca407fd341ced01df12d10a233
Fixes: OS#4375
Pau Espin
Harald Welte
Vadim Yanitskiy
Keith Whyte