Fix confusion: f_ms_establish_ul_tbf() vs f_establish_tbf(). The
new name precisely describes what the function does. Sending a
RACH.ind and receiving a DATA.req with RR Immediate Assignment
is just a part of the TBF establishment process.
Change-Id: I1e22b9936bf68ebcaf70cefbc57345995ebbbaed
I noticed that TC_egprs_pkt_chan_req_reject_exhaustion has started
to fail since the refactoring change [1]. As it turned out, this
test case does not init a GprsMs instance, which is needed by a
function it depends on - f_TC_egprs_pkt_chan_req().
I don't really see the point of using the GprsMs abstraction for
such small test cases sending just a few messages, which do not
require us to keep any state. Let's just revert those changes,
and keep using the low level API.
[1] Ib3fee37580f0ea0530a659dec83656799bf57288
Change-Id: I4f32c138e7587e68d2d35d13d13c044893fec4b2
Let's test the code path where UL TBF is requested through DL ACK/NACK
here, since we already test the usual UL TBF through CCCH approach in
most tests.
rlc_mode is changed to ACKED since that's the mode we are using so far
in tests.
Change-Id: I5a9a2e8107c87fdbf74cc2f09ae5eeafbb13ad55
Both f_ms_rx_imm_ass_ccch() and f_ms_establish_ul_tbf() functions
are actually twin brothers of good old f_pcuif_rx_imm_ass() and
f_establish_tbf() with some minor changes.
The former accepts a GprsMS parameter, that is never used. The
latter simply calls f_ultbf_new_from_rr_imm_ass() in the end.
Let's avoid code duplication:
- call f_establish_tbf() from f_ms_establish_ul_tbf(),
- remove f_ms_rx_imm_ass_ccch(), use f_pcuif_rx_imm_ass().
After the removal of f_ms_rx_imm_ass_ccch(), the implementation
of TC_ta_idle_dl_tbf_ass() does not need the GprsMS state, so
let's make it look like it was before.
Change-Id: If6c0b8796500e96525b7b1cadb61ab2fc84b4744
Before this patch, each test had to somehow keep state for all the
transactions needed. Now, most of the state is moved to generic GprsMS,
UlTbf and DlTbf structures, and APIs to maintain its state, as well as
function helpers to submit or receive messages from it. For now
specially the Tx side was improved, some of the Rx parts are left for
later and are still using the old APIs.
This will allow for more complex scenarios and more complex tests
running several MS.
All the tests were updated to use the new APIs, reworked when needed and
even totally rewritten in some cases since they were doing
inconsistent/wrong stuff from the point of view of what the scenarios
or code paths they were expected to test. There's no test regressions.
Change-Id: Ib3fee37580f0ea0530a659dec83656799bf57288
We don't (yet) support multi-BTS test cases anyway. Ideally, each
virtual BTS would be a separate component with an individual port.
Change-Id: I8b639d179db259bf0e43cf1929447a44d5736f62
Send 7 RACH indications to the IUT with EGPRS Packet Channel Request.
Since we have only one timeslot (and USF value '111'B is reserved),
the 8-th indication should be properly rejected by the IUT.
Change-Id: Ie6e5fc68e1591c57e21541ba16fbdcd3fe477ac7
Related: OS#1548
At the moment, the IUT does not support any emergency services.
Make sure that EGPRS Packet Channel Request for an emergency call
is properly rejected (RR Immediate Assignment Reject).
Note that at the time of writing this test, the IUT does not
handle EGPRS Packet Channel Request properly, so it fails.
Change-Id: I63d989e89e6235a631e024c2810a3a4b0de56ccf
Related: OS#1548
The purpose of this test case is to verify the contents of RR
Immediate Assignment Reject message (and its IAR Rest Octets)
sent in response to EGPRS Packet Channel Request (11 bit RA).
To provoke the reject message, test case crafts an incorrect
EGPRS Packet Channel Request message ('111111xxxxx'B).
Note that at the time of writing this test, the IUT does not
handle EGPRS Packet Channel Request properly, so it fails.
Change-Id: I4bfd5621085d63896e2e9b70355524cf4285036a
Related: OS#1548
This is needed since osmo-pcu.git
I9b4ef7b7277efa645bdb5becf2e9f6b32c99a9b1, where a bug was fixed in
which osmo-pcu was not sending UL ACK/NACK under some conditions.
Change-Id: I1a58b3984a96b432b2cb5300fc8a4261133a4f69
This is a first step towards refactoring of all functions to use MS and
Ul/DlTBF objects containing state.
Change-Id: Ieae27d6e707f79ec2145864ef5cd67ddbbec9314
Test Verifies that if PCU doesn't get one of the intermediate UL data blocks in a UL
TBF, it will request retransmission through UL ACK/NACK (with missing block
in its bitmap) when CV=0 is received (and hence it knows no more data is to be
transferred).
This test fails as of current osmo-pcu master, and it's fixed there by
osmo-pcu.git Change-Id I9b4ef7b7277efa645bdb5becf2e9f6b32c99a9b1.
Change-Id: I204a470e47fcc5965de758ad9a275837e0c8034d
Old code was not setting Single Block Packet Access type, and 2phase
access was not properly triggered.
Once it's triggered, message flow changes quite a lot from the 1phase
access, specially because the 2nd Ul Assignment arrives through PDCH
instead of CCCH, which means a different record is received and hence
code for 1phase cannot be easily re-used.
For similar reasons, f_tx_rlcmac_ul_n_blocks() is modified to receive
the only required tfi param instead of a full dl_block.
Some functions are also extended to support SingleBlock Allocation
instead of usual DynamicAllocation.
Change-Id: If636a4898dfa175fdbd6baf04f7f2c955a9c525d
Downlink Control blocks like Packet Uplink Assignment (PACCH)
contain rrbp + rrbp_valid in mac headers
Change-Id: I0401b0b378c7770d06a15d14dac6436303b4ccab
This function was written in a way that it tries to do as
many different (but related) things as possible:
a) send an RTS.req to the IUT, expect a DATA.ind in return,
b) decode RLC/MAC message contained in the received DATA.ind,
c) make sure that it's either GPRS or EGPRS data block,
d) calculate the last TDMA frame number of RRBP using
f_rrbp_ack_fn() regardless of its validity,
e) make sure that the block contains a given LLC payload.
Everything is ok except point d). The problem is that this is
only the case for the first block of RRBP, and not applicable
to the rest having 'rrbp_valid' flag unset. Furthermore, this
function did not match GPRS DL blocks with 'rrbp_valid' flag
unset, for some odd reason.
Let's move RRBP calculation into a separate function called
f_dl_block_ack_fn() and return TDMA frame number of the
received DATA.ind message instead.
Among with that, there are more little changes:
- simplify matching of (E)GPRS DL data blocks,
- use 'in' qualifier in parameter list where possible,
- turn parameter 'data' into a template (present).
Change-Id: I1528408b4399d0a149a23961805277eaab90d407
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
Let's make this template more flexible, so it can be used to match
any GPRS DL data blocks, not only those with rrbp_valid == true.
Note that behavior of f_rx_rlcmac_dl_block_exp_data() is
intentionally left unchanged, and will be fixed later.
Change-Id: I3940216368cdbb58fe89420675d1d8d5f5e49b05
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
The resulting frame number shall be within the period of TDMA hyperframe.
Change-Id: I794a14f69293cbbc937d62d09dd5794956b882db
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
Most of existing test cases are built on top of the PCU interface
abstraction components (see PCUIF_Components.ttcn). This means
that during the test case execution, additional components are
running in parallel, among with the MTC (Main Test Component).
When a test case terminates, either normally or due to an error,
it may happen that the virtual BTS component is stopped before
the associated TDMA clock generator. In this situation, sending
a clock indication towards the stopped BTS component would
lead to a dynamic test case error.
Let's take the process of tear down under control, and ensure that
the clock generator is stopped first. To achieve that, every test
case needs to call f_shutdown() in case of an error, as well as
in case of the normal termination.
Note we cannot use the existing f_shutdown() from Misc_Helpers,
because doing 'all component.stop' does not gurantee that the
clock generator is stopped first, and I experienced at least
one DTE while trying to integrate it.
Change-Id: I6a859687d9605cc08c51ff44d946c279b79bedfa
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
This test case would not fail even in case of Timing Advance mismatch.
Change-Id: Ife86e5e0e39f0112b854ed9a13e9c6f3c49531c9
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
We don't need a copy of the whole RR Immediate Assignment message.
Change-Id: I44417460126e507a0a47a5aee8c4a995085023fa
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
In response to a CHANNEL REQUEST received on BCCH, the network
would never allocate a Downlink TBF, so we should always expect
an Uplink TBF assignment.
Change-Id: I6b4c108bed39ba9ac9b6144827bc1e20b04333b4
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
Back in September 2019, while writing the first lines of the new
test infrastructure for OsmoPCU, I picked a random value as the
default RA (CHANNEL REQUEST message) for Uplink TBF establishment.
It has been accepted by the IUT so far, and still works, but
according to 3GPP TS 44.018, table 9.1.8.1, value '3A'O has
nothing to do with GPRS - it actually means 'Answer to paging'.
The new value belongs to range '011110xx' - one phase packet
access with request for single timeslot uplink transmission.
Change-Id: Ic036d380af3667d54a3a0a011a9d56a87e0f949b
Signed-off-by: Vadim Yanitskiy <axilirator@gmail.com>
RLCMAC blocks have a lot of fields and we will potentially require lots
of different templates, as well as functions to handle related structs.
Change-Id: I9c6597178168aa3848b21930f33be698dd2ce545
Delete PCU_Tests.ttcn along with the configs, since they are broken and
we are only adding new tests to PCU_Tests_RAW.ttcn. They are broken
because it's not possible to run multiple tests after another.
Some components which used to be in PCU_Tests.ttcn and which were used
by other PCU_Tests*.ttcn files have been moved to SGSN_Components.ttcn
Selftests have been moved to PCU_selftest.ttcn.
Small placeholder for module PCU_Tests is left in PCU_Tests.ttcn to
still be able to compile the binary as "PCU_Tests" and avoid changing
that part of docker setup. We'll eventually rename RAW tests soon
anyway.
Change-Id: Ie862a1525e9f4f9a3f2427ac3898810e3d044d2f
There's also DL_ACK_NACK message for which a template will be introduced
soon, so let's rename and fix typos/wrong descriptions to avoid
confusion later.
Change-Id: I4a2025ad282006953fcfadf429c980b77cb94371
This change is a step towards getting rid of the old test case
infrastructure. Note that a call to f_bssgp_establish() is moved
out of f_init_bssgp() to the test case's body.
Change-Id: If15339f02c5188e60fcb47ae6dc0ac289efa2896
Both 't_IMM_ASS_TBF_DL' and 't_RR_IMM_ASS_TBF_DL' templates were
introduced for a specific task - matching Packet Immediate
Assignment (Downlink TBF) by TLLI.
In the upcoming changes we will also need to match Uplink TBF
assignment, and more generic fields such as Timing Advance.
Let's add a generic template for Packet Immediate Assignment
and allow passing IaRestOctets as a parameter.
Change-Id: I492cf990820ba153ea71469b8b623e56e031e549
According to 3GPP TS 44.018, section 10.5.2.16, IA Rest Octets IE
starting with 'HH' bits may contain one of the following CSN.1
encoded components:
7 6 5 4 3 2 1 0 Bit Numbers
H H 0 0 . . . . Packet Uplink Assignment
H H 0 1 . . . . Packet Downlink Assignment
H H 1 . . . . . Second Part Packet Assignment
We already have (partial) support for the first two, while the
last type has not been supported so far. Let's add it.
Also, this change introduces several templates for IA Rest Octets
IE and some of its components mentioned above. This would allow
us to abstract the API users from dealing with further changes,
e.g. adding a coding attribute 'CSN.1 L/H' and missing fields.
Change-Id: Ib3a21e7c5fa1cad4466e3a09fa70540de7f6ecc5
Make the decoding level (BSSGP, LLC, SNDCP, L3) configurable, so the
existing PCU tests, that expect messages only decoded to the BSSGP
level, can pass again. Move the SNDCP decoding in f_dec_bssgp above the
L3 decoding, so f_dec_bssgp goes through the layers in the reverse order
of f_send_bssgp_dec.
I have verified, that all testsuites using the BSSGP Emulation (SGSN,
PCU, PCU-SNS) are still working with this patch.
Related: OS#4180
Fixes: 955aa94504 ("BSSGP_Emulation: Abandon "BssgpDecoded" intermediate structure")
Change-Id: I8f76385528c1de98c557cee451c0e0dfd182b605
It originally seemed like a great idea to define a custom record
which aggregates the decoded BSSGP, LLC, L3 and/or SNDCP and passes
it to the individual ConnHdlr. However, particularly with testcase
interoperability for IuPS in mind, this seems bogus. Also, we
never really took advantage of this.
Let's now decode as far as we can decode any PDU, and then send the
decoded version of that PDU via the ports between the BSSGP_Emulation
and the ConnHdlr component.
Change-Id: I8c1082880902dd9a04935945f0293895f4d0c53a
Some of our files didn't have a copyright notice at all, let's add
it. Also, update the notices in other files and ensure a SPDX
identifier is present in all but the most trivial files.
Change-Id: If7fa19ce484b415bc645e39b3d0d666b44b5f0fd
The PCU is supposed to forward any RR SUSPEND received by the BTS
on dedicated channels and pass it to the SGSN via BSSGP. This provides
knowledge to the SGSN that the MS is not reachable via PS/TBF anymore
until it completes its CC dedicated mode transaction and RESUMEs.
Change-Id: Iaf35e0c1087413db1dc7d169fa716df14c78f5db
Implement a basic paging test for the PCU, which is passing for paging
via TMSI (but only if osmo-pcu is started after the test is started).
Previously, this test code amounted to a debugging loop which
never terminated.
Change-Id: Id0384e0742ab91983615e4f1c883bb044c1c8b18
Related: OS#2404
* remove unused parameters
* add expected results
* use control for default test execution
* only tests which could (in theory) pass are left in control
Previously generic control had some tests which never called
"setverdict(pass)" which means they would always fail. Such tests might
be useful during development but should not be part of test set executed
by default.
Change-Id: I91910e18f108f8ee9a3b76ced16c420fdbdb665a
Related: OS#2890
The NS_Emulation has configuration values hardcoded or bound
to module parameters which prevents multiple instances.
Replace the module parameter based configuration with configuration
given when starting the NS_Emaulation.
Change-Id: I9128f9ad5c372779c38799269393137ba52576cd
Call mtc.stop after setverdict(fail), add reasons to most failures and
fail with verdict error for internal errors.
Change-Id: I9b618235939fa41160b9be6677b121963d3ec857
osmo-pcu discards PAGING-PS messages unless received on the
signalling BVCI 0. The BVCI value must be set to zero both
in the NS header and the BSSGP header. Create new ports between
the PCU_Test components which the TC_paging test case can use
to ensure that both layers send frames with BVCI value of zero.
This does not make PCU_Tests.tc_paging work yet, however the PS
paging request is now processed by osmo-pcu rather than discarded.
Change-Id: I0437123b04b7320a4f690f0646578c57abf6bc87
Related: OS#2404
Pau Espin
Vadim Yanitskiy
Vadim Yanitskiy
Oliver Smith
Harald Welte
Daniel Willmann