In gsm48_rx_mm_serv_req() we need to make sure that a given message
buffer is large enough to contain both 'gsm48_hdr' and
'gsm48_service_request' structures.
Comparing msg->data_len with size of pointer if wrong because:
- we actually need to compare with size of struct(s),
- we need msgb_l3len(), not length of the whole buffer.
Moreover, since we have to use the pointer arithmetics in order
to keep backwards compatibility with Phase1 phones, we also
need to check the length of both Classmark2 and MI IEs.
Change-Id: I6e7454d7a6f63fd5a0e12fb90d8c58688da0951e
Since in parse_umts_auth_resp() we are checking the length of
GSM48_IE_AUTH_RES_EXT TLV, we need to print its length, but
not the length of the whole L3.
Change-Id: I2bfebce6d017be834bfe7628ffa2b341eb82c11c
The MSC_A_EV_HANDOVER_END exists as parent term event for the msc_ho_fsm, but
it is not actually required as functional event, since all cleanup is handled
in msc_ho_fsm_cleanup().
That's why I never bothered to add the event to msc_a_fsm, but of course that
means we get an error message after each (successful and unsuccessful)
handover, that the MSC_A_EV_HANDOVER_END is not permitted.
Allow the event and ignore it to silence the error message.
Explain in a comment.
Change-Id: Ie8dc0c0a631b7da43111f329562007766a21b134
After neels/ho was merged, SMS over IuCS/RANAP was failing in both
MO and MT direction. The reason was that all mobile-terminated SMS-CP
layer messages were sent in RANAP with SAPI-0 instaed of SAPI-1.
Change-Id: I98e6eddb52d5c61c4e2d34bdfcd43cf460296ad7
Closes: OS#3993
The event is actually never dispatched and useless, because when an RTP stream
releases, the call_leg terminates directly anyway (which wasn't apparent when
starting to design the call_leg FSM yet).
Change-Id: I6b2fc1225c960fa2f7c46adf241520217a07821c
The SMPP 3.4 specification defines the password field as a
"Variable-length octet string with maximum length of 9", and according
to table 3-1 this means including the terminating NUL-byte.
However, OsmoMSC allows to configure longer passwords in the ESME
configuration. Those passwords will then never match, as libsmpp34
performs length validation and generates a parser error for anyone
trying to send a longer password via SMPP.
The same applies for system-id, where we have to permit only 15
characters with zero termination, but not 16 characters.
Change-Id: I81ef593e84bf1e15f6746386fc145495fae29354
Closes: OS#3166
Instead of calling trans_log_subsys() for each LOG_TRANS() log line, rather
store in trans->log_subsys once on trans_alloc() and use that.
Do not fall back to the RAN's own subsystem (DBSSAP / DIUCS), it makes little
sense and may cause logging to switch subsystems depending on the RAN state.
In trans_log_subsys(), add missing switch cases:
- Log silent call transactions also on CC.
- Log USSD on DMM.
About USSD: we currently have no dedicated USSD logging category. As a result,
after LOG_TRANS() was introduced [1], USSD logged on DBSSAP/DIUCS or DMSC,
depending on whether a RAN was associated with the trans or not. Before that
change, USSD always logged on DMM, so, until we have a separate logging
category for USSD, consistenly use DMM again.
[1] in I2e60964d7a3c06d051debd1c707051a0eb3101ba / ff7074a0c7
Related: coverity CID 198453
Change-Id: I6dfe5b98fb9e884c2dde61d603832dafceb12123
As per 3GPP TS 03.40, section 9.2.3.16 "TP-User-Data-Length (TP-UDL)",
if the TP-User-Data is coded using the GSM 7-bit default alphabet,
the TP-User-Data-Length field indicates the *number of septets*
within the TP-User-Data field to follow. Otherwise, i.e. in case
of 8-bit or UCS-2 encoded data, the *number of octets* is indicated.
Since we store the original TP-UDL value (as received), we might
need to convert septets to octets before passing it to memcpy().
Otherwise this would lead to a buffer overrun.
Also, as we receive TPDU from untrusted source (i.e. subscriber),
the TP-UDL value needs to be checked against the corresponding
maximum (160 septets or 140 octets) and truncated if needed.
Please note that buffer overrun is still possible, e.g. when an
indicated TP-UDL value is grather than the remaining TPDU length.
Preventing this would require adding an additional check.
Change-Id: I4b08db7665e854a045129e7695e2bdf296df1688
Depends-on: (core) I54f88d2908ac47228813fb8c049f4264e5145241
It was noticed that SCCP_RAN_MSG_RESET_ACK message is not freed after
sending. Since ran_peer_rx_reset() calls sccp_ran_down_l2_cl(), which
then calls osmo_sccp_user_sap_down_nofree(), which doesn't free the
message buffer (what's clear from its name).
OsmoMSC# show talloc-context application full filter msgb
full talloc report on 'osmo_msc' (total 20155 bytes in 88 blocks)
msgb contains 4640 bytes in 5 blocks (ref 0)
bssmap: reset ack contains 1160 bytes in 1 blocks (ref 0)
bssmap: reset ack contains 1160 bytes in 1 blocks (ref 0)
bssmap: reset ack contains 1160 bytes in 1 blocks (ref 0)
Let's free it after sending (or in case of error).
Change-Id: Ic174f6eecd6254af597dfbdc1c9e3d65716f0a76
The misnomed 'nas_decode' and 'nas_encode' APIs have been renamed to
'ran_decode' and 'ran_encode', which was forgotten in the large comment
explaining the message path in sccp_ran.h. Apply the rename there.
Change-Id: I742fb4844ac8a9ad76f59883ae9447eb8819b82d
This fixes the following compiler error:
msub.c: In function ‘msub_fsm_active’:
msub.c:85:35: error: ‘msc_role_a_c’ may be used uninitialized in this function
[-Werror=maybe-uninitialized]
|| (msc_role_a_c && msc_role_a_c->ran->type == OSMO_RAT_EUTRAN_SGS)))
~~~~~~~~~~~~^~~~~
msub.c:59:26: note: ‘msc_role_a_c’ was declared here
struct msc_role_common *msc_role_a_c;
^~~~~~~~~~~~
Change-Id: Id518dea77d01ed0518ca7cba6b1b363f1c8e6543
While developing the inter-MSC handover refactoring, I was annoyed by the fact
that mncc_tx_to_cc() receives an MNCC message struct containing a msg_type, as
well as a separate msg_type argument, which may deviate from each other. So, as
a first step I wanted to make sure that all callers send identical values for
both by inserting an OSMO_ASSERT(msg_type == msg->msg_type). Later I was going
to remove the separate msg_type argument.
I then forgot to
- carry on to remove the argument and
- to actually test with internal MNCC (it so happens that all of our ttcn3
tests also use external MNCC).
As a result, the "large refactoring" patch for inter-MSC Handover breaks
internal MNCC operation.
Fix that: remove the separate msg_type argument and make sure that all callers
of mncc_tx_to_cc() indeed pass the desired msg_type in msg->msg_type, and hence
also remove the odd duality of arguments.
Various functions in mncc_builtin.c also exhibit this separate msg_type
argument, which are all unused and make absolutely no sense. Remove those as
well.
Related: OS#3989
Change-Id: I966ce764796982709ea3312e76988a95257acb8d
We are just introducing smpp34_set_memory_functions() in libsmpp34
to allow applications like OsmoMSC to provide their own heap allocator
callback functions. Let's used this to integrate with talloc and
hence allow talloc tracking/debugging for libsmpp34 internal
allocations.
Depends: libsmpp34 Change-Id I3656117115e89638c093bfbcbc4369ce302f7a94
Change-Id: Ie2725ffab6a225813e65768735f01678e2022128
Related: OS#3913
Get rid of the legacy name bscconfig.h from osmo-nitb times.
Remove the #include from some of the files that aren't actually using it.
Instead of '#include "../../config.h"', use plain '#include "config.h"'
because we're anyway passing $top_srcdir as -I during compilation.
Change-Id: Id4f683be1f36f0630c83da54e02868aae847aeec
Before, I was testing with osmo-hlr patch
I01a45900e14d41bcd338f50ad85d9fabf2c61405 applied, but that patch is currently
in an abandoned state.
This is the counterpart implemented in osmo-msc: always include the terminating
nul char in the "blob" that is the MSC IPA name.
The dualities in the formats of routing between MSCs is whether to handle it as
a char*, or as a uint8_t* with explicit len (a blob).
In the VTY config to indicate target MSCs for inter-MSC handover, we have
strings. We currently even completely lack a way of configuring any blob-like
data as a VTY config item.
In osmo-hlr, the IPA names used for routing are currently received as a char*
which *includes* the terminating nul char. So in osmo-msc, if we also always
include the nul char, it works.
Instead, we could just send the char* part without the nul char, and apply
above mentioned osmo-hlr patch. That patch would magically match a name that
lacks a nul with a name that includes one. I think it is better to agree on one
format on the GSUP wire now, instead of making assumptions in osmo-hlr on the
format of the source/target names for routing. This format, from the way GSUP
so far transmits the IPA SERNO tag when a client attaches to osmo-hlr, happens
to include the terminating nul char.
Change-Id: I9ca8c9eef104519ed1ea46e2fef46dcdc0d554eb
3GPP TS 49.008 '4.3 Roles of MSC-A, MSC-I and MSC-T' defines distinct roles:
- MSC-A is responsible for managing subscribers,
- MSC-I is the gateway to the RAN.
- MSC-T is a second transitory gateway to another RAN during Handover.
After inter-MSC Handover, the MSC-I is handled by a remote MSC instance, while
the original MSC-A retains the responsibility of subscriber management.
MSC-T exists in this patch but is not yet used, since Handover is only prepared
for, not yet implemented.
Facilitate Inter-MSC and inter-BSC Handover by the same internal split of MSC
roles.
Compared to inter-MSC Handover, mere inter-BSC has the obvious simplifications:
- all of MSC-A, MSC-I and MSC-T roles will be served by the same osmo-msc
instance,
- messages between MSC-A and MSC-{I,T} don't need to be routed via E-interface
(GSUP),
- no call routing between MSC-A and -I via MNCC necessary.
This is the largest code bomb I have submitted, ever. Out of principle, I
apologize to everyone trying to read this as a whole. Unfortunately, I see no
sense in trying to split this patch into smaller bits. It would be a huge
amount of work to introduce these changes in separate chunks, especially if
each should in turn be useful and pass all test suites. So, unfortunately, we
are stuck with this code bomb.
The following are some details and rationale for this rather huge refactoring:
* separate MSC subscriber management from ran_conn
struct ran_conn is reduced from the pivotal subscriber management entity it has
been so far to a mere storage for an SCCP connection ID and an MSC subscriber
reference.
The new pivotal subscriber management entity is struct msc_a -- struct msub
lists the msc_a, msc_i, msc_t roles, the vast majority of code paths however
use msc_a, since MSC-A is where all the interesting stuff happens.
Before handover, msc_i is an FSM implementation that encodes to the local
ran_conn. After inter-MSC Handover, msc_i is a compatible but different FSM
implementation that instead forwards via/from GSUP. Same goes for the msc_a
struct: if osmo-msc is the MSC-I "RAN proxy" for a remote MSC-A role, the
msc_a->fi is an FSM implementation that merely forwards via/from GSUP.
* New SCCP implementation for RAN access
To be able to forward BSSAP and RANAP messages via the GSUP interface, the
individual message layers need to be cleanly separated. The IuCS implementation
used until now (iu_client from libosmo-ranap) did not provide this level of
separation, and needed a complete rewrite. It was trivial to implement this in
such a way that both BSSAP and RANAP can be handled by the same SCCP code,
hence the new SCCP-RAN layer also replaces BSSAP handling.
sccp_ran.h: struct sccp_ran_inst provides an abstract handler for incoming RAN
connections. A set of callback functions provides implementation specific
details.
* RAN Abstraction (BSSAP vs. RANAP)
The common SCCP implementation did set the theme for the remaining refactoring:
make all other MSC code paths entirely RAN-implementation-agnostic.
ran_infra.c provides data structures that list RAN implementation specifics,
from logging to RAN de-/encoding to SCCP callbacks and timers. A ran_infra
pointer hence allows complete abstraction of RAN implementations:
- managing connected RAN peers (BSC, RNC) in ran_peer.c,
- classifying and de-/encoding RAN PDUs,
- recording connected LACs and cell IDs and sending out Paging requests to
matching RAN peers.
* RAN RESET now also for RANAP
ran_peer.c absorbs the reset_fsm from a_reset.c; in consequence, RANAP also
supports proper RESET semantics now. Hence osmo-hnbgw now also needs to provide
proper RESET handling, which it so far duly ignores. (TODO)
* RAN de-/encoding abstraction
The RAN abstraction mentioned above serves not only to separate RANAP and BSSAP
implementations transparently, but also to be able to optionally handle RAN on
distinct levels. Before Handover, all RAN messages are handled by the MSC-A
role. However, after an inter-MSC Handover, a standalone MSC-I will need to
decode RAN PDUs, at least in order to manage Assignment of RTP streams between
BSS/RNC and MNCC call forwarding.
ran_msg.h provides a common API with abstraction for:
- receiving events from RAN, i.e. passing RAN decode from the BSC/RNC and
MS/UE: struct ran_dec_msg represents RAN messages decoded from either BSSMAP
or RANAP;
- sending RAN events: ran_enc_msg is the counterpart to compose RAN messages
that should be encoded to either BSSMAP or RANAP and passed down to the
BSC/RNC and MS/UE.
The RAN-specific implementations are completely contained by ran_msg_a.c and
ran_msg_iu.c.
In particular, Assignment and Ciphering have so far been distinct code paths
for BSSAP and RANAP, with switch(via_ran){...} statements all over the place.
Using RAN_DEC_* and RAN_ENC_* abstractions, these are now completely unified.
Note that SGs does not qualify for RAN abstraction: the SGs interface always
remains with the MSC-A role, and SGs messages follow quite distinct semantics
from the fairly similar GERAN and UTRAN.
* MGW and RTP stream management
So far, managing MGW endpoints via MGCP was tightly glued in-between
GSM-04.08-CC on the one and MNCC on the other side. Prepare for switching RTP
streams between different RAN peers by moving to object-oriented
implementations: implement struct call_leg and struct rtp_stream with distinct
FSMs each. For MGW communication, use the osmo_mgcpc_ep API that has originated
from osmo-bsc and recently moved to libosmo-mgcp-client for this purpose.
Instead of implementing a sequence of events with code duplication for the RAN
and CN sides, the idea is to manage each RTP stream separately by firing and
receiving events as soon as codecs and RTP ports are negotiated, and letting
the individual FSMs take care of the MGW management "asynchronously". The
caller provides event IDs and an FSM instance that should be notified of RTP
stream setup progress. Hence it becomes possible to reconnect RTP streams from
one GSM-04.08-CC to another (inter-BSC Handover) or between CC and MNCC RTP
peers (inter-MSC Handover) without duplicating the MGCP code for each
transition.
The number of FSM implementations used for MGCP handling may seem a bit of an
overkill. But in fact, the number of perspectives on RTP forwarding are far
from trivial:
- an MGW endpoint is an entity with N connections, and MGCP "sessions" for
configuring them by talking to the MGW;
- an RTP stream is a remote peer connected to one of the endpoint's
connections, which is asynchronously notified of codec and RTP port choices;
- a call leg is the higher level view on either an MT or MO side of a voice
call, a combination of two RTP streams to forward between two remote peers.
BSC MGW PBX
CI CI
[MGW-endpoint]
[--rtp_stream--] [--rtp_stream--]
[----------------call_leg----------------]
* Use counts
Introduce using the new osmo_use_count API added to libosmocore for this
purpose. Each use token has a distinct name in the logging, which can be a
globally constant name or ad-hoc, like the local __func__ string constant. Use
in the new struct msc_a, as well as change vlr_subscr to the new osmo_use_count
API.
* FSM Timeouts
Introduce using the new osmo_tdef API, which provides a common VTY
implementation for all timer numbers, and FSM state transitions with the
correct timeout. Originated in osmo-bsc, recently moved to libosmocore.
Depends: Ife31e6798b4e728a23913179e346552a7dd338c0 (libosmocore)
Ib9af67b100c4583342a2103669732dab2e577b04 (libosmocore)
Id617265337f09dfb6ddfe111ef5e578cd3dc9f63 (libosmocore)
Ie9e2add7bbfae651c04e230d62e37cebeb91b0f5 (libosmo-sccp)
I26be5c4b06a680f25f19797407ab56a5a4880ddc (osmo-mgw)
Ida0e59f9a1f2dd18efea0a51680a67b69f141efa (osmo-mgw)
I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw)
Change-Id: I27e4988e0371808b512c757d2b52ada1615067bd
Avoid deprecation warning: use gsm48_decode_bcd_number2() instead of
gsm48_decode_bcd_number().
Validate the return value and add error handling.
Change-Id: Ibef71c46d72d2d43123e68f73e5ed554a69243d8
When the LU is accepted and the subscriber (vsub) is not claimed as "in
use" in the ref counting system.
- Make sure vlr_subscr_get() is called when the LU is accepted.
Change-Id: Iba90be095569cc5212c61ab8e8a9bfd4ae51fd44
Related OS#3934
This looks like a rudiment from OpenBSC, where we have:
#define BSC_API __attribute__((visibility("default")))
However, we don't use this attribute in OsmoMSC.
Change-Id: Ie2f18e9b47eca478f6e4702606068814546e34ce
In smpp_openbsc.c submit_to_sms(), "get" the appropriate use count upon
assigning sms->receiver, fixing a -1 use count upon sms_free().
Also, avoid a "put" of a NULL subscriber in the same function.
Related: OS#3930
Change-Id: Idaf01cd3cfa08088ce0d543d0576db957dc94262
So far, sms_pending_failed() starts a new sms_queue_trigger() run. The
intention behind that might have been to fill up the queue when sending SMS has
failed, but the practical effect is actually bad:
As current ttcn3-msc-test runs show, a failed MT SMS gets triggered multiple
times in short succession, i.e. osmo-msc repeatedly sends Paging Requests for
the same subscriber.
This special case happens actually only when there are few SMS still in the DB
to be delivered. In the TTCN3 test, there is exactly one MT SMS for one
subscriber, and retriggering the queue brings up the same SMS every time.
See f_tc_lu_and_mt_sms_paging_and_nothing() and f_tc_sgsap_mt_sms_and_nothing()
which say:
"/* Expect the MSC to page exactly 10 times before giving up */"
This is bad because an MSC should send a Paging Request exactly once. Retrying
failed Paging is clearly the task of the BSC, not the MSC. The remaining code
around Paging correctly follows this paradigm, but this retrigger doesn't.
Do not immediately trigger the SMS queue on a failed MT SMS. Instead, leave it
up to the periodical SMS queue trigger to decide.
This patch will cause the MT SMS tests in ttcn3-msc-tests to fail, because the
test expectations are bogus. The patch fixing the test run is listed 'Related'
below.
Related: I7dce12942a65eaaf97f78ca69401c7f93faacb9e (osmo-ttcn3-hacks)
Change-Id: I24bf9f1c1167efe1080ae4cf47ed2ef0bd981e49
Start using osmo_fsm_term_safely(true), the recently added feature of
libosmocore's fsm.c. Deallocates in slightly changed order and with slightly
modified logging. Adjust test expectations.
Depends: I8eda67540a1cd444491beb7856b9fcd0a3143b18 (libosmocore)
Change-Id: I195a719d9ec1f6764ee5a361244f59f0144dc253
The function sgs_tx() is using the sgs connection pointer as context,
even though it has done a check for a nullpointer in the line before.
This is very prone to lead into a segfault when the SGs connection dies.
Change-Id: I88b95e3f8cd35241ad68f08d94c6ad7067b842e6
Related: OS#3859
The libsmpp34 build_tlv() function is allocating dynamic memory
which we need to release again by calling destroy_tlv().
Change-Id: Iacc74c9948fb10fa79c0dd7b0cb72d4adbefdeed
Closes: OS#3912
If subscriber is NULL, vlr_subscr_msisdn_or_name() returns string
"unknown", which is less informative than printing destination msisdn
expected for the queued sms.
This happens for instance if an sms was queued with Store&Forward and
destination subscriber is not currently registered
Change-Id: I4b8b54c9c41b17d4e1fa7ece63aa91a98036ef11
When the subscriber is detached from SGs services (but not from 2g
services). Then the subscriber essentially becomes a regular 2g
subscriber, which means thet the lu expiration timer needs to be
started.
Change-Id: If95c63706dc1c5a537f7cd1b6481252427cbf234
Related: OS#3614
When the subscriber is detached from non EPS services while the
SGs-association is not SGs-NULL, it needs to be removed from the VLR
database.
Change-Id: I575cf6036ad39468f590b2d57a06cd3512a4c31c
Related: OS#3614
As we don't initialize all talloc contects of libmsc, let's make
sure that there is nothing left in the NULL context after the
unit test execution is finished.
Change-Id: I99fd82750aff376e4d90eaa2402ec41f4d59ef86
A memleak has been noticed after executing some of TTCN-3 test
cases. For example, the following ones:
- MSC_Tests.TC_lu_and_mo_sms,
- MSC_Tests.TC_lu_and_mt_sms.
The key point is that MSC_Tests.TC_lu_and_mo_sms basically sends
a MO SMS to a non-attached subscriber with MSISDN 12345, so this
message is getting stored in the SMSC's database.
As soon as the SMSC's queue is triggered, sms_submit_pending() would
retrieve pending messages from the database by calling function
smsq_take_next_sms() in loop and attempt to deliver them.
This function in it's turn checks whether the subscriber is attached
or not. If not, the allocated 'gsm_sms' structure would not be
free()ed! Therefore, every time smsq_take_next_sms() is called,
one 'gsm_sms' structure for an unattached subscriber is leaked.
Furthermore, there is a unit test called 'sms_queue_test', that
actually does cover smsq_take_next_sms() and was designed to
catch some potential memory leaks, but...
In order to avoid emulating the low-level SQLite API, the unit
test by design overwrites some functions of libmsc, including
db_sms_get_next_unsent_rr_msisdn(), that is being called by
smsq_take_next_sms().
The problem is that the original function in libmsc does
allocate a 'gsm_sms' structure on heap (using talloc), while
the overwriting function did this statically, returning a
pointer to stack. This critical difference made it impossible
to spot the memleak in smsq_take_next_sms() during the
unit test execution.
Let's refactor 'sms_queue_test' to use dynamic memory allocation,
and finally fix the evil memleak in smsq_take_next_sms().
Change-Id: Iad5e4d84d8d410ea43d5907e9ddf6e5fdb55bc7a
Closes: OS#3860
The default is [yes] alert-notifications, therefore write
"no alert-notifications" in the case that this has
been set, in order to preserve configuration after
write is called from vty.
Change-Id: I079aea96ee83fbf04f782dcab344d41a4ef04657
It was observed that the SGs server is started before
the actual VTY configuration is parsed. For example:
sgs
local-port 9999
local-ip 127.0.0.1
vlr-name vlr.example.net
produces the following debug output:
<0011> sgs_server.c:185 SGs socket bound to r=NULL<->l=0.0.0.0:29118
DLSS7 NOTICE <001e> osmo_ss7.c:1284 0: ASP Restart for server not implemented yet!
DSGS NOTICE <0011> sgs_server.c:185 SGs socket bound to r=NULL<->l=0.0.0.0:9999
DSGS NOTICE <0011> sgs_server.c:185 SGs socket bound to r=NULL<->l=127.0.0.1:9999
DMNCC DEBUG <0004> msc_main.c:604 Using internal MNCC handler.
The first startup is triggered by sgs_iface_init(), before reading
the VTY configuration, so the logging style is different. The next
two calls to sgs_server_open() are triggered during reading of the
VTY configuration by cfg_sgs_local_port() and cfg_sgs_local_ip().
Let's avoid starting the SGs server three times, and do it once,
after the VTY configuration is parsed. Also, keep the possibility
to change the binding parameters at run-time.
Change-Id: Ie0c31205ac48be7e50d0380a89833771b2708da4