Place all code related to the object into the related file.
Having all the data model in one file made sense in early stage of
development to make progress quickly, but nowadays it hurts more than
helps, due to constantly growing size and more and more bits being
added to the model, gaining in complexity.
Currently, having lots of different objects mixed up in gsm_data.h is a hole
of despair, where nobody can make any sense were to properly put new stuff
in, ending up with functions related to same object in different files
or with wrong prefixes, declarations of non-existing functions, etc.
because people cannot make up their mind on strict relation to objects
in the data model.
Splitting them in files really helps finding code operating on a
specific object and helping with logically splitting in the future.
Change-Id: I00c15f5285b5c1a0109279b7ab192d5467a04ece
There are a number of OML messages which are not seen on IP based
BTSs. Those are perfectly normal and expected on E1 based BTS.
Change-Id: Icd87fc9f3652b21f9d569af2572d080c9ac89e8b
Closes: OS#4665
Found while playing with "rf_locked 1" on a 2TRX setup with channel
allocator descend. After applying the setting, the 1st TRX is still used
to allocate the channels. After this patch is applied, the BSC correctly
allocates channels from TRX0.
Change-Id: I5201d2749363c9cbd0706177bde09117b163cbe3
This message may contain optional IEs (HSN, MAIO, ARFCN list),
so we cannot know the final length in advance. Let's set both
msg->{l2h,l3h} pointers and use msgb_l3len() to get the length.
Change-Id: I948ad4b847921324794a6eabd95d5583324da6e4
Related: OS#4545
3GPP TS 12.21 defines coding of 'ARFCN List' attribute as follows:
+---------------------------+--------------------+
| Attribute Identifier | 1st octet |
+---------------------------+--------------------+
| Length | 2-3 octets |
+---------------------------+--------------------+
| ARFCN1 | 4-5 octets |
+---------------------------+--------------------+
| ... | ... |
+---------------------------+--------------------+
| ARFCNn | (n * 2 - 3) octets |
+---------------------------+--------------------+
so this is basically TL16V, where L16 is the length of V.
In the Siemens dialect of OML coding rules are different though:
+---------------------------+--------------------+
| Attribute Identifier | 1st octet |
+---------------------------+--------------------+
| ARFCN count | 2nd octet |
+---------------------------+--------------------+
| ARFCN1 | 4-5 octets |
+---------------------------+--------------------+
| ... | ... |
+---------------------------+--------------------+
| ARFCNn | (n * 2 - 2) octets |
+---------------------------+--------------------+
so this is TCV, where C is the amount of ARFCNs in V.
This change fixes encoding of 'ARFCN List' for other dialects,
in particular encoding of the 'Length' field (1 vs 2 octets).
I verified the results in Wireshark (generic 3GPP TS 12.21
and ip.access dialect), everything looks good.
Change-Id: Iec1826f55459ac8e9355328a1a6bb0949874db60
Related: OS#4545
The value of the feature vector can not only be greater, but also
shorter than size of the buffer! This would potentially result
in a buffer overrun. Let's fix this.
Change-Id: I65e3228022865ea73de2e4821985df3097b9448b
Save OML failure reports from each BTS. Add a VTY command to display them
conveniently and optionally clear the list.
OsmoBSC> show bts 0 fail-rep
[2020-03-23 14:51:22] Type=processing failure, Severity=minor failure, Probable cause=Manufacturer specific values: Software warning, Additional text=test message sent from VTY
[2020-03-23 14:51:19] Type=processing failure, Severity=minor failure, Probable cause=Manufacturer specific values: Software warning, Additional text=test message sent from VTY
Related: OS#1605
Change-Id: I18aa17a721cd5eb1c98926dc2367229c0a50bc78
Separate raw input parsing from handling the failure report. A follow-up
patch will use the new parsing function to print saved failure reports
to the VTY.
While at it, put struct tlv_parsed inside struct nm_fail_rep_signal_data
instead of a pointer, so we don't need an additional alloc. Also add
error handling to the abis_nm_tlv_parse() call.
Related: OS#1605
Change-Id: Ia51004faf620aa4d40435d58c70d758c9d0054d8
Use the extra bts pointer instead of mb->trx->bts, which does not point
to an allocated bts.
Related: OS#1605
Change-Id: Ie61512f5690763fa380bdf0e7fb4763dbda019d2
Let's not just pass around the raw msgb, but also all other metadata,
such as the decoded parts of the TS 12.21 message.
As there's no current consumer of that signal, this creates no
compatibility issues.
Change-Id: I5d4d9d422b4e23348ffbe69c6e87a31d5574f90d
After sending of NM_MT_IPACC_RSL_CONNECT message, we start a timer,
and stop it on receipt of NM_MT_IPACC_RSL_CONNECT_{ACK,NACK}. When
running a multi-trx setup, one can see the following warnings:
DRSL NOTICE abis_nm.c:2852 (bts=0,trx=1) RSL connection request timed out
DRSL NOTICE abis_nm.c:2852 (bts=0,trx=2) RSL connection request timed out
even despite NM_MT_IPACC_RSL_CONNECT is actually being acknowledged.
The problem is in abis_nm_rx_ipacc(): we cannot just use sign_link->trx,
because the message itself was received over the OML link, so this
pointer always gives us C0/TRX0. Instead, we must find a TRX by its
number from the FOH header using gsm_bts_trx_by_nr().
Change-Id: Ib4b9a198da11c88a51cfa78ffb7e7235a6365ef4
The LOGPC() is usually used for continuation when printing complex
logging messages (e.g. where using format string is not enough).
In this case, nothing is being printed before calling LOGPC(), so
the logging messages appear without the meta info (time-stamp,
level, category, etc.), for example:
BTS 0 reported connected PCU version 0.7.0.1-2585-dirty
Change-Id: I868633ad3e50f2cb3ebfb2c566d16c4710f17563
It's quite ugly to have manual "bts=%d" printf-statements all over
the BSC code. Let's change this to use shared logging helper functions
all over the place, whenever we need to log something related to one
BTS or one TRX.
This can also help us as the first step to later add alternative logging
of BTS identities, e.g. by printing the Cell Global Identifier or
LAC+CI, or even a human-readable/vty-defined 'name' of the BTS, rather
than its numeric bts number. With this change in place, we can
introduce such changes at a single location in the code.
Change-Id: I4a7814d164384eecfb6913c31802cf2faead6e6c
the 'p_val' variable should represent he 'probabl cause value' and
nothing else. Let's use other local variables for other things.
Change-Id: Ie79279fc1d42cd57e7478820493b250fe38c2f4f
Before this patch, all OML GET ATTRIBUTES messages were encoded with
an erroneous OML header length value. The length value was always three
bytes less than the actual message length. This patch fixes the
problem.
Change-Id: I56068de0bb14a99ec39be587e542e27cddb7d1df
Closes: OS#3799
This signal can be used for tools willing to request and parse Attribute Response
and do something with the information. ipaccess-config tool will use
this signal in later patch Change-Id Ida416a969a3309868d6f4e50f34b34f224c32dd6.
Related: OS#3624
Change-Id: I9a121bbfe1b96904d4e16845abc90bb6ef20d2c9
Coverity points out that abis_nm_rcvmsg_sw() contains a switch
statement with suspicious looking missing break statements.
It is unclear to me if the code intends to process some
types of messages in more than one state, or of all messages
which affect a particular state already appear in the state's
corresponding switch block.
Can someone else tell what is supposed to happen here?
If this code is falling through intentionally, I will suggest
a patch adding /* fallthrough */ comments for clarity.
Change-Id: I1ea4221fadf30074156e9d17d94a5cb065242584
Related: CID#57703
Related: CID#57704
See following specs for related information:
* 3GPP TS 52.021 §8.11.3 Get Attribute Response
* 3GPP TS 52.021 §9.4.64 Get Attribute Response Info
Related: OS#3624
Change-Id: Ie61d70bc28427d5d879638516a36f590ce98bdc7
Spec compliant format is defined in:
* 3GPP TS 52.021 §8.11.3 "Get Attribute Response"
* 3GPP TS 52.021 §9.4.64 "Get Attribute Response Info".
On nanoBTS, however, reported attribute list is provided directly inside/after
the foh header instead of being enveloped inside the Get Attributes Response Info.
Furthermore, The Get Attributes Response Info can still appear and be at any position
in the reported attribute list, and it only contains the unreported
attribute ID list inside.
Change-Id: I81a613d53bddf432a79fa5cb0bf9d847b4bdee37
nanoBTS actually supports regular formatting. There are a few differences
with spec though:
* The attributes are listed directly in the message instead of being inside
the Get Attributes Response Info after the unsupported attribute ID list.
* The Get Attributes Response Info can be at any position in the
attribute list, and it only contains the unsupported attribute ID list.
As a result, parsing is currently split into 3 main parts or functions:
* Parsing regular (per spec) Get Attributes Response Info attr and get a
pointer to the list of attributes.
* A function that parses the list of attributes, called directly in case
of nanoBTS, and called by the former parser of Get Attributes Response
Info for regular (per spec) OML endpoints.
* A function to parse the unsupported attribute ID list, also used in the
first function to get a pointer to the list of attributes.
Related: OS#3624
Change-Id: I52e9f177c14fec1ec3f5c4ddb244594409008357
* Allow sending Get Attributes message in abis_nm_get_attr.
* Don't try to decode Get Attribute Response Info for nanoBTS, since it
uses a different formatting than the one defined in specs.
Related: OS#3624
Change-Id: I53d01e73791cf5450aa34b1ac8f051730e3a70f9
nanoBTS uses same format for the attribute list from Attribute Response
Info, but without using the later as an evelope. By splitting we can
later reuse the code handling the Attribute list.
While at it, take the chance to remove functions parse_attr_resp_info_*
which expect TLVs following an specific order, which is not mandatory.
Related: OS#3624
Change-Id: Iec7a6e4d27639d0e5adc0d9a01cd3c3e7a46f558
In future commits, nanoBTS support will be added, which implements its
own format not exactly equal to specs Attribute Response Info.
Related: OS#3624
Change-Id: I346dacc58faac70e6d224ca49484f9211cb8a046
This function will be merged into another using a "len" variable. This
change makes diffs easier to follow in future patches.
Change-Id: I1be03e31901ccf284e31237a04bb7448d1f22c07
Seen while operating a setup with 2 TRX (2 nanobts):
DNM <0004> abis_nm.c:703 OC=CHANNEL(03) INST=(00,01,04): bts=0 trx=0 Opstart ACK
DTS <0011> bts_ipaccess_nanobts.c:308 timeslot(0-0-4-TCH_F)[0x612000008aa0]{UNUSED}: Received Event TS_EV_OML_READY
DTS <0011> bts_ipaccess_nanobts.c:308 timeslot(0-0-4-TCH_F)[0x612000008aa0]{UNUSED}: Event TS_EV_OML_READY not permitte
As it can be seen, OML log header state correctly it comes from TRX1,
but later content in line expresses TRX0, and that failure is forwarded
up the stack and channels for TRX different than TRX0 are never
initialized.
Related: OS#3560
Fixes: f0ff9a6711
Change-Id: I27e992e419422051247777b048175b724c05323e
LOGPFOH uses the msgb through "foh", so we have to free msgb after
calling it, not before.
Fixes following ASAN report:
[0;m[1;36m20180822120155990 [1;34mDNM[0;m[1;36m <0004> abis_nm.c:1889 OC=CHANNEL(03) INST=(00,01,06): Set Chan Attr (bts=0,trx=1,ts=6)
[0;m=================================================================
==16465==ERROR: AddressSanitizer: heap-use-after-free on address 0x61a00002b3f0 at pc 0x7f587f44c0db bp 0x7ffc59e31df0 sp 0x7ffc59e31de8
READ of size 1 at 0x61a00002b3f0 thread T0
#0 0x7f587f44c0da in abis_nm_dump_foh libosmocore/src/gsm/abis_nm.c:937
#1 0x561e09e1532c in abis_nm_set_channel_attr osmo-bsc/src/osmo-bsc/abis_nm.c:1892
#2 0x561e09efd269 in nm_statechg_event osmo-bsc/src/osmo-bsc/bts_ipaccess_nanobts.c:168
#3 0x561e09efd269 in bts_ipa_nm_sig_cb osmo-bsc/src/osmo-bsc/bts_ipaccess_nanobts.c:335
#4 0x7f587efb3d16 in osmo_signal_dispatch libosmocore/src/signal.c:120
#5 0x561e09e18e31 in abis_nm_rx_statechg_rep osmo-bsc/src/osmo-bsc/abis_nm.c:255
#6 0x561e09e18e31 in abis_nm_rcvmsg_report osmo-bsc/src/osmo-bsc/abis_nm.c:380
#7 0x561e09e18e31 in abis_nm_rcvmsg_fom osmo-bsc/src/osmo-bsc/abis_nm.c:778
#8 0x561e09e1dc19 in abis_nm_rcvmsg osmo-bsc/src/osmo-bsc/abis_nm.c:926
#9 0x7f587ec90cc2 in handle_ts1_read input/ipaccess.c:274
#10 0x7f587ec90cc2 in ipaccess_fd_cb input/ipaccess.c:389
#11 0x7f587efb1ca8 in osmo_fd_disp_fds libosmocore/src/select.c:217
#12 0x7f587efb1ca8 in osmo_select_main libosmocore/src/select.c:257
#13 0x561e09e049d6 in main osmo-bsc/src/osmo-bsc/osmo_bsc_main.c:922
#14 0x7f587d53a2e0 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x202e0)
#15 0x561e09e054e9 in _start (/bin/osmo-bsc+0x34d4e9)
Fixes: OS#3494
Change-Id: I030117abfdcee387516a4dea7e1e6a9bae8055f6
Commit "dyn ts, bts_ipaccess_nanobts.c: init PDCH on Chan OPSTART ACK"
bf7099262a Icf6e25ff068e8a2600562d52726ead65e864ec02
introduced signal S_NM_OPSTART_ACK and passed the FOM header to identify the BTS
by. But the FOM header's BTS number is zero on each Abis/IP link, and the BTS
and TRX are actually identified by msgb->dst == e1inp_sign_link, member trx. So
the initial implementation associated *all* Channel OPSTART ACKs with BTS 0.
Pass the entire msgb as S_NM_OPSTART_ACK signal argument, implement a
abis_nm_get_ts() to retrieve the proper timeslot and use that during timeslot
init.
Related: OS#3351 OS#3205
Change-Id: I45ce5c24cb62d00f350df1af1be6c11104d74193
Move all of libbsc/ into osmo-bsc/, and separate/move some implementations to
allow linking from utils/* and ipaccess/* without pulling in unccessary
dependencies.
Some utilities use gsm_network and gsm_bts structs, which already include data
structures for fairly advanced uses. Move initialization that only osmo-bsc
needs into new bsc_network_init() and bsc_bts_alloc_register() functions, so
that the leaner tools can use the old gsm_* versions without the need to link
everything (e.g. handover and lchan alloc code).
In some instances, there need to be stubs if to cut off linking "just before
the RSL level" and prevent dependencies from creeping in.
- abis_rsl_rcvmsg(): the only program currently interpreting RSL messages is
osmo-bsc, the utils are merely concerned with OML, if at all.
- paging_flush_bts(): ip.access nanobts models call this when the RSL link is
dropped. Only osmo-bsc actually needs to do anything there.
- on_gsm_ts_init(): the mechanism to trigger timeslot initialization is related
to OML, while this action to take on init would pull in RSL dependencies.
utils/ and ipaccess/ each have a stubs.c file to implement these stubs. Tests
implement stubs inline where required.
From src/utils/, src/ipaccess/ and tests/*/, link in .o files from osmo-bsc/.
In order for this to work, the osmo-bsc subdir must be built before the other
source trees. (An alternative would be to include the .c files as sources, but
that would re-compile them in every source tree. Not a large burden really, but
unless linking .o files gives problems, let's have the quicker build.)
Minor obvious cleanups creep in with this patch, I will not bother to name them
individually now unless code review asks me to.
Rationale:
1) libbsc has been separate to use it for osmo-nitb and osmo-bsc in the old
openbsc.git. This is no longer required, and spreading over libbsc and osmo-bsc
is distracting.
2) Recently, ridiculous linking requirements have made adding new functions
cumbersome, because libbsc has started depending on osmo-bsc/*.c
implementations: on gscon FSM and bssap functions. For example, neither
bs11_config nor ipaccess-config nor bts_test need handover_cfg or BSSMAP
message composition. It makes no sense to link the entire osmo-bsc to it, nor
do we want to keep adding stubs to each linking realm.
Change-Id: I36a586726f5818121abe54d25654819fc451d3bf