aboutsummaryrefslogtreecommitdiffstats
path: root/openbsc/src/libbsc
AgeCommit message (Collapse)AuthorFilesLines
2018-01-30Work fix gcc maybe-uninitialized warnings/errorsDaniel Willmann3-6/+6
2017-11-27vty: Add cmd to configure 3g Early Classmark SendingPau Espin Pedrol3-3/+30
In state prior to this patch, "3G Early Classmark Sending Restriction" bit in SI3 rest octets was always set to H, which is a sane default as the policy to send the information is then controlled by "Early Classmark Sending Control" bit in the same octet. However, it seems Quortus SoftCore can have some issues decoding the option, so let's add a vty cmd to be able to disable it for those having any issues. Related: SYS#4021 Change-Id: Ic1afe071038a3bb5871d7ff40f665c8644f801ec
2017-11-27Use type bool for boolean fields in gsm48_si_ro_infoPau Espin Pedrol1-7/+7
Change-Id: Ic0981fca96f4927717ca335be1dab00a5d17fd6c
2017-11-01vty: skip installing cmds now always installed by defaultNeels Hofmeyr3-6/+0
vty_install_default() and install_default() will soon be deprecated. Depends: I5021c64a787b63314e0f2f1cba0b8fc7bff4f09b Change-Id: I4951982fc78ae167d8e16a672d7af44d703721a9
2017-10-23Make TRX rf locking more visibleMax3-7/+11
* log administrative state transitions * log what's caused it * while at it, mark boolean variable as such Cherry-picked from osmo-bsc be356ed32fbd28dc8d1460371d9e47511b20ac63 Change-Id: I3e25a19fac4d0b4886d825c9876771b1f66efe58 Related: SYS#3864
2017-10-23libbsc: Use correct printf formatting for uint64_tPau Espin Pedrol1-4/+4
unsigned long can be 32 bits on some arch/OS, while "current" field is always 64 bit because it's a uint64_t. Change-Id: I9719c69ef661bb46d8bb43cf8d6537c3e8d47826
2017-10-17bsc_api: Fix NULL secondary_lchan access in handle_ass_failPau Espin Pedrol1-2/+4
Related: OW#3893 Program terminated with signal SIGSEGV, Segmentation fault. 0 gsm_lchan_name (lchan=lchan@entry=0x0) at gsm_data_shared.c:342 (gdb) bt 0 gsm_lchan_name (lchan=lchan@entry=0x0) at gsm_data_shared.c:342 1 0x0805ab80 in lchan_release (lchan=0x0, sacch_deact=sacch_deact@entry=0, mode=mode@entry=RSL_REL_LOCAL_END) at chan_alloc.c:410 2 0x0805c1dd in handle_ass_fail (msg=0x94142b8, conn=0x9251048) at bsc_api.c:459 3 dispatch_dtap (msg=0x94142b8, link_id=0 '\000', conn=0x9251048) at bsc_api.c:598 4 gsm0408_rcvmsg (msg=msg@entry=0x94142b8, link_id=0 '\000') at bsc_api.c:658 5 0x08058ca2 in abis_rsl_rx_rll (msg=0x94142b8) at abis_rsl.c:1686 6 abis_rsl_rcvmsg (msg=0x94142b8) at abis_rsl.c:2097 7 0xb7e8cf9a in handle_ts1_read (bfd=0x94e8e08) at input/ipaccess.c:271 8 ipaccess_fd_cb (bfd=0x94e8e08, what=1) at input/ipaccess.c:386 9 0xb7ee8434 in osmo_select_main (polling=polling@entry=0) at select.c:158 10 0x0804bd7c in main (argc=6, argv=0xbfc27144) at osmo_bsc_main.c:272 (gdb) print lchan $2 = (const struct gsm_lchan *) 0x0 Possible scenario in which this crash can appear: 1- gsm0808_assign_req() calls handle_new_assignment() which sends an CHAN ACTIVATE msg and arms T10 timer. 2- ACTIVATE ACK is received (handle_chan_ack), which calls gsm48_send_rr_ass_cmd() which sends an ASSIGNMENT CMD, and doesn't disable/modify T10 timer. 3- T10 timeout is triggered (assignment_t10_timeout()), which sets conn->secondary_lchan = NULL 4- Immediately after, the ASSIGNMENT FAILURE message (which might have been already queued) is processed in handle_ass_fail, and then the crash occurs. This race condition is not an issue for handle_ass_compl() path because there's this check there which would trigger most probably if secondary_lchan is NULL: "if (conn->secondary_lchan != msg->lchan)" Change-Id: Ied5bd90b9c06f27135a2e3c46e40d49d27d9a387
2017-10-01Make sure BA-IND in all SI2xxx is '0' and in all SI5xxx is '1'Harald Welte2-3/+21
In masurement reports sent by the MS, this can then be used to correlate if a given measurement report was in response to a BCCH/neighbor list received on BCCH (SI2xxx) or on dowlink SACCH (SI5xxx). Closes: OS#2525 Change-Id: I1cd0dc51026dcd0e508e63eea4e333e6b184787a
2017-10-01libbsc: document arguments of generate_bcch_chan_list()Harald Welte1-1/+7
Change-Id: I5afc6e6a5a1d6b6a8ee73fdb60cc28074cf8585b
2017-09-27Show OML link uptime in vtyMax3-3/+29
Save the time when OML link to BTS was established and show it in vty. That's useful when troubleshooting issues like periodic/sporadic BTS restart. Related: SYS#3889 Change-Id: I9e4e8504afe8ca467b68d41826f61654e24d9600
2017-09-04handover_decision: Fix condition for power budget handover attemptIvan Kluchnikov1-1/+1
Handover attempt for power budget case should be performed every N SACCH frames, where N = Power Budget Interval. First measurement report with mr->nr = 0 was used for the first handover attempt in this case, which is not correct, because first usable report should have mr->nr = net->handover.pwr_interval-1. Moreover using the first measurement report with mr->nr = 0 for handover attempt could lead to unnecessary handover, because usually av_rxlev for first measurement report from MS is worse than for following reports. Change-Id: If7f54a4cb179eaa9e5eb147b9477633ac618e69e
2017-08-30SI13: drop PBCCH-related bitsMax2-44/+16
According to 3GPP TS 44.018 §1.8 the "network shall never enable PBCCH and PCCCH". Change-Id: I319e71a4b0c682361529e9c21377398a826b934b Related: OS#2400
2017-08-14Fix gsm_pchan2chan_nr() to use RSL_CHAN_OSMO_PDCHHarald Welte1-1/+1
When converting from GSM_PCHAN_PDCH, we should generate a RSL channel number IE with the osmocom extension RSL_CHAN_OSMO_PDCH rather than claiming it is a regular TCH/F channel. This is important as this function is used by osmo-bts, too - and it decides which channel number IE is put in the GSMTAP header for both GSMTAP tracing as well as the GSMTAP based osmo-bts-virtual. In order to avoid any unintended effect on libbsc, we make sure to modify rsl_ipacc_pdch_activate() to always use GSM_PCHAN_TCH_F in related RSL message. Change-Id: Ie34219e64a6d89da4a79f2db8ec73d1909fb8280
2017-08-13handover_logic: set correct link to bts for subscriber_connection in case of ↵Ivan Kluchnikov1-0/+1
moving this connection to another bts In case of successful completion of handover gsm_subscriber_connection could be moved from one bts to another, so connection link to bts should be replaced by link to bts, which owns new_lchan. This bug was detected, because conn->bts->nr is used in call control log messages and wrong number of bts was observed in these messages after handover. Change-Id: Idc7dd412b7580c451e716b73ef7549826c60b0d9
2017-08-09timer vty: also print the default value in cmd docNeels Hofmeyr1-13/+13
Rationale: allows seeing all timer defaults at once by doing OsmoBSC(config-net)# timer ? Before, defaults are visible only by doing on each timer: OsmoBSC(config-net)# timer t1234 <tab> Change-Id: I8259234e5c62e058dde56d531071440bbab11462
2017-08-09vty: add 'default' keyword to timer configNeels Hofmeyr1-3/+14
Change-Id: I4e837e8bedfad7ac4fd50048ecb016ddb37c2397
2017-08-09cosmetic: vty for timers: remove obsolete range checkNeels Hofmeyr1-6/+0
The VTY parsing already ensures the parameter range being 1..65535, no need to check the range again. Change-Id: I1cffa5b01cd5c589f1e42998e32135f1da8c960b
2017-07-20remove code disabling T3109 if configured to 0Harald Welte1-4/+0
We no longer permit timers with a 0 value, so this case can never happen. Also, if it should happen, I'd rather have a timter expiring immediately (and breaking something) than not being started in the first place. Change-Id: Ibfcdd3ddc0155caee89c501498329bde247621a0
2017-07-20bsc_vty: Don't allow timers of zero (0)Harald Welte1-2/+2
It typically doesn't make sense to configure any of the GSM RR timer to 0 (Seconds). In fact, accidentially configuring any of the timers to zero might have severe side effects, such as "stuck channels" described in https://osmocom.org/issues/2380 Change-Id: I517828f2f0c80ec01cb63648db2626f17a67fe57
2017-07-20GSM timers: User reasonable defaults; don't save if equal defaultHarald Welte2-13/+25
A number of the GSM timers (including T3109) had no reasonable default values if not specified in the VTY / config file. Together with unconditional writing to the config file, this created config files with a persistent setting for important timers as '0'. To make things worse, many of our example cofig files suffered from the same problem. Let's avoid this from happening by * having reasonable defaults if nothing specified in the config file * conditionally savingg timers only if they differ from default * reject any timer values that state zero during start-up (see previous commit) Change-Id: Iaac0bfca423852b61d8b9eb1438157ef00d0d8c8 Closes: OS#2380
2017-07-19bsc_vty: Add VTY command to test CTRL TRAP featureHarald Welte1-0/+16
Using this new command (introduced in OsmoBSC + OsmoNITB), you can simulate the generation of TRAP events for testin purposes. start the control interface monitor as an example client program: ./openbsc/contrib/bsc_control.py -m -d localhost -p 4249 then start OsmoBSC or OsmoNITB, telnet to the VTY and enter 'enable' mode and issue the following (example) command: ctrl-interface generate-trap my.foo.var 2342 As a result, on the bsc_control.py you will see: Got message: TRAP 0 my.foo.var 2342 Change-Id: Ib1d2ec38290dc94797c1b365d9b733e5215ab7d1
2017-07-16gsm_bts_trx_set_system_infos(): Disable non-existing SIHarald Welte1-6/+17
If we previously had a given SI present/active, we must send a zero-length BCCH FILLING for that SI type to the BTS to stop it from further transmitting this SI. Change-Id: I33e356e2fa3a69efac9080813e3e9ef4e6438ed1 Closes: OS#2368
2017-07-15RSL: Allow disabling of BCCH/SACCH filling for given SI typeHarald Welte1-4/+8
If we want to instruct the BTS to stop sending a given SI, we must be able to send the respective BCCH INFO / SACCH FILLING with a header but without any L3 data IE. This patch enables the related functions to do this whenever their data argument points to NULL. Change-Id: I88b85614951a108574f05db3b706884afe7e87a9
2017-07-15Fix regression causing loss of static system-information messagesHarald Welte1-2/+2
In commit 8b1a2f8cd7a81c6b8c7cdb0963dcf89de7c46100 we started to initialize bts->si_valid to 0. This means we are skipping the manually configured static system information. Instead, we have to initialize bts->si_valid to bts->si_mode_static, i.e. start with those that are static and not to be auto-generated. Found while developing http://git.osmocom.org/osmo-ttcn3-hacks/tree/sysinfo Change-Id: Iab9cc93cf6d54560a72cc393cc3721a8d10e04bf Closes: #2367
2017-07-12check for missing result of rate_ctr_group_alloc()Harald Welte1-0/+4
In case the counter group allocation fails, we must handle this gracefully and fail the allocation of the parent object, too. RelateD: OS#2361 Change-Id: I7dad4a4d52fe05f6b990359841b4408df5990e21
2017-07-11libbsc: Add VTY command to re-send the SYSTEM INFORMATION to BTSHarald Welte1-0/+33
This is useful if you are updating some configuration parameters which affect the content of the SYSTEM INFORMATION messages. Currently, we only send them at the time the RSL connection is established (i.e. when the BTS is initialized), so if you change something, you need to bring down and re-start the BTS. Using the newly-introduced "bts <0-255> resend-system-information" command, you can re-generate + re-send SYSTEM INFORMATION without bringing the BTS down, i.e. without any radio carrier downtime. Change-Id: I326df47de98f6d36c9a4d2d5475225d1e62bafb5
2017-07-11bsc_api: Fix copy+paste error in printing name of RR STATUS PDUHarald Welte1-1/+1
Change-Id: I0ef78ef046e4850346569f750693e12938b50ab5
2017-07-10Fix BTS attribute requestsMax2-3/+10
* fix BTS numbers: use 0 to indicate given BTS and 0xFF to indicate all BTS' as it's explained in 3GPP TS 52.021 §9.3. * only request attributes from supported (OsmoBTS) types Change-Id: I8f43055c38000248033a8ff9ddaf0910d68d794b Related: OS#2317
2017-06-22Fix BTS features length checkMax1-4/+4
While fixing potentially incorrect memory access, the check for maximum number of supported BTS features was incorrectly adjusted instead of feature vectore length check next to it. Fix this by adjusting checks properly and adding comments to avoid future confusion. The error was introduced in a60bb3dd28ce9e3720f8ee1b262893f3e233e2e6. Change-Id: I06d2498d730624d5da535f6add6fa98d004714ae
2017-06-18Add vty command "radio-link-timeout infinite" for uplink rx testingHarald Welte2-7/+36
When we are performing Rx sensitivity testing on a BTS, we want to deactivate the connection failure criterion / radio link timeout, i.e. no matter how many SACCH frames in uplink are failed to decode, the BTS should never close the channel. OsmoBTS Change-Id I736f21f6528db5c16fa80cdb905af20673797be5 covers a way how this behavior can be requested from the BTS via an OML attribute. This patch adds support to the BSC to actually set that attribute. Do not use this in production networks, as the BTS will keep open radio channels indefinitely even if the phone is gone and no longer transmitting anything. This is a pure testing feature. Change-Id: I6cb94e0f024934f7baeeb728ca9ed3042fbf16d2
2017-06-15bsc_init: Forget which SIs are valid for the trxMax2-4/+11
Previously the SI generation lead to setting the BCCH SIs for all TRX in a multi-trx setup. This is because we create the SIs globally but si_valid appears to be limited to the 'current' trx. Warn if we attempt to set SIs for the BCCH on a trx that does not have a BCCH. Change-Id: Ie0e288252a2e7709c4dae16b96a0b1512278847f Tweaked-by: Max <msuraev@sysmocom.de>
2017-06-15Update SI data structures and generationMax4-130/+186
To support segmented SI2quater as per 3GPP TS 44.018 we'll have to support multiple SI messages (up to 16 for SI2q) for a given type in contrast to existing 1:1 mapping: * expand storage space to hold up to 16 SI messages (spec limit) * add assertions for budget calculations * generate multiple SI2q messages * adjust SI2q-related tests * use precise check for number of SIq messages instead of approximate estimation Change-Id: Ic516ec9f0b821557d9461ae9f1c0afdd786f3b05 Related: OS#1660
2017-06-15Move SI-related definesMax2-3/+3
* move SI2quater related defines to shared header * add define from OsmoBTS which checks for presence of a given SI message in gsm_bts struct. Rename it to avoid conflicts with OsmoBTS code and to match naming conventions of similar macros. Change-Id: I11432c93c772d1ead6d45a7bb0f1d13d492c82f1 Related: OS#1660
2017-06-12OML: fix potential OOB memory accessMax1-2/+2
Use sizeof target BTS feature storage to make sure we always fit into pre-allocated memory. Also use it for log check. Change-Id: Ib107daa6e8b9bc397a10756071849f8ff82455d5 Fixes: CID 170581
2017-06-09Request and parse BTS feature list via OMLMax1-0/+40
Request features supported by BTS when getting attributes over OML. Change-Id: Ic35b2865998bca3c2c0cb4b77e4e73d12e08bd7e Related: OS#1614
2017-06-09Get TRX attributesMax2-37/+75
Request per-TRX attributes in addition to BTS attributes. Change-Id: I2b61131b9930afd03357c0b66947ee856d58cc46 Related: OS#1614
2017-06-09don't re-implement osmo_talloc_replace_string()Harald Welte1-1/+1
osmo_talloc_replace_string() was introducd into libosmocore in 2014, see commit f3c7e85d05f7b2b7bf093162b776f71b2bc6420d There's no reason for us to re-implement this as bsc_replace_string here. Change-Id: I6d2fcaabbc74730f6f491a2b2d5c784ccafc6602
2017-05-31Adjust BTS model feature checkMax1-2/+2
Rename gsm_bts_has_feature() -> gsm_btsmodel_has_feature() and adjust type signature to match gsm_btsmodel_set_feature() function and avoid confusion with upcoming functions to check/set BTS features reported over OML. Change-Id: I97abdedbef568e0c2fbd37c110f7d658cf20e100 Related: OS#1614
2017-05-29libbsc: Create pcu-socket only as specified in config fileHarald Welte2-14/+26
Since commit b4999b60d48bcbb5aa575973d068e07ab672e095 we created PCU sockets at hard-coded paths in the filesystem by default for all BTSs. This is inflexible and prevents the use of multiple BSC instances on a single filesystem, or the placement of the sockets in a more secure location than /tmp. The new approach with this patch is that * no PCU sockets are created by default * only for those BTSs where a 'pcu-socket' is configured via VTY, the socket will actually be created Change-Id: Ie9079470584777dcc31f85f9bf0808f479156ccb Closes: OS#2293
2017-05-28bsc_vty: Add command to manually issue IPAC MDCXHarald Welte1-1/+31
Using this command, one can modify the RTP stream associated with a given logical channel and (re)direct it to a specified IP:Port. Change-Id: I63e03b932038a4e2f6d51c5541b52e4a42df27bf
2017-05-27bsc_vty.c: Add command for manual [de]actiovation of logical channelsHarald Welte1-0/+119
Sometimes it is useful to manually activate (or decativate) a given logical channel from the VTY. Doing this on the BSC (rather than the BTS) ensures that the BSC knows that this timeslot / channel is allocated and there is no risk to have clashes between the BSC "owning" the resources and the BTS allocating some by itself. Change-Id: I44fc3904678eb48bd3ab1a3da8c0c265fa082e0d
2017-05-27bsc_vty.c: Further simplify vty_get_ts()Harald Welte1-6/+7
We can also move the string-to-numeric conversion inside vty_get_ts() to reduce the amount of work required in the caller. Change-Id: I2a74ed06e90e39d39f53fff39bb96df172728c0e
2017-05-27bsc_vty: Factor vty_get_ts() out of pdch_act_cmd()Harald Welte1-14/+26
Resolving a timeslot based on its numeric identities is a generally useful function, so lets' factor that out. Change-Id: Id2570232f82542487a1133be7efb1dc1eb3029a8
2017-05-27bsc_api.c: Documentation for handle_mr_config()Harald Welte1-2/+13
Change-Id: I6edd442afc7c20f8f097198941bc592d429ba5da
2017-05-27costmetic: Document gsm48_multirate_config() + const-ify inputHarald Welte1-1/+6
We generally use const pointers for input arguments. Also, document input/output arguments of function and add spec reference. Change-Id: I2532cde69a18e3b021f7371e68f67a28a43d8b5f
2017-05-25Use new e1inp_signal_names from libosmo-abis to print input signal namesHarald Welte2-7/+4
Change-Id: I156027ecdd1456228c9f8776577edd48e70c19da
2017-05-25abis_rsl: use msgb_pull to parse tlli from msgAlexander Couzens1-4/+2
Change-Id: I971bf6c8821689f9d8a1294a9b3bf1af9c4091f7
2017-05-25abis_rsl: fix off-by-one length check when parsing ericson tlli fieldAlexander Couzens1-1/+1
Change-Id: I658f6d82a67944345ddda5534fa996dca9e990ab
2017-05-25pcu_sock: Fix broken paging commandPhilipp Maier1-7/+20
The pcu sends us an already made up MAC-Block that contains the paging request. pcu_sock.c is parsing this paging request wrongly and fails silently, which results into a dropping of the request. This commit fixes the parsing problems. Change-Id: Iefef08123bdc351afd8287d3f27ebf0ae58a6e7d
2017-05-25pcu_sock: Forward imm.ass PCU originated messagesPhilipp Maier2-0/+38
The PCU sends imm.ass messages in response to a rach request. Those messages need to be forwarded to RSL in order to get them send. This commit introduces the required functionality for that Change-Id: Ice099c4ed7008200ed179e581aba1899c6c29455
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-13 01:37:46 +0000