aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2021-12-28phy_instance_destroy(): fix NULL pointer dereferenceHEADmasterVadim Yanitskiy1-4/+6
It's possible that a PHY instance has no associated TRX instance. Change-Id: Id028577ef92d1b3ce63ba62b63b8b29edb2ae5a4 Fixes: OS#5377
2021-12-14treewide: remove FSF address2021q4Oliver Smith1-4/+0
Remove the paragraph about writing to the Free Software Foundation's mailing address. The FSF has changed addresses in the past, and may do so again. In 2021 this is not useful, let's rather have a bit less boilerplate at the start of source files. Change-Id: I12b250e16426b125124def49e124e40ae93d0d58
2021-12-12cbch: cosmetic: use talloc_zero() in bts_process_smscb_cmd()Vadim Yanitskiy1-1/+1
Change-Id: Ibb5090b7cd53c1a485d0b9474c6470fa5831de67
2021-12-12osmo-bts-trx: fix a memleak in trx_sched_set_lchan()Vadim Yanitskiy1-0/+4
An additional burst buffer is allocated in rx_data_fn() for Uplink SACCH repetition, but unlike the main buffer it never gets free()d. Change-Id: I93310e7aed91a49c0511e2ed30e000440617f95a Fixes: I7e4cc33cc010866e41e3b594351a7f7bf93e08ac Related: OS#4795, SYS#5114
2021-12-12osmo-bts-trx: use l1ts as talloc context for burst buffersVadim Yanitskiy5-23/+10
Before a massive refactoring of the scheduler structures [1] it was impossible to have a clean hierarchy, in which the burst buffers get allocated as children of their parent timeslot structures. This change makes it easier to read talloc reports and simplifies memory management, offloading free()ing of the burst bufferes to talloc. Change-Id: Idb1ceaf83c433d2d0eb84d77c2187a00a657905c Related: [1] I7c4379e43a25e9d858d582a99bf6c4b65c9af481
2021-12-12oml: assign unique names to 'struct tlv_parsed' chunksVadim Yanitskiy1-0/+4
Change-Id: Ie248d2a65a11da914324a7e65b767b894ed8644a
2021-12-12oml: fix copy-pasted comments in oml_rx_set_*_attr()Vadim Yanitskiy1-4/+4
Change-Id: Iaa0e630b28b2a86312c069e7e55dd8bc5743759a
2021-12-12oml: use ts->trx as talloc-context in oml_rx_set_chan_attr()Vadim Yanitskiy1-1/+1
Change-Id: I6dc3706a486726afa9eec5d2e5369bd62c511455
2021-12-12oml: use proper talloc context in oml_rx_set_radio_attr()Vadim Yanitskiy1-1/+1
Change-Id: I69b6c72d85f68e4225d7786646f24259a5964998
2021-11-30debian/control: add osmo-bts meta-packageOliver Smith1-2/+17
Add an osmo-bts meta-package that depends on the subpackages osmo-bts-trx and osmo-bts-virtual. This is important, because the downstream packaging of debian does not split osmo-bts in subpackages, it has only one osmo-bts package. So without this, when installing "osmo-bts" in debian while having the Osmocom binary package repository installed and expecting to get the newer version, the user will actually get the older version from debian. Closes: OS#5319 Related: https://salsa.debian.org/debian-mobcom-team/osmo-bts/-/blob/master/debian/control Change-Id: I1a6edf2eb146616ac51b7957368a8d057caa685f
2021-11-25[lc15,oc2g,octphy] Fix memory leak on write queue overflowHarald Welte3-3/+16
Callers of osmo_wqueue_enqueue must always check the return value and free the message themselves in the error case. Change-Id: Ic67fbef23e419c0c8a7bcfb16d134e1bf649de72 Related: OS#5329
2021-11-24cbch: Fix dangling cur_msg leading to double-free in bts_cbch_reset()Harald Welte1-3/+3
If a new default message is installed via RSL, and the old default message is currently being transmitted, we must set cur_msg to NULL. The old default message must be talloc_free()d unconditionally whenever a new default message is being set. We can do that by using the TALLOC_FREE macro. Change-Id: Id32c2074b61cd1f09957b9d1558ffb3a7691a8e0 Related: OS#5325
2021-11-24cbch: Fix bts_smscb_state_reset() to avoid double-freeHarald Welte1-1/+4
If the currently transmitted message is the default message, bts_ss->cur_msg == bts_ss->derfault_msg. In this case we cannot simply talloc_free() both of them, as it would result in a boudle-free. Change-Id: I2d3645e34d31507b012a53ffe12d14223682f808 Closes: OS#5325 Fixes: Ib01d38c59ba9fa083fcc0682009c13d2db3664fe
2021-11-23bts: Properly free bts->shutdown_fi when struct gsm_bts is freedPau Espin Pedrol1-0/+4
Change-Id: I8c59f030c39c83a91b8504d4c5ce3b2058206fcf
2021-11-23bts_shutdown_fsm.h: Remove wrong comment describing enumPau Espin Pedrol1-1/+0
Change-Id: I963234255ce5a648a842ae6dbfd9966a694d0c09
2021-11-16Bump version: 1.3.0.348-f42287-dirty → 1.4.01.4.0Pau Espin Pedrol5-23/+391
Change-Id: Ibf3ce0bfd0cf67148229dd988ebde6e6b8d744cc
2021-11-12gsm_lchan_interf_meas_calc_avg(): adapt to the order of boundariesNeels Hofmeyr1-3/+12
The order of interference level boundaries is not clearly defined by 3GPP, so we should support both ascending and descending variants. Change-Id: I88d841d8d835bde8392c7b606b28c9070b7adc6e Related: SYS#5313
2021-11-11gsm_lchan_interf_meas_calc_avg(): fix band calculationVadim Yanitskiy1-2/+5
This patch makes osmo-bts bahave similar to ip.access nanoBTS. Change-Id: I1bcc6d6ba154f82aef95d05fb9af0eab490923c9 Related: SYS#5313
2021-11-11oml: use ARRAY_SIZE() in oml_rx_set_bts_attr()Vadim Yanitskiy1-1/+1
Change-Id: I89dab55e39fe49c8b4d3afb4b46d1b7f2bf3663c
2021-11-09rsl: exclude disabled timeslots from interference reportsVadim Yanitskiy3-0/+14
It may happen after the A-bis connection recovery that the RF RESource INDication message gets sent too early, while some timeslots are not yet configured. This confuses the BSC and provokes error messages. Change-Id: I00bc6fe67ea1bbedcd5d8640e73bd8b16b9e667f Related: SYS#5313, SYS#4971
2021-11-07common/Makefile.am: reformat {AM_CPPFLAGS,AM_CFLAGS,LDADD}Vadim Yanitskiy1-3/+18
Change-Id: If475999cccc215b6792f276b9cc6c494a0c1ad81
2021-11-05gsm_ts_release(): Make sure pchan{,is_want} is reset to NONEPau Espin Pedrol1-0/+4
Let's make sure all ts->pchan* related data is reset when we release the TS. This is important for pchan_is, since in bts-trx upon shutdown finish the PDCH lchan of a osmo dyn TS is set to lchan->state=LCHAN_S_NONE, and as a result when the PCU tries to release it by means of PCUIF act_req later, it may be ignored since the lchan is already in erased state. Related: SYS#4971 Change-Id: Ic7c34c0de23101ce094ffd03e00b4d2f6a551152
2021-11-04measurement: make use of gsm48_meas_res_is_valid()Vadim Yanitskiy1-3/+3
Change-Id: Iea6ab7b69defd7ce88c2aa92fcf2b84370a4c135 Depends: Iae2bd508a08c4b5093d36e514c22218763e11edf
2021-11-04[overpower] Turn it on and off depending on DL RxQualVadim Yanitskiy4-3/+57
Change-Id: Iaa812d4661ee17c4cd4a8c4ae4bd3e94c1a2e6cc Depends: Ia28293a12de0af71f55e701fb65c46e905dae217 Related: SYS#5319
2021-11-04measurement: pass *mr to lchan_bs_pwr_ctrl()Vadim Yanitskiy6-112/+19
As a side effect, we have to sacrifice a unit test (TC_inval_dummy) because it becomes impossible to pass a dummy or invalid SACCH block to lchan_bs_pwr_ctrl(). Change-Id: I937117cf26fb718d57920382f6972390ad498c51 Related: SYS#4918
2021-11-04measurement: pass *mr to repeated_dl_facch_active_decision()Vadim Yanitskiy1-10/+3
Change-Id: Idbf5f95d632aa2270c49b351ad5561ca2182bf9a Related: SYS#5114
2021-11-04measurement: get rid of *le in lchan_meas_handle_sacch()Vadim Yanitskiy1-6/+3
The current Timing Advance value can be obtained either from the L1 SACCH header (if received) or from lchan->ta_ctrl.current. Change-Id: I2b3693a0e49f03f2b4496c9dbd30cf47e9bc86b5
2021-11-04lchan_meas_handle_sacch(): check if Measurement Result is validVadim Yanitskiy1-1/+2
We should not rely on measurement data marked as invalid. Change-Id: I4aaac742674ce3ac15e9a4a32fe7c72db81d32d2
2021-11-04l1sap: rework handling of DATA.ind on SACCHVadim Yanitskiy4-74/+84
Currently an Uplink SACCH block is being passed to LAPDm first, and then gets forwareded to the BSC in handle_ms_meas_report(), together with the Uplink measurements collected so far. This approach has a serious flaw: handle_ms_meas_report() won't be called if an Uplink block contains SAPI=3 data (SMS) or was not decoded at all (len=0) fow whatever reason. Therefore, no RSL MEASurement RESult message will be sent to the BSC. Rename handle_ms_meas_report() to lchan_meas_handle_sacch(), and call it from l1sap_ph_data_ind(). This way perioduc RSL MEASurement RESult messages will be sent regardless of what happens on Uplink SACCH. Change-Id: Ifed91f87fd653debc87a09da3fd31ad64a13f330 Fixes: TC_meas_res_speech_{tchf,tchh}_sapi3 Related: SYS#5319
2021-11-03abis: Try one reconnect to previously connected BSC before trying next onePau Espin Pedrol1-0/+11
This way we keep all BTS connected to the same BSC if there was a spurious network problem. Related: SYS#4971 Change-Id: I16b75da5987584d099edc3a640f3a5cd61f3ad69
2021-11-03abis: Drop unneded if condition in else clausePau Espin Pedrol1-1/+1
priv->current_bsc will for sure be != last, otherwise it would have entered the if clause above it. Change-Id: I0a6519f7b93f0b45c67d19fef4f50daeeefc7340
2021-11-03bts-trx: sched_lchan_tchf: Change log level to debug for line informing ↵Pau Espin Pedrol1-1/+1
about missing dl prim This scenario is actually expected when the call is being set up or torn down, since we may not be receiving RTP from the MGW to send to the MS. Hence, let's lower the log level to DEBUG to avoid having log clogged for each call start/stop if INFO is used. Related: SYS#5676 Change-Id: Ib7f274b97cc66d671316eae429ee4baf16831534
2021-11-03scheduler: Fix FACCH msg with l2len==0 going to lower layers and logging errorsPau Espin Pedrol1-1/+1
Recent commit (see below) changed a check to avoid weird results for msg which had l2h=NULL, since l2len would return unexpected results there. However, some code branches like FACCH or TCH seem to be always setting l2h even if len=0. Hence, we must test either of the 2 scenarios (null pointer and l2len()). This commit fixes the following message appearing all the time during calls: """ TCH/F: Prim has odd len=0 != 23 """ Fixes: fb905b8d235ff2efe6c1cd9fde2b488b311c1cd7 Related: SYS#5676 Related: SYS#4919 Change-Id: I43152bc8484a35cd004d3303d3a6e6efcdefa890
2021-11-03bts-trx: sched_lchan_tchf: Drop impossible code pathPau Espin Pedrol1-6/+0
It cannot happen that msg1 is NULL and msg2 is not null, since they are deuqueed from same place. Only possible combinations are: msg1!=NULL && msg2!=NULL msg1!=NULL && msg2==NULL Change-Id: Ifd789844b1a7dfba596664de440d4c20b9f4c78f
2021-11-01osmo-bts-trx-calypso.cfg: Adjust settings to work with current osmo-bts versionsMartin Hauke1-4/+5
* Remove deprecated configuration options: 'osmotrx timing-advance-loop' is deprecated, Timing Advance loop is now active by default 'osmotrx ms-power-loop <-127-127>' is deprecated, MS Power Control is now managed by BSC * Adjust 'ipa unit-id' (1801 -> 6969) to match the one from the example osmo-bsc.cfg * Set clock advance values to known working values: + 'osmotrx fn-advance' -> 20 + 'osmotrx rts-advance' -> 5 * Set 'nominal-tx-power' since the calypso-bts tranceiver does not support NOMTXPOWER Change-Id: I69436b914cf1bec57f9fe7acea4a896e7c46b3a9
2021-10-28rsl: fix a memory leak in handle_gprs_susp_req()Vadim Yanitskiy1-0/+1
Change-Id: I65d9c12888aa5e5112680b3b3f38817e322ecc1c
2021-10-28l1sap: make 'l1sap' argument of process_l1sap_meas_data() constVadim Yanitskiy1-4/+4
Change-Id: Idc3004b0c74f7b98c96f20560c8b60a1fb4eb9c8
2021-10-27l1sap: process_l1sap_meas_data() accepts pointer to lchanVadim Yanitskiy1-27/+21
In 2/3 cases when calling process_l1sap_meas_data() we already have a pointer to the logical channel, so let's pass it as the first argument instead of a pointer to the transceiver. This way we avoid calling get_active_lchan_by_chan_nr() two times. In l1sap_ph_data_ind(), call process_l1sap_meas_data() below the conditional branch handling PDCH, so it won't be called for GSM_LCHAN_PDTCH anymore. GPRS specific measurements are handled by the PCU and not of interest for the BSC. Change-Id: I9de67a0b2d2b18923f2c2003b400387a0f1af411
2021-10-27l1sap: use designated initializers in process_l1sap_meas_data()Vadim Yanitskiy1-30/+24
Change-Id: I5169a6c5f6865655dbfebb6b68d5f67941d9cdb1
2021-10-27l1sap: move false PTCCH/U detection into PDCH branchVadim Yanitskiy1-9/+8
This check is only relevant for PDCH timeslots. Change-Id: I187fef8f3de0b41b502b0b18acfb11c56c5551f0
2021-10-27l1sap: fix handling of lchan->pending_rel_ind_msgVadim Yanitskiy1-11/+7
After merging the patch [1] fixing handling of the RLL RELease INDication message in lapdm_rll_tx_cb(), ttcn3-bts-test shows several regressions: pass->fail: BTS_Tests.TC_rll_rel_ind_DCCH_0 pass->fail: BTS_Tests.TC_rll_rel_ind_DCCH_3 pass->fail: BTS_Tests.TC_rll_rel_ind_ACCH_0 pass->fail: BTS_Tests.TC_rll_rel_ind_ACCH_3 pass->fail: BTS_Tests_LAPDm.TC_sabm_dm [1] I823c9101bcca72d5792e16379b02d3602ffc2726 991020c049c63768e147d49bd2918c2d2e0f6dcb The problem is actually *not* in patch [1], but in the older one [2] which we attempted to fix. While a logical channel is in signalling mode, the lower layers do not produce PRIM_TCH_RTS, and thus the l1sap_tch_rts_ind() is not being called. Unlike l1sap_tch_rts_ind(), the l1sap_ph_rts_ind() is being called regardless of the channel mode (signalling vs speech), so let's move handling of lchan->pending_rel_ind_msg there. Change-Id: I2c380f9045624f0a0a8f988bb207bc73d8354857 Fixes: [2] Ie4f70c75f0137b4bd72d579b3a32575bac2fca3
2021-10-26scheduler: Avoid crash upon call to trx_sched_set_lchan if l1ts is uninitializedPau Espin Pedrol1-0/+6
It could happen if for instance l1 code called trx_sched_clean() when closing the phy, and after that some code called (erroneously) trx_sched_set_lchan(). Let's make sure we don't allow other modules to crash the process when using this function. Related: SYS#4971 Change-Id: I93af7c3dcf0e34e9317eec0ee183dbf18b8d2f3b
2021-10-26l1sap: Avoid re-(de)activating already (de)active lchansPau Espin Pedrol1-0/+13
This avoids triggering all sorts of unexpected paths where one tries to release an already released lchan, etc. This can happen for instance if BTS shuts down due to BSC link going down, and hence resets all lchans, announcing it to the PCU. Then the PCU may try to deactivate the channel sending act_req (disable), but the BTS already unilaterally dropped the channels. That code path seems to trigger some crash, probably because something in lchan has been freed. Related: SYS#4971 Change-Id: I093e4d4e23b527b10bf5d6ff538460626c30a8f8
2021-10-25bts-trx: sched: tx_pdtch_fn: Drop log line clogging logsPau Espin Pedrol1-1/+5
The burst is properly pre-filled to _sched_dummy_burst in bts_sched_init_buffers(), so we are fine doing nothing on C0. Related: SYS#5676 Related: SYS#4919 Fixes: 300e31ed135c674cd44526b7503d4664a45a9ec3 Change-Id: Ia7045724a1a3206f5890c0b12843711ad2360ed8
2021-10-25scheduler: Fix check against empty PDCH blocksPau Espin Pedrol1-1/+1
msgb_l2len() will return nonzero values if msg->l2h is NULL. Let's check against msg->l2h being NULL instead. Related: SYS#5676 Related: SYS#4919 Fixes: 300e31ed135c674cd44526b7503d4664a45a9ec3 Change-Id: I18b61fcaa858a53887191a18d560c2e929478c64
2021-10-25Revert "bts-trx: sched: tx_pdtch_fn: Handle PCU idle blocks properly"Pau Espin Pedrol1-5/+2
This reverts commit 8a85b71167c4580ddf9e6c59ea602475fd462b3a. This is actually the wrong fix. Proper fix comes as a follow-up patch, since the root cause happens earlier in upper layers. No msg with l2h=NULL should reach this point of code ever. Change-Id: Ie23d94924f824bd7e89437ccf73b260f542477c6
2021-10-25bts-trx: sched: tx_pdtch_fn: Handle PCU idle blocks properlyPau Espin Pedrol1-2/+5
PCU idle blocks are identified by being 0 length, and hence msg->l2h being null. Don't try to encode those, but simply discard them silently. In case there's a missing block in C0, we want to log the event since the BTS is expected to send something on C0. Related: SYS#5676 Related: SYS#4919 Fixes: 300e31ed135c674cd44526b7503d4664a45a9ec3 Change-Id: I57e215fedeb415db4e67fdc56bf0f1410b5f7130
2021-10-25bts-trx: sched_lchan_pdtch: Refactor tx_pdtch_fn to get rid of goto tagPau Espin Pedrol1-12/+11
With this change the error case is moved at the end of the function, which is more usual. At the same time, one goto tag can be removed, simplifying the function. This is also a preparation for next patch improvinga bit the logic around same place. Related: SYS#5676 Related: SYS#4919 Change-Id: Ifbd95ccbebf4d810b1fe0a162722e63fe69106b8
2021-10-25[overpower] scheduler: handle {sacch,facch}_enabled flagsVadim Yanitskiy1-5/+5
The new [bit-]fields in the RSL_IE_OSMO_TEMP_OVP_ACCH_CAP allow more fine-grained control over the overpower feature, which can be enabled: * for both SACCH and FACCH, * for SACCH only, or * for FACCH only. Change-Id: Iaaab675a20bbefece832d913963c8c5ae32ff80c Depends: Ia28293a12de0af71f55e701fb65c46e905dae217 Related: SYS#5319
2021-10-25[overpower] lchan_dump_full_vty(): print overpower stateVadim Yanitskiy1-0/+31
Change-Id: I052f1d68b27b2dc7203835b4a93d40c94b0c8d82 Depends: Ia28293a12de0af71f55e701fb65c46e905dae217 Related: SYS#5319