cellular-infrastructure
/
osmo-bsc
mirror of https://gerrit.osmocom.org/osmo-bsc
9
0
Fork 0
Code Issues Releases Wiki Activity
osmo-bsc/tests/handover/handover_test.c

1818 lines
52 KiB
C
Raw Normal View History

HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
/*
* (C) 2013 by Andreas Eversberg <jolly@eversberg.eu>
* All Rights Reserved
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published by
* the Free Software Foundation; either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#include <stdio.h>
#include <stdlib.h>
#include <errno.h>
#include <assert.h>
#include <osmocom/core/application.h>
#include <osmocom/core/select.h>
#include <osmocom/core/talloc.h>
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
#include <osmocom/mgcp_client/mgcp_client_endpoint_fsm.h>
tests: handover_test.c: Add missing header Fixes following compilation warning: osmo-bsc/tests/handover/handover_test.c:48:83: warning: ‘struct mgcp_conn_peer’ declared inside parameter list will not be visible outside of this definition or declaration int __real_mgcp_conn_modify(struct osmo_fsm_inst *fi, uint32_t parent_evt, struct mgcp_conn_peer *conn_peer); ^~~~~~~~~~~~~~ Change-Id: I1b5e91adee3daaec71122a7691a214aa8240bce2
2018-03-17 00:16:54 +00:00
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
#include <osmocom/bsc/abis_rsl.h>
#include <osmocom/bsc/debug.h>
#include <osmocom/bsc/bsc_subscriber.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/lchan_select.h>
#include <osmocom/bsc/lchan_fsm.h>
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
#include <osmocom/bsc/handover_decision.h>
#include <osmocom/bsc/system_information.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/handover.h>
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
#include <osmocom/bsc/handover_cfg.h>
#include <osmocom/bsc/handover_decision_2.h>
#include <osmocom/bsc/bss.h>
cosmetic: rename bsc_api.h to gsm_08_08.h See also I91922f557072d0fb8cfe213a8a7b50f3bb23dea0, which renames osmo_bsc_api.c to gsm_08_08.c. Change-Id: I7179eb27183ee213f8fc8d548895b67aa43dc6a2
2018-07-24 16:10:05 +00:00
#include <osmocom/bsc/gsm_08_08.h>
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
#include <osmocom/bsc/osmo_bsc.h>
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
#include <osmocom/bsc/bsc_subscr_conn_fsm.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/timeslot_fsm.h>
#include <osmocom/bsc/lchan_fsm.h>
#include <osmocom/bsc/handover_fsm.h>
#include <osmocom/bsc/bsc_msc_data.h>
Move struct gsm_bts: gsm_data.* => bts.* 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
2020-07-15 18:53:16 +00:00
#include <osmocom/bsc/bts.h>
refactor paging: introduce bsc_paging_start() Allow starting a paging from elsewhere than a BSSMAP Paging Request. For upcoming Location Services (LCS), a BSSLAP TA Request from the SMLC may require triggering a Paging. Change-Id: Iaff91584699d163bd1963927280ff3a8ddd43073
2020-09-18 00:49:32 +00:00
#include <osmocom/bsc/paging.h>
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
ctx cleanup: use non-NULL talloc ctx for osmo_init_logging2() Fix various sanitizer complaints about memory leaks using a sanitizer build with gcc (Debian 7.3.0-12) 7.3.0. Also fix deprecation warnings on osmo_init_logging(). Depends: I216837780e9405fdaec8059c63d10699c695b360 (libosmocore) Change-Id: I970c6f8a0e36a8b63e42349dbc92baff649e5cef
2018-03-05 04:31:14 +00:00
void *ctx;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
struct gsm_network *bsc_gsmnet;
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
/* override, requires '-Wl,--wrap=osmo_mgcpc_ep_ci_request'.
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
* Catch modification of an MGCP connection. */
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
void __real_osmo_mgcpc_ep_ci_request(struct osmo_mgcpc_ep_ci *ci,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
enum mgcp_verb verb, const struct mgcp_conn_peer *verb_info,
struct osmo_fsm_inst *notify,
uint32_t event_success, uint32_t event_failure,
void *notify_data);
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
void __wrap_osmo_mgcpc_ep_ci_request(struct osmo_mgcpc_ep_ci *ci,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
enum mgcp_verb verb, const struct mgcp_conn_peer *verb_info,
struct osmo_fsm_inst *notify,
uint32_t event_success, uint32_t event_failure,
void *notify_data)
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
{
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
struct mgcp_conn_peer fake_data = {};
/* All MGCP shall be successful */
if (!notify)
return;
osmo_fsm_inst_dispatch(notify, event_success, &fake_data);
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
}
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
/* measurement report */
uint8_t meas_rep_ba = 0, meas_rep_valid = 1, meas_valid = 1, meas_multi_rep = 0;
uint8_t meas_dl_rxlev = 0, meas_dl_rxqual = 0;
uint8_t meas_ul_rxlev = 0, meas_ul_rxqual = 0;
uint8_t meas_tx_power_ms = 0, meas_tx_power_bs = 0, meas_ta_ms = 0;
uint8_t meas_dtx_ms = 0, meas_dtx_bs = 0, meas_nr = 0;
uint8_t meas_num_nc = 0, meas_rxlev_nc[6], meas_bsic_nc[6], meas_bcch_f_nc[6];
static void gen_meas_rep(struct gsm_lchan *lchan)
{
struct msgb *msg = msgb_alloc_headroom(256, 64, "RSL");
struct abis_rsl_dchan_hdr *dh;
uint8_t chan_nr = gsm_lchan2chan_nr(lchan);
uint8_t ulm[3], l1i[2], *buf;
struct gsm48_hdr *gh;
struct gsm48_meas_res *mr;
dh = (struct abis_rsl_dchan_hdr *) msgb_put(msg, sizeof(*dh));
dh->c.msg_discr = ABIS_RSL_MDISC_DED_CHAN;
dh->c.msg_type = RSL_MT_MEAS_RES;
dh->ie_chan = RSL_IE_CHAN_NR;
dh->chan_nr = chan_nr;
msgb_tv_put(msg, RSL_IE_MEAS_RES_NR, meas_nr++);
ulm[0] = meas_ul_rxlev | (meas_dtx_bs << 7);
ulm[1] = meas_ul_rxlev;
ulm[2] = (meas_ul_rxqual << 3) | meas_ul_rxqual;
msgb_tlv_put(msg, RSL_IE_UPLINK_MEAS, sizeof(ulm), ulm);
msgb_tv_put(msg, RSL_IE_BS_POWER, meas_tx_power_bs);
l1i[0] = 0;
l1i[1] = meas_ta_ms;
msgb_tv_fixed_put(msg, RSL_IE_L1_INFO, sizeof(l1i), l1i);
buf = msgb_put(msg, 3);
buf[0] = RSL_IE_L3_INFO;
buf[1] = (sizeof(*gh) + sizeof(*mr)) >> 8;
buf[2] = (sizeof(*gh) + sizeof(*mr)) & 0xff;
gh = (struct gsm48_hdr *) msgb_put(msg, sizeof(*gh));
mr = (struct gsm48_meas_res *) msgb_put(msg, sizeof(*mr));
gh->proto_discr = GSM48_PDISC_RR;
gh->msg_type = GSM48_MT_RR_MEAS_REP;
/* measurement results */
mr->rxlev_full = meas_dl_rxlev;
mr->rxlev_sub = meas_dl_rxlev;
mr->rxqual_full = meas_dl_rxqual;
mr->rxqual_sub = meas_dl_rxqual;
mr->dtx_used = meas_dtx_ms;
mr->ba_used = meas_rep_ba;
mr->meas_valid = !meas_valid; /* 0 = valid */
if (meas_rep_valid) {
mr->no_nc_n_hi = meas_num_nc >> 2;
mr->no_nc_n_lo = meas_num_nc & 3;
} else {
/* no results for serving cells */
mr->no_nc_n_hi = 1;
mr->no_nc_n_lo = 3;
}
mr->rxlev_nc1 = meas_rxlev_nc[0];
mr->rxlev_nc2_hi = meas_rxlev_nc[1] >> 1;
mr->rxlev_nc2_lo = meas_rxlev_nc[1] & 1;
mr->rxlev_nc3_hi = meas_rxlev_nc[2] >> 2;
mr->rxlev_nc3_lo = meas_rxlev_nc[2] & 3;
mr->rxlev_nc4_hi = meas_rxlev_nc[3] >> 3;
mr->rxlev_nc4_lo = meas_rxlev_nc[3] & 7;
mr->rxlev_nc5_hi = meas_rxlev_nc[4] >> 4;
mr->rxlev_nc5_lo = meas_rxlev_nc[4] & 15;
mr->rxlev_nc6_hi = meas_rxlev_nc[5] >> 5;
mr->rxlev_nc6_lo = meas_rxlev_nc[5] & 31;
mr->bsic_nc1_hi = meas_bsic_nc[0] >> 3;
mr->bsic_nc1_lo = meas_bsic_nc[0] & 7;
mr->bsic_nc2_hi = meas_bsic_nc[1] >> 4;
mr->bsic_nc2_lo = meas_bsic_nc[1] & 15;
mr->bsic_nc3_hi = meas_bsic_nc[2] >> 5;
mr->bsic_nc3_lo = meas_bsic_nc[2] & 31;
mr->bsic_nc4 = meas_bsic_nc[3];
mr->bsic_nc5 = meas_bsic_nc[4];
mr->bsic_nc6 = meas_bsic_nc[5];
mr->bcch_f_nc1 = meas_bcch_f_nc[0];
mr->bcch_f_nc2 = meas_bcch_f_nc[1];
mr->bcch_f_nc3 = meas_bcch_f_nc[2];
mr->bcch_f_nc4 = meas_bcch_f_nc[3];
mr->bcch_f_nc5_hi = meas_bcch_f_nc[4] >> 1;
mr->bcch_f_nc5_lo = meas_bcch_f_nc[4] & 1;
mr->bcch_f_nc6_hi = meas_bcch_f_nc[5] >> 2;
mr->bcch_f_nc6_lo = meas_bcch_f_nc[5] & 3;
msg->dst = lchan->ts->trx->bts->c0->rsl_link;
msg->l2h = (unsigned char *)dh;
msg->l3h = (unsigned char *)gh;
abis_rsl_rcvmsg(msg);
}
static struct gsm_bts *create_bts(int arfcn)
{
struct gsm_bts *bts;
struct e1inp_sign_link *rsl_link;
int i;
handover_test: use 'unknown' BTS type instead of 'sysmobts' This unit test does not really need a BTS of such specific type. Change-Id: Id676042518d06e94a9fb20112334280e2b91074b
2020-05-30 18:56:20 +00:00
bts = bsc_bts_alloc_register(bsc_gsmnet, GSM_BTS_TYPE_UNKNOWN, 0x3f);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
if (!bts) {
printf("No resource for bts1\n");
return NULL;
}
bts->location_area_code = 23;
bts->c0->arfcn = arfcn;
bts->codec.efr = 1;
bts->codec.hr = 1;
bts->codec.amr = 1;
ctx cleanup: use non-NULL talloc ctx for osmo_init_logging2() Fix various sanitizer complaints about memory leaks using a sanitizer build with gcc (Debian 7.3.0-12) 7.3.0. Also fix deprecation warnings on osmo_init_logging(). Depends: I216837780e9405fdaec8059c63d10699c695b360 (libosmocore) Change-Id: I970c6f8a0e36a8b63e42349dbc92baff649e5cef
2018-03-05 04:31:14 +00:00
rsl_link = talloc_zero(ctx, struct e1inp_sign_link);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
rsl_link->trx = bts->c0;
bts->c0->rsl_link = rsl_link;
bts->c0->mo.nm_state.operational = NM_OPSTATE_ENABLED;
bts->c0->mo.nm_state.availability = NM_AVSTATE_OK;
Avoid selecting channels from administratively locked trx 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
2020-06-22 15:57:01 +00:00
bts->c0->mo.nm_state.administrative = NM_STATE_UNLOCKED;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
bts->c0->bb_transc.mo.nm_state.operational = NM_OPSTATE_ENABLED;
bts->c0->bb_transc.mo.nm_state.availability = NM_AVSTATE_OK;
Avoid selecting channels from administratively locked trx 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
2020-06-22 15:57:01 +00:00
bts->c0->bb_transc.mo.nm_state.administrative = NM_STATE_UNLOCKED;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
/* 4 full rate and 4 half rate channels */
for (i = 1; i <= 6; i++) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
bts->c0->ts[i].pchan_from_config = (i < 5) ? GSM_PCHAN_TCH_F : GSM_PCHAN_TCH_H;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
bts->c0->ts[i].mo.nm_state.operational = NM_OPSTATE_ENABLED;
bts->c0->ts[i].mo.nm_state.availability = NM_AVSTATE_OK;
Avoid selecting channels from administratively locked trx 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
2020-06-22 15:57:01 +00:00
bts->c0->ts[i].mo.nm_state.administrative = NM_STATE_UNLOCKED;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
for (i = 0; i < ARRAY_SIZE(bts->c0->ts); i++) {
/* make sure ts->lchans[] get initialized */
fix nanobts: timeslot FSM: use flags to remember OML,RSL status Before this patch, the timeslot FSM receives OML and RSL ready events. Afterwards, it relies on examining the RSL and OML status to match the received events. This doesn't work for the ip.access nanobts, which fails to change the CHANNEL OM's operational status even though it has sent an Opstart ACK. We receive OML CHANNEL Opstart ACK, but the mo's state left at OP_STATE=Disabled. We apparently cannot rely on the gsm_abis_mo state as assumed before this patch, since changing the state depends on each BTS vendor's OML implementation. Also, implementation wise, it is better to not include assumptions on RSL and OML implementations in the timeslot FSM. Simply receive the OML and RSL ready events and remember that they arrived in dedicated flags. Remove the no longer needed oml_is_ts_ready() callback from struct gsm_bts_model added in: commit 91aa68f762218906e45be4817c6ea54b480da5e1 "dyn TS: init only when both RSL and the Channel OM are established" I99f29d2ba079f6f4b77f0af12d9784588d2f56b3 This keeps osmo-bts operational while fixing ip.access nanobts, where the CHANNEL OM's state prevented the timeslot FSM from entering operation. Change-Id: I4843d03b3237cdcca0ad2041ef6895ff253d8419
2018-07-26 18:33:15 +00:00
osmo_fsm_inst_dispatch(bts->c0->ts[i].fi, TS_EV_RSL_READY, 0);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
osmo_fsm_inst_dispatch(bts->c0->ts[i].fi, TS_EV_OML_READY, 0);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
}
return bts;
}
void create_conn(struct gsm_lchan *lchan)
{
cosmetic: handover_test: add IMSI to subscr for logging Change-Id: I52fa8f3183c93ed362d8252d20daf00c60a117c8
2018-05-24 16:43:58 +00:00
static unsigned int next_imsi = 0;
char imsi[sizeof(lchan->conn->bsub->imsi)];
struct gsm_network *net = lchan->ts->trx->bts->network;
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
struct gsm_subscriber_connection *conn;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
struct mgcp_client *fake_mgcp_client = (void*)talloc_zero(net, int);
cosmetic: handover_test: add IMSI to subscr for logging Change-Id: I52fa8f3183c93ed362d8252d20daf00c60a117c8
2018-05-24 16:43:58 +00:00
conn = bsc_subscr_con_allocate(net);
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
conn->user_plane.mgw_endpoint = osmo_mgcpc_ep_alloc(conn->fi,
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
GSCON_EV_FORGET_MGW_ENDPOINT,
move mgw endpoint FSM to osmo-mgw.git osmo-mgw.git also includes fixes of the MGW endpoint FSM, for example I92a9944acc96398acd6649f9c3c5badec5dd6dcc. Depends: I9a3effd38e72841529df6c135c077116981dea36 (osmo-mgw) Change-Id: I03e6b48d9b0a5370310d5f56809259ff7909cf9d
2019-04-23 16:37:37 +00:00
fake_mgcp_client,
net->mgw.tdefs,
"test",
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
"fake endpoint");
handover_test: Properly allocate MSC data struct. Without a properly allocated MSC struct, touching counters crashes the test. Change-Id: I7498d2891259be9b532845627f14ac75e98e703e
2020-05-10 19:48:01 +00:00
conn->sccp.msc = osmo_msc_data_alloc(net, 0);
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
lchan->conn = conn;
conn->lchan = lchan;
cosmetic: handover_test: add IMSI to subscr for logging Change-Id: I52fa8f3183c93ed362d8252d20daf00c60a117c8
2018-05-24 16:43:58 +00:00
/* Make up a new IMSI for this test, for logging the subscriber */
next_imsi ++;
snprintf(imsi, sizeof(imsi), "%06u", next_imsi);
introduce osmo_use_count for bsc_subscr During LCS development, I'm getting use count bugs and would like to see use token strings to figure it out. Change-Id: I29bf60059d4cf7bb99a00753e6cdc149baf95f94
2020-09-15 01:03:58 +00:00
lchan->conn->bsub = bsc_subscr_find_or_create_by_imsi(net->bsc_subscribers, imsi, BSUB_USE_CONN);
cosmetic: handover_test: add IMSI to subscr for logging Change-Id: I52fa8f3183c93ed362d8252d20daf00c60a117c8
2018-05-24 16:43:58 +00:00
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
/* kick the FSM from INIT through to the ACTIVE state */
osmo_fsm_inst_dispatch(conn->fi, GSCON_EV_A_CONN_REQ, NULL);
osmo_fsm_inst_dispatch(conn->fi, GSCON_EV_A_CONN_CFM, NULL);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
}
/* create lchan */
struct gsm_lchan *create_lchan(struct gsm_bts *bts, int full_rate, char *codec)
{
struct gsm_lchan *lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan = lchan_select_by_type(bts, (full_rate) ? GSM_LCHAN_TCH_F : GSM_LCHAN_TCH_H);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
if (!lchan) {
printf("No resource for lchan\n");
exit(EXIT_FAILURE);
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* serious hack into osmo_fsm */
lchan->fi->state = LCHAN_ST_ESTABLISHED;
lchan->ts->fi->state = TS_ST_IN_USE;
LOG_LCHAN(lchan, LOGL_DEBUG, "activated by handover_test.c\n");
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
create_conn(lchan);
if (!strcasecmp(codec, "FR") && full_rate)
lchan->tch_mode = GSM48_CMODE_SPEECH_V1;
else if (!strcasecmp(codec, "HR") && !full_rate)
lchan->tch_mode = GSM48_CMODE_SPEECH_V1;
else if (!strcasecmp(codec, "EFR") && full_rate)
lchan->tch_mode = GSM48_CMODE_SPEECH_EFR;
handover_fsm: copy old S15_S0 to new lchan When a new lchan is selected during handover, some of the properties of the old lchan are inherited by the new lchan. At the moment S15-S0 is not not inherited so that the value for those bits will always be 0x0000 for the new lchan. Since those bits also define the active set AMR codec the channel activation will fail because 0x0000 is invalid (active set with zero rates) Change-Id: Ifd470397e99985394634da1bb13ccfc5041984d2 Related: OS#3503
2019-03-11 13:10:26 +00:00
else if (!strcasecmp(codec, "AMR")) {
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
lchan->tch_mode = GSM48_CMODE_SPEECH_AMR;
handover_fsm: copy old S15_S0 to new lchan When a new lchan is selected during handover, some of the properties of the old lchan are inherited by the new lchan. At the moment S15-S0 is not not inherited so that the value for those bits will always be 0x0000 for the new lchan. Since those bits also define the active set AMR codec the channel activation will fail because 0x0000 is invalid (active set with zero rates) Change-Id: Ifd470397e99985394634da1bb13ccfc5041984d2 Related: OS#3503
2019-03-11 13:10:26 +00:00
lchan->activate.info.s15_s0 = 0x0002;
} else {
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
printf("Given codec unknown\n");
exit(EXIT_FAILURE);
}
lchan->conn->codec_list = (struct gsm0808_speech_codec_list){
.codec = {
{ .fi=true, .type=GSM0808_SCT_FR1, },
{ .fi=true, .type=GSM0808_SCT_FR2, },
{ .fi=true, .type=GSM0808_SCT_FR3, },
{ .fi=true, .type=GSM0808_SCT_HR1, },
{ .fi=true, .type=GSM0808_SCT_HR3, },
},
.len = 5,
};
return lchan;
}
/* parse channel request */
static int got_chan_req = 0;
static struct gsm_lchan *chan_req_lchan = NULL;
static int parse_chan_act(struct gsm_lchan *lchan, uint8_t *data)
{
chan_req_lchan = lchan;
return 0;
}
static int parse_chan_rel(struct gsm_lchan *lchan, uint8_t *data)
{
chan_req_lchan = lchan;
return 0;
}
/* parse handover request */
static int got_ho_req = 0;
static struct gsm_lchan *ho_req_lchan = NULL;
static int parse_ho_command(struct gsm_lchan *lchan, uint8_t *data, int len)
{
struct gsm48_hdr *gh = (struct gsm48_hdr *) data;
struct gsm48_ho_cmd *ho = (struct gsm48_ho_cmd *) gh->data;
int arfcn;
struct gsm_bts *neigh;
switch (gh->msg_type) {
case GSM48_MT_RR_HANDO_CMD:
arfcn = (ho->cell_desc.arfcn_hi << 8) | ho->cell_desc.arfcn_lo;
/* look up trx. since every dummy bts uses different arfcn and
* only one trx, it is simple */
llist_for_each_entry(neigh, &bsc_gsmnet->bts_list, list) {
if (neigh->c0->arfcn != arfcn)
continue;
ho_req_lchan = lchan;
return 0;
}
break;
case GSM48_MT_RR_ASS_CMD:
ho_req_lchan = lchan;
return 0;
break;
default:
fprintf(stderr, "Error, expecting HO or AS command\n");
return -EINVAL;
}
return -1;
}
/* send channel activation ack */
static void send_chan_act_ack(struct gsm_lchan *lchan, int act)
{
struct msgb *msg = msgb_alloc_headroom(256, 64, "RSL");
struct abis_rsl_dchan_hdr *dh;
dh = (struct abis_rsl_dchan_hdr *) msgb_put(msg, sizeof(*dh));
dh->c.msg_discr = ABIS_RSL_MDISC_DED_CHAN;
dh->c.msg_type = (act) ? RSL_MT_CHAN_ACTIV_ACK : RSL_MT_RF_CHAN_REL_ACK;
dh->ie_chan = RSL_IE_CHAN_NR;
dh->chan_nr = gsm_lchan2chan_nr(lchan);
msg->dst = lchan->ts->trx->bts->c0->rsl_link;
msg->l2h = (unsigned char *)dh;
abis_rsl_rcvmsg(msg);
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* Send RLL Est Ind for SAPI[0] */
static void send_est_ind(struct gsm_lchan *lchan)
{
struct msgb *msg = msgb_alloc_headroom(256, 64, "RSL");
struct abis_rsl_rll_hdr *rh;
uint8_t chan_nr = gsm_lchan2chan_nr(lchan);
rh = (struct abis_rsl_rll_hdr *) msgb_put(msg, sizeof(*rh));
rh->c.msg_discr = ABIS_RSL_MDISC_RLL;
rh->c.msg_type = RSL_MT_EST_IND;
rh->ie_chan = RSL_IE_CHAN_NR;
rh->chan_nr = chan_nr;
rh->ie_link_id = RSL_IE_LINK_IDENT;
rh->link_id = 0x00;
msg->dst = lchan->ts->trx->bts->c0->rsl_link;
msg->l2h = (unsigned char *)rh;
abis_rsl_rcvmsg(msg);
}
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
/* send handover complete */
static void send_ho_complete(struct gsm_lchan *lchan, bool success)
{
struct msgb *msg = msgb_alloc_headroom(256, 64, "RSL");
struct abis_rsl_rll_hdr *rh;
uint8_t chan_nr = gsm_lchan2chan_nr(lchan);
uint8_t *buf;
struct gsm48_hdr *gh;
struct gsm48_ho_cpl *hc;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
send_est_ind(lchan);
lchan_fsm: split off lchan_rtp_fsm, establish RTP a bit earlier Change-Id: Id7a4407d9b63be05ce63f5f2768b7d7e3d5c86fb
2018-07-12 19:23:26 +00:00
osmo_fsm_inst_dispatch(lchan->fi, LCHAN_EV_RTP_READY, 0);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
rh = (struct abis_rsl_rll_hdr *) msgb_put(msg, sizeof(*rh));
rh->c.msg_discr = ABIS_RSL_MDISC_RLL;
rh->c.msg_type = RSL_MT_DATA_IND;
rh->ie_chan = RSL_IE_CHAN_NR;
rh->chan_nr = chan_nr;
rh->ie_link_id = RSL_IE_LINK_IDENT;
rh->link_id = 0x00;
buf = msgb_put(msg, 3);
buf[0] = RSL_IE_L3_INFO;
buf[1] = (sizeof(*gh) + sizeof(*hc)) >> 8;
buf[2] = (sizeof(*gh) + sizeof(*hc)) & 0xff;
gh = (struct gsm48_hdr *) msgb_put(msg, sizeof(*gh));
hc = (struct gsm48_ho_cpl *) msgb_put(msg, sizeof(*hc));
gh->proto_discr = GSM48_PDISC_RR;
gh->msg_type =
success ? GSM48_MT_RR_HANDO_COMPL : GSM48_MT_RR_HANDO_FAIL;
msg->dst = lchan->ts->trx->bts->c0->rsl_link;
msg->l2h = (unsigned char *)rh;
msg->l3h = (unsigned char *)gh;
abis_rsl_rcvmsg(msg);
}
handover_test: explicitly wrap abis_rsl_sendmsg() Until now, handover_test simply defines an abis_rsl_sendmsg() function to override the one from libosmo-babis. To me this amounts to implicit linker magic (aka luck). The common pattern we're using for this scenario is to explicitly wrap an existing function by instructing the linker. Also do this for abis_rsl_sendmsg() in handover_test.c to clarify. Change-Id: I05ce33a4e42bc82b9ce5060e2f811033f5a1f8d8
2018-03-05 20:53:18 +00:00
/* override, requires '-Wl,--wrap=abis_rsl_sendmsg'.
* Catch RSL messages sent towards the BTS. */
int __real_abis_rsl_sendmsg(struct msgb *msg);
int __wrap_abis_rsl_sendmsg(struct msgb *msg)
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
{
struct abis_rsl_dchan_hdr *dh = (struct abis_rsl_dchan_hdr *) msg->data;
struct e1inp_sign_link *sign_link = msg->dst;
int rc;
struct gsm_lchan *lchan = rsl_lchan_lookup(sign_link->trx, dh->chan_nr, &rc);
if (rc) {
printf("rsl_lchan_lookup() failed\n");
exit(1);
}
switch (dh->c.msg_type) {
case RSL_MT_CHAN_ACTIV:
rc = parse_chan_act(lchan, dh->data);
if (rc == 0)
got_chan_req = 1;
break;
case RSL_MT_RF_CHAN_REL:
rc = parse_chan_rel(lchan, dh->data);
if (rc == 0)
send_chan_act_ack(chan_req_lchan, 0);
break;
case RSL_MT_DATA_REQ:
rc = parse_ho_command(lchan, msg->l3h, msgb_l3len(msg));
if (rc == 0)
got_ho_req = 1;
break;
case RSL_MT_IPAC_CRCX:
break;
lchan release: always Deact SACCH If an lchan is being released and had a SACCH active, there is no reason to omit the Deact SACCH message ever. All of the callers that passed do_deact_sacch = false did so for no good reason. Drop the do_deact_sacch flag everywhere and, when the lchan type matches and SAPI[0] is still active, simply always send a Deact SACCH message. The do_deact_sacch flag was carried over from legacy code, by me, mainly because I never really understood why it was there. I do hope I'm correct now, asserting that having this flag makes no sense. Change-Id: Id3301df059582da2377ef82feae554e94fa42035
2018-11-06 21:24:07 +00:00
case RSL_MT_DEACTIVATE_SACCH:
break;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
default:
printf("unknown rsl message=0x%x\n", dh->c.msg_type);
}
return 0;
}
/* test cases */
static char *test_case_0[] = {
"2",
"Stay in better cell\n\n"
"There are many neighbor cells, but only the current cell is the best\n"
"cell, so no handover is performed\n",
"create-bts", "7",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "30","0",
"6","0","20","1","21","2","18","3","20","4","23","5","19",
"expect-no-chan",
NULL
};
static char *test_case_1[] = {
"2",
"Handover to best better cell\n\n"
"The best neighbor cell is selected\n",
"create-bts", "7",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "10","0",
"6","0","20","1","21","2","18","3","20","4","23","5","19",
"expect-chan", "5", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_2[] = {
"2",
"Handover and Assignment must be enabled\n\n"
"This test will start with disabled assignment and handover. A\n"
"better neighbor cell (assignment enabled) will not be selected and \n"
"also no assignment from TCH/H to TCH/F to improve quality. There\n"
"will be no handover nor assignment. After enabling assignment on the\n"
"current cell, the MS will assign to TCH/F. After enabling handover\n"
"in the current cell, but disabling in the neighbor cell, handover\n"
"will not be performed, until it is enabled in the neighbor cell too.\n",
"create-bts", "2",
"afs-rxlev-improve", "0", "5",
"create-ms", "0", "TCH/H", "AMR",
"as-enable", "0", "0",
"ho-enable", "0", "0",
"meas-rep", "0", "0","0", "1","0","30",
"expect-no-chan",
"as-enable", "0", "1",
"meas-rep", "0", "0","0", "1","0","30",
"expect-chan", "0", "1",
"ack-chan",
"expect-ho", "0", "5",
"ho-complete",
"ho-enable", "0", "1",
"ho-enable", "1", "0",
"meas-rep", "0", "0","0", "1","0","30",
"expect-no-chan",
"ho-enable", "1", "1",
"meas-rep", "0", "0","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_3[] = {
"2",
"Penalty timer must not run\n\n"
"The MS will try to handover to a better cell, but this will fail.\n"
"Even though the cell is still better, handover will not be performed\n"
"due to penalty timer after handover failure\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-failed",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
NULL
};
static char *test_case_4[] = {
"2",
"TCH/H keeping with HR codec\n\n"
"The MS is using half rate V1 codec, but the better cell is congested\n"
"at TCH/H slots. As the congestion is removed, the handover takes\n"
"place.\n",
"create-bts", "2",
"set-min-free", "1", "TCH/H", "4",
"create-ms", "0", "TCH/H", "HR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
"set-min-free", "1", "TCH/H", "3",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "5",
"ack-chan",
"expect-ho", "0", "5",
"ho-complete",
NULL
};
static char *test_case_5[] = {
"2",
"TCH/F keeping with FR codec\n\n"
"The MS is using full rate V1 codec, but the better cell is congested\n"
"at TCH/F slots. As the congestion is removed, the handover takes\n"
"place.\n",
"create-bts", "2",
"set-min-free", "1", "TCH/F", "4",
"create-ms", "0", "TCH/F", "FR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
"set-min-free", "1", "TCH/F", "3",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_6[] = {
"2",
"TCH/F keeping with EFR codec\n\n"
"The MS is using full rate V2 codec, but the better cell is congested\n"
"at TCH/F slots. As the congestion is removed, the handover takes\n"
"place.\n",
"create-bts", "2",
"set-min-free", "1", "TCH/F", "4",
"create-ms", "0", "TCH/F", "EFR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
"set-min-free", "1", "TCH/F", "3",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_7[] = {
"2",
"TCH/F to TCH/H changing with AMR codec\n\n"
"The MS is using AMR V3 codec, the better cell is congested at TCH/F\n"
"slots. The handover is performed to non-congested TCH/H slots.\n",
"create-bts", "2",
"set-min-free", "1", "TCH/F", "4",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "5",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_8[] = {
"2",
"No handover to a cell with no slots available\n\n"
"If no slot is available, no handover is performed\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"meas-rep", "0", "0","0", "1","0","30",
"expect-no-chan",
NULL
};
static char *test_case_9[] = {
"2",
"No more parallel handovers, if max_unsync_ho is defined\n\n"
"There are tree mobiles that want to handover, but only two can do\n"
"it at a time, because the maximum number is limited to two.\n",
"create-bts", "2",
"set-max-ho", "1", "2",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "0","0", "1","0","30",
"expect-chan", "1", "1",
"meas-rep", "1", "0","0", "1","0","30",
"expect-chan", "1", "2",
"meas-rep", "2", "0","0", "1","0","30",
"expect-no-chan",
NULL
};
static char *test_case_10[] = {
"2",
"Hysteresis\n\n"
"If neighbor cell is better, handover is only performed if the\n"
Fix some typos Fix typos and common misspellings in code comments and in the manual. Change-Id: I46fc9d424620c77ae9ccf78b58081bd303386d7c
2019-11-13 21:10:41 +00:00
"amount of improvement is greater or equal hyteresis\n",
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "27","0", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "26","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_11[] = {
"2",
"No Hysteresis and minimum RX level\n\n"
"If current cell's RX level is below mimium level, handover must be\n"
"performed, no matter of the hysteresis. First do not perform\n"
"handover to better neighbor cell, because the hysteresis is not\n"
"met. Second do not perform handover because better neighbor cell is\n"
"below minimum RX level. Third perform handover because current cell\n"
"is below minimum RX level, even if the better neighbor cell (minimum\n"
"RX level reached) does not meet the hysteresis.\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "10","0", "1","0","11",
"expect-no-chan",
"meas-rep", "0", "8","0", "1","0","9",
"expect-no-chan",
"meas-rep", "0", "9","0", "1","0","10",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_12[] = {
"2",
"No handover to congested cell\n\n"
"The better neighbor cell is congested, so no handover is performed.\n"
"After the congestion is over, handover will be performed.\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
"set-min-free", "1", "TCH/F", "3",
"set-min-free", "1", "TCH/H", "3",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_13[] = {
"2",
"Handover to balance congestion\n\n"
"The current and the better cell are congested, so no handover is\n"
"performed. This is because handover would congest the neighbor cell\n"
"more. After congestion raises in the current cell, the handover is\n"
"performed to balance congestion\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "0", "TCH/H", "4",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "20","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_14[] = {
"2",
"Handover to congested cell, if RX level is below minimum\n\n"
"The better neighbor cell is congested, so no handover is performed.\n"
"If the RX level of the current cell drops below minimum acceptable\n"
"level, the handover is performed.\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
"meas-rep", "0", "10","0", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "9","0", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_15[] = {
"2",
"Handover to cell with worse RXLEV, if RXQUAL is below minimum\n\n"
"The neighbor cell has worse RXLEV, so no handover is performed.\n"
"If the RXQUAL of the current cell drops below minimum acceptable\n"
"level, the handover is performed. It is also required that 10\n"
"reports are received, before RXQUAL is checked.\n",
/* (See also test 28, which tests for RXQUAL triggering HO to congested cell.) */
/* TODO: bad RXQUAL may want to prefer assignment within the same cell to avoid interference.
Fix some typos Fix typos and common misspellings in code comments and in the manual. Change-Id: I46fc9d424620c77ae9ccf78b58081bd303386d7c
2019-11-13 21:10:41 +00:00
* See Performance Enhancements in a Frequency Hopping GSM Network (Nielsen Wigard 2002), Chapter
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
* 2.1.1, "Interference" in the list of triggers on p.157. */
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-no-chan",
"meas-rep", "0", "40","6", "1","0","30",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_16[] = {
"2",
"Handover due to maximum TA exceeded\n\n"
"The MS in the current (best) cell has reached maximum allowed timing\n"
"advance. No handover is performed until the timing advance exceeds\n"
"it. The originating cell is still the best, but no handover is\n"
"performed back to that cell, because the penalty timer (due to\n"
"maximum allowed timing advance) is running.\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"set-max-ta", "0", "5", /* of cell */
"set-ta", "0", "5", /* of ms */
"meas-rep", "0", "30","0", "1","0","20",
"expect-no-chan",
"set-ta", "0", "6", /* of ms */
"meas-rep", "0", "30","0", "1","0","20",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
"meas-rep", "0", "20","0", "1","0","30",
"expect-no-chan",
NULL
};
static char *test_case_17[] = {
"2",
"Congestion check: No congestion\n\n"
"Three cells have different number of used slots, but there is no\n"
"congestion in any of these cells. No handover is performed.\n",
"create-bts", "3",
"set-min-free", "0", "TCH/F", "2",
"set-min-free", "0", "TCH/H", "2",
"set-min-free", "1", "TCH/F", "2",
"set-min-free", "1", "TCH/H", "2",
"set-min-free", "2", "TCH/F", "2",
"set-min-free", "2", "TCH/H", "2",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "1", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "2", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "3", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "4", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "5", "30","0", "2","0","20","1","20",
"expect-no-chan",
"congestion-check",
"expect-no-chan",
NULL
};
static char *test_case_18[] = {
"2",
"Congestion check: One out of three cells is congested\n\n"
"Three cells have different number of used slots, but there is\n"
"congestion at TCH/F in the first cell. Handover is performed with\n"
"the best candidate.\n",
"create-bts", "3",
"set-min-free", "0", "TCH/F", "2",
"set-min-free", "0", "TCH/H", "2",
"set-min-free", "1", "TCH/F", "2",
"set-min-free", "1", "TCH/H", "2",
"set-min-free", "2", "TCH/F", "2",
"set-min-free", "2", "TCH/H", "2",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "1", "TCH/F", "AMR",
"create-ms", "1", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "1", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "2", "30","0", "2","0","21","1","20",
"expect-no-chan",
"meas-rep", "3", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "4", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "5", "30","0", "2","0","20","1","20",
"expect-no-chan",
"meas-rep", "6", "30","0", "2","0","20","1","20",
"expect-no-chan",
"congestion-check",
"expect-chan", "1", "2",
"ack-chan",
"expect-ho", "0", "3", /* best candidate is MS 2 at BTS 1, TS 3 */
"ho-complete",
NULL
};
static char *test_case_19[] = {
"2",
"Congestion check: Balancing over congested cells\n\n"
handover_test.c: typo s/more/less in test 19 description Change-Id: Ia46fa9bb4f340d35515168f82dd23a1270f6fcd5
2020-08-10 19:13:55 +00:00
"Two cells are congested, but the second cell is less congested.\n"
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
"Handover is performed to solve the congestion.\n",
"create-bts", "2",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "1", "TCH/F", "4",
"create-ms", "0", "TCH/F", "FR",
"create-ms", "0", "TCH/F", "FR",
"create-ms", "0", "TCH/F", "FR",
"create-ms", "1", "TCH/F", "FR",
"meas-rep", "0", "30","0", "1","0","20",
"expect-no-chan",
"meas-rep", "1", "30","0", "1","0","21",
"expect-no-chan",
"meas-rep", "2", "30","0", "1","0","20",
"expect-no-chan",
"meas-rep", "3", "30","0", "1","0","20",
"expect-no-chan",
"congestion-check",
"expect-chan", "1", "2",
"ack-chan",
"expect-ho", "0", "2", /* best candidate is MS 1 at BTS 0, TS 2 */
"ho-complete",
NULL
};
static char *test_case_20[] = {
"2",
"Congestion check: Solving congestion by handover TCH/F -> TCH/H\n\n"
"Two BTS, one MS in the first congested BTS must handover to\n"
"non-congested TCH/H of second BTS, in order to solve congestion\n",
"create-bts", "2",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "0", "TCH/H", "4",
"set-min-free", "1", "TCH/F", "4",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "30","0", "1","0","30",
"expect-no-chan",
"congestion-check",
"expect-chan", "1", "5",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_21[] = {
"2",
"Congestion check: Balancing congestion by handover TCH/F -> TCH/H\n\n"
"Two BTS, one MS in the first congested BTS must handover to\n"
"less-congested TCH/H of second BTS, in order to balance congestion\n",
"create-bts", "2",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "0", "TCH/H", "4",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "1","0","30",
"expect-no-chan",
"congestion-check",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_22[] = {
"2",
"Congestion check: Upgrading worst candidate from TCH/H -> TCH/F\n\n"
"There is only one BTS. The TCH/H slots are congested. Since\n"
"assignment is performed to less-congested TCH/F, the candidate with\n"
"the worst RX level is chosen.\n",
"create-bts", "1",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "0", "TCH/H", "4",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "0",
"meas-rep", "1", "34","0", "0",
"meas-rep", "2", "20","0", "0",
"expect-no-chan",
"congestion-check",
"expect-chan", "0", "1",
"ack-chan",
"expect-ho", "0", "6",
"ho-complete",
NULL
};
static char *test_case_23[] = {
"2",
"Story: 'A neighbor is your friend'\n",
"create-bts", "3",
"print",
"Andreas is driving along the coast, on a sunny june afternoon.\n"
"Suddenly he is getting a call from his friend and neighbor Axel.\n"
"\n"
"What happens: Two MS are created, #0 for Axel, #1 for Andreas.",
/* Axel */
"create-ms", "2", "TCH/F", "AMR",
/* andreas */
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "1", "40","0", "1","0","30",
"expect-no-chan",
"print",
"Axel asks Andreas if he would like to join them for a barbecue.\n"
"Axel's house is right in the neighborhood and the weather is fine.\n"
"Andreas agrees, so he drives to a close store to buy some barbecue\n"
"skewers.\n"
"\n"
"What happens: While driving, a different cell (mounted atop the\n"
"store) becomes better.",
/* drive to bts 1 */
"meas-rep", "1", "20","0", "1","0","35",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
"print",
"While Andreas is walking into the store, Axel asks, if he could also\n"
"bring some beer. Andreas has problems understanding him: \"I have a\n"
"bad reception here. The cell tower is right atop the store, but poor\n"
"coverage inside. Can you repeat please?\"\n"
"\n"
"What happens: Inside the store the close cell is so bad, that\n"
"handover back to the previous cell is required.",
/* bts 1 becomes bad, so bts 0 helps out */
"meas-rep", "1", "5","0", "1","0","20",
"expect-chan", "0", "1",
"ack-chan",
"expect-ho", "1", "1",
"ho-complete",
"print",
"After Andreas bought skewers and beer, he leaves the store.\n"
"\n"
"What happens: Outside the store the close cell is better again, so\n"
"handover back to the that cell is performed.",
/* bts 1 becomes better again */
"meas-rep", "1", "20","0", "1","0","35",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
"print",
/* bts 2 becomes better */
"Andreas drives down to the lake where Axel's house is.\n"
"\n"
"What happens: There is a small cell at Axel's house, which becomes\n"
"better, because the current cell has no good comverage at the lake.",
"meas-rep", "1", "14","0", "2","0","2","1","63",
"expect-chan", "2", "2",
"ack-chan",
"expect-ho", "1", "1",
"ho-complete",
"print",
"Andreas wonders why he still has good radio coverage: \"Last time it\n"
"was so bad\". Axel says: \"I installed a pico cell in my house,\n"
"now we can use our mobile phones down here at the lake.\"",
NULL
};
static char *test_case_24[] = {
"2",
"No (or not enough) measurements for handover\n\n"
"Do not solve congestion in cell, because there is no measurement.\n"
Fix some typos Fix typos and common misspellings in code comments and in the manual. Change-Id: I46fc9d424620c77ae9ccf78b58081bd303386d7c
2019-11-13 21:10:41 +00:00
"As soon as enough measurements available (1 in our case), perform\n"
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
"handover. Afterwards the old cell becomes congested and the new\n"
"cell is not. Do not perform handover until new measurements are\n"
"received.\n",
/* two cells, first in congested, but no handover */
"create-bts", "2",
"set-min-free", "0", "TCH/F", "4",
"set-min-free", "0", "TCH/H", "4",
"create-ms", "0", "TCH/F", "AMR",
"congestion-check",
"expect-no-chan",
/* send measurement and trigger congestion check */
"meas-rep", "0", "20","0", "1","0","20",
"expect-no-chan",
"congestion-check",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
/* congest the first cell and remove congestion from second cell */
"set-min-free", "0", "TCH/F", "0",
"set-min-free", "0", "TCH/H", "0",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
/* no handover until measurements applied */
"congestion-check",
"expect-no-chan",
"meas-rep", "0", "20","0", "1","0","20",
"expect-no-chan",
"congestion-check",
"expect-chan", "0", "1",
"ack-chan",
"expect-ho", "1", "1",
"ho-complete",
NULL
};
static char *test_case_25[] = {
"1",
"Stay in better cell\n\n"
"There are many neighbor cells, but only the current cell is the best\n"
"cell, so no handover is performed\n",
"create-bts", "7",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "30","0",
"6","0","20","1","21","2","18","3","20","4","23","5","19",
"expect-no-chan",
NULL
};
static char *test_case_26[] = {
"1",
"Handover to best better cell\n\n"
"The best neighbor cell is selected\n",
"create-bts", "7",
"create-ms", "0", "TCH/F", "AMR",
"meas-rep", "0", "10","0",
"6","0","20","1","21","2","18","3","20","4","23","5","19",
"expect-chan", "5", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
static char *test_case_27[] = {
"2",
"Congestion check: Upgrading worst candidate from TCH/H -> TCH/F\n\n"
"There is only one BTS. The TCH/H slots are congested. Since\n"
"assignment is performed to less-congested TCH/F, the candidate with\n"
"the worst RX level is chosen. (So far like test 22.)\n"
"After that, trigger more congestion checks to ensure stability.\n",
"create-bts", "1",
"set-min-free", "0", "TCH/F", "2",
"set-min-free", "0", "TCH/H", "4",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "0",
"meas-rep", "1", "34","0", "0",
"meas-rep", "2", "20","0", "0",
"expect-no-chan",
"congestion-check",
"expect-chan", "0", "1",
"ack-chan",
"expect-ho", "0", "6",
"ho-complete",
"congestion-check",
"expect-chan", "0", "2",
"ack-chan",
"expect-ho", "0", "5",
"ho-complete",
"congestion-check",
"expect-no-chan",
"congestion-check",
"expect-no-chan",
NULL
};
static char *test_case_28[] = {
"2",
"Handover to congested cell, if RX quality is below minimum\n\n"
"The better neighbor cell is congested, so no handover is performed.\n"
"If the RX quality of the current cell drops below minimum acceptable\n"
"level, the handover is performed. It is also required that 10\n"
"resports are received, before RX quality is checked.\n",
"create-bts", "2",
"create-ms", "0", "TCH/F", "AMR",
"set-min-free", "1", "TCH/F", "4",
"set-min-free", "1", "TCH/H", "4",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-no-chan",
"meas-rep", "0", "30","6", "1","0","40",
"expect-chan", "1", "1",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
handover_test.c: add test case 29: TCH/F -> TCH/H Change-Id: I5439beec4d295a8acaa1c49a0fa575dc2f7b973d
2020-08-02 00:50:17 +00:00
static char *test_case_29[] = {
"2",
"Congestion check: Balancing congestion by handover TCH/F -> TCH/H\n\n"
"One BTS, and TCH/F are considered congested, TCH/H are not.\n"
,
"create-bts", "1",
"set-min-free", "0", "TCH/F", "3",
"set-min-free", "0", "TCH/H", "0",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/F", "AMR",
"create-ms", "0", "TCH/H", "AMR",
"meas-rep", "0", "30","0", "1","0","30",
"expect-no-chan",
"congestion-check",
"expect-chan", "0", "5",
"ack-chan",
"expect-ho", "0", "1",
"ho-complete",
NULL
};
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
static char **test_cases[] = {
test_case_0,
test_case_1,
test_case_2,
test_case_3,
test_case_4,
test_case_5,
test_case_6,
test_case_7,
test_case_8,
test_case_9,
test_case_10,
test_case_11,
test_case_12,
test_case_13,
test_case_14,
test_case_15,
test_case_16,
test_case_17,
test_case_18,
test_case_19,
test_case_20,
test_case_21,
test_case_22,
test_case_23,
test_case_24,
test_case_25,
test_case_26,
test_case_27,
test_case_28,
handover_test.c: add test case 29: TCH/F -> TCH/H Change-Id: I5439beec4d295a8acaa1c49a0fa575dc2f7b973d
2020-08-02 00:50:17 +00:00
test_case_29,
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
};
static const struct log_info_cat log_categories[] = {
[DHO] = {
.name = "DHO",
.description = "Hand-Over Process",
.color = "\033[1;38m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DHODEC] = {
.name = "DHODEC",
.description = "Hand-Over Decision",
.color = "\033[1;38m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DMEAS] = {
.name = "DMEAS",
.description = "Radio Measurement Processing",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DREF] = {
.name = "DREF",
.description = "Reference Counting",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DRSL] = {
.name = "DRSL",
Cosmetic: Fix typo: Siganlling->Signalling Change-Id: I92b39eebfba396ee7690e99de09ee20593b7139d
2018-04-16 11:40:07 +00:00
.description = "A-bis Radio Signalling Link (RSL)",
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
.color = "\033[1;35m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
[DRR] = {
.name = "DRR",
.description = "RR",
.color = "\033[1;35m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DRLL] = {
.name = "DRLL",
.description = "RLL",
.color = "\033[1;35m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
[DMSC] = {
.name = "DMSC",
.description = "Mobile Switching Center",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
create separate logging categories for lchan,ts,as FSMs Change-Id: Ie889b8860a4a63c7c22ef65025f690d64cd7330c
2018-07-11 17:53:39 +00:00
[DCHAN] = {
.name = "DCHAN",
.description = "lchan FSM",
.color = "\033[1;32m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DTS] = {
.name = "DTS",
.description = "timeslot FSM",
.color = "\033[1;31m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
[DAS] = {
.name = "DAS",
.description = "assignment FSM",
.color = "\033[1;33m",
.enabled = 1, .loglevel = LOGL_DEBUG,
},
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
};
const struct log_info log_info = {
.cat = log_categories,
.num_cat = ARRAY_SIZE(log_categories),
};
int main(int argc, char **argv)
{
char **test_case;
struct gsm_bts *bts[256];
int bts_num = 0;
struct gsm_lchan *lchan[256];
int lchan_num = 0;
int i;
dissolve libbsc: move all to src/osmo-bsc, link .o files 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
2018-05-26 23:26:31 +00:00
int algorithm;
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
int test_case_i;
int last_test_i;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
ctx cleanup: use non-NULL talloc ctx for osmo_init_logging2() Fix various sanitizer complaints about memory leaks using a sanitizer build with gcc (Debian 7.3.0-12) 7.3.0. Also fix deprecation warnings on osmo_init_logging(). Depends: I216837780e9405fdaec8059c63d10699c695b360 (libosmocore) Change-Id: I970c6f8a0e36a8b63e42349dbc92baff649e5cef
2018-03-05 04:31:14 +00:00
ctx = talloc_named_const(NULL, 0, "handover_test");
msgb_talloc_ctx_init(ctx, 0);
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
test_case_i = argc > 1? atoi(argv[1]) : -1;
last_test_i = ARRAY_SIZE(test_cases) - 1;
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
if (test_case_i < 0 || test_case_i > last_test_i) {
for (i = 0; i <= last_test_i; i++) {
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
printf("Test #%d (algorithm %s):\n%s\n", i,
test_cases[i][0], test_cases[i][1]);
}
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
printf("\nPlease specify test case number 0..%d\n", last_test_i);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
return EXIT_FAILURE;
}
ctx cleanup: use non-NULL talloc ctx for osmo_init_logging2() Fix various sanitizer complaints about memory leaks using a sanitizer build with gcc (Debian 7.3.0-12) 7.3.0. Also fix deprecation warnings on osmo_init_logging(). Depends: I216837780e9405fdaec8059c63d10699c695b360 (libosmocore) Change-Id: I970c6f8a0e36a8b63e42349dbc92baff649e5cef
2018-03-05 04:31:14 +00:00
osmo_init_logging2(ctx, &log_info);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
log_set_print_category(osmo_stderr_target, 1);
log_set_print_category_hex(osmo_stderr_target, 0);
log_set_print_filename2(osmo_stderr_target, LOG_FILENAME_BASENAME);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
osmo_fsm_log_addr(false);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
dissolve libbsc: move all to src/osmo-bsc, link .o files 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
2018-05-26 23:26:31 +00:00
bsc_network_alloc();
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
if (!bsc_gsmnet)
exit(1);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
ts_fsm_init();
lchan_fsm_init();
bsc_subscr_conn_fsm_init();
handover_fsm_init();
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
ho_set_algorithm(bsc_gsmnet->ho, 2);
ho_set_ho_active(bsc_gsmnet->ho, true);
ho_set_hodec2_as_active(bsc_gsmnet->ho, true);
ho_set_hodec2_min_rxlev(bsc_gsmnet->ho, -100);
ho_set_hodec2_rxlev_avg_win(bsc_gsmnet->ho, 1);
ho_set_hodec2_rxlev_neigh_avg_win(bsc_gsmnet->ho, 1);
ho_set_hodec2_rxqual_avg_win(bsc_gsmnet->ho, 10);
ho_set_hodec2_pwr_hysteresis(bsc_gsmnet->ho, 3);
ho_set_hodec2_pwr_interval(bsc_gsmnet->ho, 1);
ho_set_hodec2_afs_bias_rxlev(bsc_gsmnet->ho, 0);
ho_set_hodec2_min_rxqual(bsc_gsmnet->ho, 5);
ho_set_hodec2_afs_bias_rxqual(bsc_gsmnet->ho, 0);
ho_set_hodec2_max_distance(bsc_gsmnet->ho, 9999);
ho_set_hodec2_ho_max(bsc_gsmnet->ho, 9999);
ho_set_hodec2_penalty_max_dist(bsc_gsmnet->ho, 300);
ho_set_hodec2_penalty_failed_ho(bsc_gsmnet->ho, 60);
ho_set_hodec2_penalty_failed_as(bsc_gsmnet->ho, 60);
handover_test: use 'unknown' BTS type instead of 'sysmobts' This unit test does not really need a BTS of such specific type. Change-Id: Id676042518d06e94a9fb20112334280e2b91074b
2020-05-30 18:56:20 +00:00
/* We don't really need any specific model here */
bts_model_unknown_init();
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
test_case = test_cases[test_case_i];
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
fprintf(stderr, "--------------------\n");
fprintf(stderr, "Performing the following test %d (algorithm %s):\n%s",
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
test_case_i, test_case[0], test_case[1]);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
algorithm = atoi(test_case[0]);
test_case += 2;
fprintf(stderr, "--------------------\n");
/* Disable the congestion check timer, we will trigger manually. */
bsc_gsmnet->hodec2.congestion_check_interval_s = 0;
handover_decision_1_init();
hodec2_init(bsc_gsmnet);
while (*test_case) {
if (!strcmp(*test_case, "create-bts")) {
static int arfcn = 870;
int n = atoi(test_case[1]);
fprintf(stderr, "- Creating %d BTS (one TRX each, "
"TS(1-4) are TCH/F, TS(5-6) are TCH/H)\n", n);
for (i = 0; i < n; i++)
bts[bts_num + i] = create_bts(arfcn++);
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
for (i = 0; i < n; i++) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (gsm_generate_si(bts[bts_num + i], SYSINFO_TYPE_2) <= 0)
HO: fix minor issues found by coverity handover_decision_2: - lchan->conn should never be NULL, but if it is, don't crash in LOGPHO* macros. - make certain to avoid modulo-zero. It's cosmetic since that config item is not allowed to be zero. handover_test: - check return value of gsm_generate_si(). - safer evaluation of main()'s argv. Fixes: CID#182929, CID#182928, CID#182927, CID#182926 Change-Id: I68e8ed3a1e8659edb47ac3e8a10508f17a3f5670
2018-02-21 13:33:15 +00:00
fprintf(stderr, "Error generating SI2\n");
}
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
bts_num += n;
test_case += 2;
} else
if (!strcmp(*test_case, "as-enable")) {
fprintf(stderr, "- Set assignment enable state at "
"BTS %s to %s\n", test_case[1], test_case[2]);
ho_set_hodec2_as_active(bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "ho-enable")) {
fprintf(stderr, "- Set handover enable state at "
"BTS %s to %s\n", test_case[1], test_case[2]);
ho_set_ho_active(bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "afs-rxlev-improve")) {
fprintf(stderr, "- Set afs RX level improvement at "
"BTS %s to %s\n", test_case[1], test_case[2]);
ho_set_hodec2_afs_bias_rxlev(bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "afs-rxqual-improve")) {
fprintf(stderr, "- Set afs RX quality improvement at "
"BTS %s to %s\n", test_case[1], test_case[2]);
ho_set_hodec2_afs_bias_rxqual(bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "set-min-free")) {
fprintf(stderr, "- Setting minimum required free %s "
"slots at BTS %s to %s\n", test_case[2],
test_case[1], test_case[3]);
if (!strcmp(test_case[2], "TCH/F"))
ho_set_hodec2_tchf_min_slots(bts[atoi(test_case[1])]->ho, atoi(test_case[3]));
else
ho_set_hodec2_tchh_min_slots(bts[atoi(test_case[1])]->ho, atoi(test_case[3]));
test_case += 4;
} else
if (!strcmp(*test_case, "set-max-ho")) {
fprintf(stderr, "- Setting maximum parallel handovers "
"at BTS %s to %s\n", test_case[1],
test_case[2]);
ho_set_hodec2_ho_max( bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "set-max-ta")) {
fprintf(stderr, "- Setting maximum timing advance "
"at BTS %s to %s\n", test_case[1],
test_case[2]);
ho_set_hodec2_max_distance(bts[atoi(test_case[1])]->ho, atoi(test_case[2]));
test_case += 3;
} else
if (!strcmp(*test_case, "create-ms")) {
fprintf(stderr, "- Creating mobile #%d at BTS %s on "
"%s with %s codec\n", lchan_num, test_case[1],
test_case[2], test_case[3]);
lchan[lchan_num] = create_lchan(bts[atoi(test_case[1])],
!strcmp(test_case[2], "TCH/F"), test_case[3]);
if (!lchan[lchan_num]) {
printf("Failed to create lchan!\n");
return EXIT_FAILURE;
}
fprintf(stderr, " * New MS is at BTS %d TS %d\n",
lchan[lchan_num]->ts->trx->bts->nr,
lchan[lchan_num]->ts->nr);
lchan_num++;
test_case += 4;
} else
if (!strcmp(*test_case, "set-ta")) {
fprintf(stderr, "- Setting maximum timing advance "
"at MS %s to %s\n", test_case[1],
test_case[2]);
meas_ta_ms = atoi(test_case[2]);
test_case += 3;
} else
if (!strcmp(*test_case, "meas-rep")) {
/* meas-rep <lchan-nr> <rxlev> <rxqual> <nr-of-neighbors> [<cell-idx> <rxlev> [...]] */
int n = atoi(test_case[4]);
struct gsm_lchan *lc = lchan[atoi(test_case[1])];
fprintf(stderr, "- Sending measurement report from "
"mobile #%s (rxlev=%s, rxqual=%s)\n",
test_case[1], test_case[2], test_case[3]);
meas_dl_rxlev = atoi(test_case[2]);
meas_dl_rxqual = atoi(test_case[3]);
meas_num_nc = n;
test_case += 5;
for (i = 0; i < n; i++) {
int nr = atoi(test_case[0]);
/* since our bts is not in the list of neighbor
* cells, we need to shift */
if (nr >= lc->ts->trx->bts->nr)
nr++;
fprintf(stderr, " * Neighbor cell #%s, actual "
"BTS %d (rxlev=%s)\n", test_case[0], nr,
test_case[1]);
meas_bcch_f_nc[i] = atoi(test_case[0]);
/* bts number, not counting our own */
meas_rxlev_nc[i] = atoi(test_case[1]);
meas_bsic_nc[i] = 0x3f;
test_case += 2;
}
got_chan_req = 0;
gen_meas_rep(lc);
} else
if (!strcmp(*test_case, "congestion-check")) {
fprintf(stderr, "- Triggering congestion check\n");
got_chan_req = 0;
if (algorithm == 2)
hodec2_congestion_check(bsc_gsmnet);
test_case += 1;
} else
if (!strcmp(*test_case, "expect-chan")) {
fprintf(stderr, "- Expecting channel request at BTS %s "
"TS %s\n", test_case[1], test_case[2]);
if (!got_chan_req) {
printf("Test failed, because no channel was "
"requested\n");
return EXIT_FAILURE;
}
fprintf(stderr, " * Got channel request at BTS %d "
"TS %d\n", chan_req_lchan->ts->trx->bts->nr,
chan_req_lchan->ts->nr);
if (chan_req_lchan->ts->trx->bts->nr
!= atoi(test_case[1])) {
printf("Test failed, because channel was not "
"requested on expected BTS\n");
return EXIT_FAILURE;
}
if (chan_req_lchan->ts->nr != atoi(test_case[2])) {
printf("Test failed, because channel was not "
"requested on expected TS\n");
return EXIT_FAILURE;
}
test_case += 3;
} else
if (!strcmp(*test_case, "expect-no-chan")) {
fprintf(stderr, "- Expecting no channel request\n");
if (got_chan_req) {
fprintf(stderr, " * Got channel request at "
"BTS %d TS %d\n",
chan_req_lchan->ts->trx->bts->nr,
chan_req_lchan->ts->nr);
printf("Test failed, because channel was "
"requested\n");
return EXIT_FAILURE;
}
fprintf(stderr, " * Got no channel request\n");
test_case += 1;
} else
if (!strcmp(*test_case, "expect-ho")) {
fprintf(stderr, "- Expecting handover/assignment "
"request at BTS %s TS %s\n", test_case[1],
test_case[2]);
if (!got_ho_req) {
printf("Test failed, because no handover was "
"requested\n");
return EXIT_FAILURE;
}
fprintf(stderr, " * Got handover/assignment request at "
"BTS %d TS %d\n",
ho_req_lchan->ts->trx->bts->nr,
ho_req_lchan->ts->nr);
if (ho_req_lchan->ts->trx->bts->nr
!= atoi(test_case[1])) {
printf("Test failed, because "
"handover/assignment was not commanded "
"at the expected BTS\n");
return EXIT_FAILURE;
}
if (ho_req_lchan->ts->nr != atoi(test_case[2])) {
printf("Test failed, because "
"handover/assignment was not commanded "
"at the expected TS\n");
return EXIT_FAILURE;
}
test_case += 3;
} else
if (!strcmp(*test_case, "ack-chan")) {
fprintf(stderr, "- Acknowledging channel request\n");
if (!got_chan_req) {
printf("Cannot ack channel, because no "
"request\n");
return EXIT_FAILURE;
}
test_case += 1;
got_ho_req = 0;
send_chan_act_ack(chan_req_lchan, 1);
} else
if (!strcmp(*test_case, "ho-complete")) {
fprintf(stderr, "- Acknowledging handover/assignment "
"request\n");
if (!got_chan_req) {
printf("Cannot ack handover/assignment, "
"because no chan request\n");
return EXIT_FAILURE;
}
if (!got_ho_req) {
printf("Cannot ack handover/assignment, "
"because no ho request\n");
return EXIT_FAILURE;
}
test_case += 1;
got_chan_req = 0;
got_ho_req = 0;
/* switch lchan */
for (i = 0; i < lchan_num; i++) {
if (lchan[i] == ho_req_lchan) {
fprintf(stderr, " * MS %d changes from "
"BTS=%d TS=%d to BTS=%d "
"TS=%d\n", i,
lchan[i]->ts->trx->bts->nr,
lchan[i]->ts->nr,
chan_req_lchan->ts->trx->bts->nr,
chan_req_lchan->ts->nr);
lchan[i] = chan_req_lchan;
}
}
send_ho_complete(chan_req_lchan, true);
} else
if (!strcmp(*test_case, "ho-failed")) {
fprintf(stderr, "- Making handover fail\n");
if (!got_chan_req) {
printf("Cannot fail handover, because no chan "
"request\n");
return EXIT_FAILURE;
}
test_case += 1;
got_chan_req = 0;
got_ho_req = 0;
send_ho_complete(ho_req_lchan, false);
} else
if (!strcmp(*test_case, "print")) {
fprintf(stderr, "\n%s\n\n", test_case[1]);
test_case += 2;
} else {
printf("Unknown test command '%s', please fix!\n",
*test_case);
return EXIT_FAILURE;
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
/* Help the lchan out of releasing states */
struct gsm_bts *bts;
llist_for_each_entry(bts, &bsc_gsmnet->bts_list, list) {
struct gsm_bts_trx *trx;
llist_for_each_entry(trx, &bts->trx_list, list) {
int ts_nr;
for (ts_nr = 0; ts_nr < TRX_NR_TS; ts_nr++) {
struct gsm_lchan *lchan;
ts_for_each_lchan(lchan, &trx->ts[ts_nr]) {
if (lchan->fi && lchan->fi->state == LCHAN_ST_WAIT_BEFORE_RF_RELEASE) {
osmo_fsm_inst_state_chg(lchan->fi, LCHAN_ST_WAIT_RF_RELEASE_ACK, 0, 0);
osmo_fsm_inst_dispatch(lchan->fi, LCHAN_EV_RSL_RF_CHAN_REL_ACK, 0);
}
}
}
}
}
}
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
}
for (i = 0; i < lchan_num; i++) {
struct gsm_subscriber_connection *conn = lchan[i]->conn;
lchan[i]->conn = NULL;
conn->lchan = NULL;
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
osmo_fsm_inst_term(conn->fi, OSMO_FSM_TERM_REGULAR, NULL);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
}
fprintf(stderr, "--------------------\n");
printf("Test OK\n");
fprintf(stderr, "--------------------\n");
ctx cleanup: use non-NULL talloc ctx for osmo_init_logging2() Fix various sanitizer complaints about memory leaks using a sanitizer build with gcc (Debian 7.3.0-12) 7.3.0. Also fix deprecation warnings on osmo_init_logging(). Depends: I216837780e9405fdaec8059c63d10699c695b360 (libosmocore) Change-Id: I970c6f8a0e36a8b63e42349dbc92baff649e5cef
2018-03-05 04:31:14 +00:00
talloc_free(ctx);
HO: Implement load based handover, as handover_decision_2.c Change-Id: Ie597eae82722baf32546331e443dd9d94f1f25e6
2017-12-07 02:54:01 +00:00
return EXIT_SUCCESS;
}
void rtp_socket_free() {}
void rtp_send_frame() {}
void rtp_socket_upstream() {}
void rtp_socket_create() {}
void rtp_socket_connect() {}
void rtp_socket_proxy() {}
void trau_mux_unmap() {}
void trau_mux_map_lchan() {}
void trau_recv_lchan() {}
void trau_send_frame() {}
introduce an osmo_fsm for gsm_subscriber_connection In the current implementation of osmo-bsc, the subscriber connection is not handled (very) statefully. However, there is some state keeping in the code that handles the mgcp connection, but there are still to much loose ends which allow odd situations to happen, which then lead severe error situations (see also closes tags at the end) This commit adds a number of improvements to fix those problems. - Use an osmo-fsm to control the gsm_subscriber_connection state and make sure that certain operations can only take place at certain states (e.g let connection oriented SCCP traffic only pass when an SCCP connection actually exists. Remove the old osmo_bsc_mgcp.c code. Use the recently developed MGCP client FSM to handle the MGCP connections. Also make sure that stuff that already works does not break. This in particular refers to the internal handover capability and the respective unit-tests. See also OS#2823, OS#2768 and OS#2898 - Fix logic to permit assignment to a signalling channel. (OS#2762) - Introduce T993210 to release lchan + subscr_conn if MSC fails to respond The GSM specs don't have an explicit timer for this, so let's introdcue a custom timer (hence starting with 99). This timeout catches the following situation: * we send a SCCP CR with COMPL_L3_INFO from the MS to the MSC, * the MSC doesn't respond (e.g. SCCP routing failure, program down, ...) The MS is supposed to timeout with T3210, 3220 or 3230. But the BSC shouldn't trust the MS but have some timer on its own. SCCP would have a timer T(conn est), but that one is specified to be 1-2min and hence rather long. See also: OS#2775 - Terminate bsc_subscr_conn_fsm on SCCP N-DISC.ind from MSC If the MSC is disconnecting the SCCP channel, we must terminate the FSM which in turn will release all lchan's and other state. This makes TC_chan_rel_hard_rlsd pass, see also OS#2731 As a side-effect, this fixes TC_chan_act_ack_est_ind_refused(), where the MSC is answering with CREF to our CR/COMPL_L3. - Release subscriber connection on RLL RELEASE IND of SAPI0 on main DCCH The subscriber connection isn't really useful for anything after the SAPI0 main signalling link has been released. We could try to re-establish, but our best option is probably simply releasing the subscriber_conn and anything related to it. This will make TC_chan_rel_rll_rel_ind pass, see also OS#2730 This commit has been tested using the BSC_Tests TTCN3 testsuit and the following tests were passed: TC_chan_act_noreply TC_chan_act_ack_noest TC_chan_act_ack_est_ind_noreply TC_chan_act_ack_est_ind_refused TC_chan_act_nack TC_chan_exhaustion TC_ctrl TC_chan_rel_conn_fail TC_chan_rel_hard_clear TC_chan_rel_hard_rlsd TC_chan_rel_a_reset TC_rll_est_ind_inact_lchan TC_rll_est_ind_inval_sapi1 TC_rll_est_ind_inval_sapi3 TC_rll_est_ind_inval_sacch TC_assignment_cic_only TC_assignment_csd TC_assignment_ctm TC_assignment_fr_a5_0 TC_assignment_fr_a5_1_codec_missing TC_assignment_fr_a5_1 TC_assignment_fr_a5_3 TC_assignment_fr_a5_4 TC_paging_imsi_nochan TC_paging_tmsi_nochan TC_paging_tmsi_any TC_paging_tmsi_sdcch TC_paging_tmsi_tch_f TC_paging_tmsi_tch_hf TC_paging_imsi_nochan_cgi TC_paging_imsi_nochan_lac_ci TC_paging_imsi_nochan_ci TC_paging_imsi_nochan_lai TC_paging_imsi_nochan_lac TC_paging_imsi_nochan_all TC_paging_imsi_nochan_plmn_lac_rnc TC_paging_imsi_nochan_rnc TC_paging_imsi_nochan_lac_rnc TC_paging_imsi_nochan_lacs TC_paging_imsi_nochan_lacs_empty TC_paging_imsi_a_reset TC_paging_counter TC_rsl_drop_counter TC_classmark TC_unsol_ass_fail TC_unsol_ass_compl TC_unsol_ho_fail TC_err_82_short_msg TC_ho_int Authors: Harald Welte <laforge@gnumonks.org> Philipp Maier <pmaier@sysmocom.de> Neels Hofmeyr <neels@hofmeyr.de> Closes: OS#2730 Closes: OS#2731 Closes: OS#2762 Closes: OS#2768 Closes: OS#2775 Closes: OS#2823 Closes: OS#2898 Closes: OS#2936 Change-Id: I68286d26e2014048b054f39ef29c35fef420cc97
2018-01-28 02:04:16 +00:00
int osmo_bsc_sigtran_send(struct gsm_subscriber_connection *conn, struct msgb *msg) { return 0; }
int osmo_bsc_sigtran_open_conn(struct gsm_subscriber_connection *conn, struct msgb *msg) { return 0; }
bsc_subscr_conn_fsm: use proper cause values in SAPI N REJECT As 3GPP TS 48.058 clearly states, an RLL RELease INDication is only sent if the datalink is released by the MS. Thre's no cause value giving further diagnostics, but if the MS releases the link, it is clearly not the BSS's fault, and hence "BSS not equipped" is wrong. Change-Id: I38222e60071841abcd06046a472ddb35907164a4 Related: OS#4728
2020-08-26 11:49:08 +00:00
void bsc_sapi_n_reject(struct gsm_subscriber_connection *conn, uint8_t dlci, enum gsm0808_cause cause) {}
remove struct bsc_api struct bsc_api was used to provide an abstract API for both osmo-bsc and osmo-msc, between BSC and MSC. That's no longer needed, so get rid of it, to prevent code turbulences it creates for no reason. Change-Id: I3fd5888c63a0b4f95520a498320aa105a6d60579
2018-05-27 01:05:18 +00:00
void bsc_cipher_mode_compl(struct gsm_subscriber_connection *conn, struct msgb *msg, uint8_t chosen_encr) {}
compl l3: allocate conn in gsm_08_08.c, not gsm_04_08_rr.c Move conn allocation to bsc_compl_l3(), from gsm0408_rcvmsg(). Drop dispatch of GSCON_EV_A_DISC_IND, because a) we did not receive such DISC.IND, and b) the lchan release will discard the conn in the regular fashion. In upcoming LCS patch, bsc_compl_l3() will decide whether to allocate a new conn or whether a conn from a Perform Location Request already exists for the subscriber. In this patch, it becomes clear that the conn->bsub is always NULL in bsc_compl_l3(), and that the 'log_set_context(LOG_CTX_BSC_SUBSCR, conn->bsub);' never has the intended effect. An upcoming patch will change that. Change-Id: I92af0f0d54c4282d782f2b29d524a64006c3b674
2020-09-11 23:35:28 +00:00
int bsc_compl_l3(struct gsm_lchan *lchan, struct msgb *msg, uint16_t chosen_channel)
remove struct bsc_api struct bsc_api was used to provide an abstract API for both osmo-bsc and osmo-msc, between BSC and MSC. That's no longer needed, so get rid of it, to prevent code turbulences it creates for no reason. Change-Id: I3fd5888c63a0b4f95520a498320aa105a6d60579
2018-05-27 01:05:18 +00:00
{ return 0; }
refactor paging: introduce bsc_paging_start() Allow starting a paging from elsewhere than a BSSMAP Paging Request. For upcoming Location Services (LCS), a BSSLAP TA Request from the SMLC may require triggering a Paging. Change-Id: Iaff91584699d163bd1963927280ff3a8ddd43073
2020-09-18 00:49:32 +00:00
int bsc_paging_start(struct bsc_paging_params *params)
{ return 0; }
remove struct bsc_api struct bsc_api was used to provide an abstract API for both osmo-bsc and osmo-msc, between BSC and MSC. That's no longer needed, so get rid of it, to prevent code turbulences it creates for no reason. Change-Id: I3fd5888c63a0b4f95520a498320aa105a6d60579
2018-05-27 01:05:18 +00:00
void bsc_dtap(struct gsm_subscriber_connection *conn, uint8_t link_id, struct msgb *msg) {}
void bsc_assign_compl(struct gsm_subscriber_connection *conn, uint8_t rr_cause) {}
void bsc_cm_update(struct gsm_subscriber_connection *conn,
const uint8_t *cm2, uint8_t cm2_len,
const uint8_t *cm3, uint8_t cm3_len) {}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
struct gsm0808_handover_required;
int bsc_tx_bssmap_ho_required(struct gsm_lchan *lchan, const struct gsm0808_cell_id_list2 *target_cells)
{ return 0; }
int bsc_tx_bssmap_ho_request_ack(struct gsm_subscriber_connection *conn, struct msgb *rr_ho_command)
{ return 0; }
int bsc_tx_bssmap_ho_detect(struct gsm_subscriber_connection *conn) { return 0; }
enum handover_result bsc_tx_bssmap_ho_complete(struct gsm_subscriber_connection *conn,
struct gsm_lchan *lchan) { return HO_RESULT_OK; }
void bsc_tx_bssmap_ho_failure(struct gsm_subscriber_connection *conn) {}