osmo-bsc/src/osmo-bsc/bsc_subscr_conn_fsm.c

1158 lines
37 KiB
C
Raw Normal View History

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
/* (C) 2017 by Harald Welte <laforge@gnumonks.org>
* 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 Affero 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 <osmocom/core/fsm.h>
#include <osmocom/core/logging.h>
#include <osmocom/gsm/gsm0808.h>
#include <osmocom/sigtran/sccp_sap.h>
#include <osmocom/gsm/gsm0808_utils.h>
#include <osmocom/bsc/debug.h>
#include <osmocom/bsc/a_reset.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_api.h>
#include <osmocom/bsc/gsm_data.h>
#include <osmocom/bsc/handover.h>
#include <osmocom/bsc/chan_alloc.h>
#include <osmocom/bsc/bsc_subscriber.h>
#include <osmocom/bsc/osmo_bsc_sigtran.h>
#include <osmocom/bsc/osmo_bsc_lcls.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>
#include <osmocom/bsc/osmo_bsc.h>
#include <osmocom/bsc/penalty_timers.h>
#include <osmocom/mgcp_client/mgcp_client_fsm.h>
#include <osmocom/core/byteswap.h>
#define S(x) (1 << (x))
#define MGCP_MGW_TIMEOUT 4 /* in seconds */
#define MGCP_MGW_TIMEOUT_TIMER_NR 1
#define MGCP_MGW_HO_TIMEOUT 4 /* in seconds */
#define MGCP_MGW_HO_TIMEOUT_TIMER_NR 2
#define ENDPOINT_ID "rtpbridge/*@mgw"
enum gscon_fsm_states {
ST_INIT,
/* waiting for CC from MSC */
ST_WAIT_CC,
/* active connection */
ST_ACTIVE,
/* during assignment; waiting for ASS_CMPL */
ST_WAIT_ASS_CMPL,
/* BSSMAP CLEAR has been received */
ST_CLEARING,
/* MGW handling */
/* during assignment; waiting for MGW response to CRCX for BTS */
ST_WAIT_CRCX_BTS,
/* during assignment; waiting for MGW response to MDCX for BTS */
ST_WAIT_MDCX_BTS,
/* during assignment; waiting for MGW response to CRCX for MSC */
ST_WAIT_CRCX_MSC,
/* MT (inbound) handover */
/* Wait for Handover Access from MS/BTS */
ST_WAIT_MT_HO_ACC,
/* Wait for RR Handover Complete from MS/BTS */
ST_WAIT_MT_HO_COMPL,
/* MO (outbound) handover */
/* Wait for Handover Command / Handover Required Reject from MSC */
ST_WAIT_MO_HO_CMD,
/* Wait for Clear Command from MSC */
ST_MO_HO_PROCEEDING,
/* Internal HO handling */
/* Wait for the handover logic to complete the handover */
ST_WAIT_HO_COMPL,
/* during handover; waiting for MGW response to MDCX for BTS */
ST_WAIT_MDCX_BTS_HO,
};
static const struct value_string gscon_fsm_event_names[] = {
{GSCON_EV_A_CONN_IND, "MT-CONNECT.ind"},
{GSCON_EV_A_CONN_REQ, "MO-CONNECT.req"},
{GSCON_EV_A_CONN_CFM, "MO-CONNECT.cfm"},
{GSCON_EV_A_ASSIGNMENT_CMD, "ASSIGNMENT_CMD"},
{GSCON_EV_A_CLEAR_CMD, "CLEAR_CMD"},
{GSCON_EV_A_DISC_IND, "DISCONNET.ind"},
{GSCON_EV_A_HO_REQ, "HANDOVER_REQUEST"},
{GSCON_EV_RR_ASS_COMPL, "RR_ASSIGN_COMPL"},
{GSCON_EV_RR_ASS_FAIL, "RR_ASSIGN_FAIL"},
{GSCON_EV_RLL_REL_IND, "RLL_RELEASE.ind"},
{GSCON_EV_RSL_CONN_FAIL, "RSL_CONN_FAIL.ind"},
{GSCON_EV_RSL_CLEAR_COMPL, "RSL_CLEAR_COMPLETE"},
{GSCON_EV_MO_DTAP, "MO-DTAP"},
{GSCON_EV_MT_DTAP, "MT-DTAP"},
{GSCON_EV_TX_SCCP, "TX_SCCP"},
{GSCON_EV_MGW_FAIL_BTS, "MGW_FAILURE_BTS"},
{GSCON_EV_MGW_FAIL_MSC, "MGW_FAILURE_MSC"},
{GSCON_EV_MGW_CRCX_RESP_BTS, "MGW_CRCX_RESPONSE_BTS"},
{GSCON_EV_MGW_MDCX_RESP_BTS, "MGW_MDCX_RESPONSE_BTS"},
{GSCON_EV_MGW_CRCX_RESP_MSC, "MGW_CRCX_RESPONSE_MSC"},
{GSCON_EV_MGW_MDCX_RESP_MSC, "MGW_MDCX_RESPONSE_MSC"},
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
{GSCON_EV_HO_START, "HO_START"},
{GSCON_EV_HO_TIMEOUT, "HO_TIMEOUT"},
{GSCON_EV_HO_FAIL, "HO_FAIL"},
{GSCON_EV_HO_COMPL, "HO_COMPL"},
{GSCON_EV_LCLS_FAIL, "LCLS_FAIL"},
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
{0, NULL}
};
/* Depending on the channel mode and rate, set the codec type that is signalled
* towards the MGW. */
void bsc_subscr_pick_codec(struct mgcp_conn_peer *conn_peer, struct gsm_subscriber_connection *conn)
{
switch (conn->user_plane.chan_mode) {
case GSM48_CMODE_SPEECH_V1:
if (conn->user_plane.full_rate)
conn_peer->codecs[0] = CODEC_GSM_8000_1;
else
conn_peer->codecs[0] = CODEC_GSMHR_8000_1;
conn_peer->codecs_len = 1;
break;
case GSM48_CMODE_SPEECH_EFR:
conn_peer->codecs[0] = CODEC_GSMEFR_8000_1;
conn_peer->codecs_len = 1;
break;
case GSM48_CMODE_SPEECH_AMR:
conn_peer->codecs[0] = CODEC_AMR_8000_1;
conn_peer->codecs_len = 1;
break;
default:
conn_peer->codecs_len = 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
/* Send data SCCP message through SCCP connection. All sigtran messages
* that are send from this FSM must use this function. Never use
* osmo_bsc_sigtran_send() directly since this would defeat the checks
* provided by this function. */
static void sigtran_send(struct gsm_subscriber_connection *conn, struct msgb *msg, struct osmo_fsm_inst *fi)
{
int rc;
/* Make sure that we only attempt to send SCCP messages if we have
* a life SCCP connection. Otherwise drop the message. */
if (fi->state == ST_INIT || fi->state == ST_WAIT_CC) {
LOGPFSML(fi, LOGL_ERROR, "No active SCCP connection, dropping message!\n");
msgb_free(msg);
return;
}
rc = osmo_bsc_sigtran_send(conn, msg);
if (rc < 0)
LOGPFSML(fi, LOGL_ERROR, "Unable to deliver SCCP message!\n");
}
/* Add the LCLS BSS Status IE to a BSSMAP message. We assume this is
* called on a msgb that was returned by gsm0808_create_ass_compl() */
static void bssmap_add_lcls_status(struct msgb *msg, enum gsm0808_lcls_status status)
{
OSMO_ASSERT(msg->l3h[0] == BSSAP_MSG_BSS_MANAGEMENT);
OSMO_ASSERT(msg->l3h[2] == BSS_MAP_MSG_ASSIGMENT_COMPLETE ||
msg->l3h[2] == BSS_MAP_MSG_HANDOVER_RQST_ACKNOWLEDGE ||
msg->l3h[2] == BSS_MAP_MSG_HANDOVER_COMPLETE ||
msg->l3h[2] == BSS_MAP_MSG_HANDOVER_PERFORMED);
OSMO_ASSERT(msgb_tailroom(msg) >= 2);
/* append IE to end of message */
msgb_tv_put(msg, GSM0808_IE_LCLS_BSS_STATUS, status);
/* increment the "length" byte in the BSSAP header */
msg->l3h[1] += 2;
}
/* Add (append) the LCLS BSS Status IE to a BSSMAP message, if there is any LCLS
* active on the given \a conn */
static void bssmap_add_lcls_status_if_needed(struct gsm_subscriber_connection *conn,
struct msgb *msg)
{
enum gsm0808_lcls_status status = lcls_get_status(conn);
if (status != 0xff) {
LOGPFSM(conn->fi, "Adding LCLS BSS-Status (%s) to %s\n",
gsm0808_lcls_status_name(status),
gsm0808_bssmap_name(msg->l3h[2]));
bssmap_add_lcls_status(msg, status);
}
}
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
/* Generate and send assignment complete message */
static void send_ass_compl(struct gsm_lchan *lchan, struct osmo_fsm_inst *fi, bool voice)
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 msgb *resp;
struct gsm0808_speech_codec sc;
struct gsm0808_speech_codec *sc_ptr = 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
struct gsm_subscriber_connection *conn;
struct sockaddr_storage *addr_local = NULL;
int perm_spch = 0;
uint8_t chosen_channel;
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
conn = lchan->conn;
OSMO_ASSERT(conn);
/* apply LCLS configuration (if any) */
lcls_apply_config(conn);
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
LOGPFSML(fi, LOGL_DEBUG, "Sending assignment complete message... (id=%i)\n", conn->sccp.conn_id);
/* Generate voice related fields */
if (voice) {
perm_spch = gsm0808_permitted_speech(lchan->type, lchan->tch_mode);
switch (conn->sccp.msc->a.asp_proto) {
case OSMO_SS7_ASP_PROT_IPA:
/* don't add any AoIP specific fields. CIC allocated by MSC */
break;
default:
OSMO_ASSERT(lchan->abis_ip.ass_compl.valid);
addr_local = &conn->user_plane.aoip_rtp_addr_local;
/* Extrapolate speech codec from speech mode */
gsm0808_speech_codec_from_chan_type(&sc, perm_spch);
sc_ptr = &sc;
break;
}
/* FIXME: AMR codec configuration must be derived from lchan1! */
}
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
chosen_channel = gsm0808_chosen_channel(lchan->tch_mode, lchan->type);
if (!chosen_channel)
LOGP(DMSC, LOGL_ERROR, "Unknown lchan type or TCH mode: %s\n", gsm_lchan_name(lchan));
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
/* Generate message */
resp = gsm0808_create_ass_compl(lchan->abis_ip.ass_compl.rr_cause,
chosen_channel,
lchan->encr.alg_id, perm_spch,
addr_local, sc_ptr, 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
if (!resp) {
LOGPFSML(fi, LOGL_ERROR, "Failed to generate assignment completed message! (id=%i)\n",
conn->sccp.conn_id);
}
/* Add LCLS BSS-Status IE in case there is any LCLS status for this connection */
bssmap_add_lcls_status_if_needed(conn, resp);
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
sigtran_send(conn, resp, fi);
}
/* forward MT DTAP from BSSAP side to RSL side */
static void submit_dtap(struct gsm_subscriber_connection *conn, struct msgb *msg, struct osmo_fsm_inst *fi)
{
int rc;
struct msgb *resp = NULL;
OSMO_ASSERT(fi);
OSMO_ASSERT(msg);
OSMO_ASSERT(conn);
rc = gsm0808_submit_dtap(conn, msg, OBSC_LINKID_CB(msg), 1);
if (rc != 0) {
LOGPFSML(fi, LOGL_ERROR, "Tx BSSMAP CLEAR REQUEST to MSC\n");
resp = gsm0808_create_clear_rqst(GSM0808_CAUSE_EQUIPMENT_FAILURE);
sigtran_send(conn, resp, fi);
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 0);
return;
}
}
/* forward MO DTAP from RSL side to BSSAP side */
static void forward_dtap(struct gsm_subscriber_connection *conn, struct msgb *msg, struct osmo_fsm_inst *fi)
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 msgb *resp = NULL;
OSMO_ASSERT(msg);
OSMO_ASSERT(conn);
resp = gsm0808_create_dtap(msg, OBSC_LINKID_CB(msg));
sigtran_send(conn, resp, fi);
}
/* In case there are open MGCP connections, toss
* those connections */
static void toss_mgcp_conn(struct gsm_subscriber_connection *conn, struct osmo_fsm_inst *fi)
{
LOGPFSML(fi, LOGL_ERROR, "tossing all MGCP connections...\n");
if (conn->user_plane.fi_bts) {
mgcp_conn_delete(conn->user_plane.fi_bts);
conn->user_plane.fi_bts = NULL;
}
if (conn->user_plane.fi_msc) {
mgcp_conn_delete(conn->user_plane.fi_msc);
conn->user_plane.fi_msc = NULL;
}
if (conn->user_plane.mgw_endpoint) {
talloc_free(conn->user_plane.mgw_endpoint);
conn->user_plane.mgw_endpoint = NULL;
}
}
static void gscon_fsm_init(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct osmo_scu_prim *scu_prim = NULL;
struct msgb *msg = NULL;
int rc;
switch (event) {
case GSCON_EV_A_CONN_REQ:
/* RLL ESTABLISH IND with initial L3 Message */
msg = data;
/* FIXME: Extract Mobile ID and update FSM using osmo_fsm_inst_set_id()
* i.e. we will probably extract the mobile identity earlier, where the
* imsi filter code is. Then we could just use it here.
* related: OS#2969 */
rc = osmo_bsc_sigtran_open_conn(conn, msg);
if (rc < 0) {
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_ERROR, NULL);
} else {
/* SCCP T(conn est) is 1-2 minutes, way too long. The MS will timeout
* using T3210 (20s), T3220 (5s) or T3230 (10s) */
osmo_fsm_inst_state_chg(fi, ST_WAIT_CC, 20, 993210);
}
break;
case GSCON_EV_A_CONN_IND:
scu_prim = data;
if (!conn->sccp.msc) {
LOGPFSML(fi, LOGL_NOTICE, "N-CONNECT.ind from unknown MSC %s\n",
osmo_sccp_addr_dump(&scu_prim->u.connect.calling_addr));
osmo_sccp_tx_disconn(conn->sccp.msc->a.sccp_user, scu_prim->u.connect.conn_id,
&scu_prim->u.connect.called_addr, 0);
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_REGULAR, NULL);
}
/* FIXME: Extract optional IMSI and update FSM using osmo_fsm_inst_set_id()
* related: OS2969 (same as above) */
LOGPFSML(fi, LOGL_NOTICE, "No support for MSC-originated SCCP Connections yet\n");
osmo_sccp_tx_disconn(conn->sccp.msc->a.sccp_user, scu_prim->u.connect.conn_id,
&scu_prim->u.connect.called_addr, 0);
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_REGULAR, NULL);
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* We've sent the CONNECTION.req to the SCCP provider and are waiting for CC from MSC */
static void gscon_fsm_wait_cc(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
switch (event) {
case GSCON_EV_A_CONN_CFM:
/* MSC has confirmed the connection, we now change into the
* active state and wait there for further operations */
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 0);
/* if there's user payload, forward it just like EV_MT_DTAP */
/* FIXME: Question: if there's user payload attached to the CC, forward it like EV_MT_DTAP? */
break;
default:
OSMO_ASSERT(false);
break;
}
}
static const char *get_mgw_ep_name(struct gsm_subscriber_connection *conn)
{
static char ep_name[256];
struct bsc_msc_data *msc = conn->sccp.msc;
switch (conn->sccp.msc->a.asp_proto) {
case OSMO_SS7_ASP_PROT_IPA:
/* derive endpoint name from CIC on A interface side */
snprintf(ep_name, sizeof(ep_name), "%x@mgw",
mgcp_port_to_cic(conn->user_plane.rtp_port, msc->rtp_base));
break;
default:
/* use dynamic RTPBRIDGE endpoint allocation in MGW */
osmo_strlcpy(ep_name, ENDPOINT_ID, sizeof(ep_name));
break;
}
return ep_name;
}
#define assignment_failed(fi, cause) \
_assignment_failed(fi, cause, __FILE__, __LINE__)
static void _assignment_failed(struct osmo_fsm_inst *fi, enum gsm0808_cause cause,
const char *file, int line)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct msgb *resp = NULL;
LOGPFSMLSRC(fi, LOGL_ERROR, file, line, "Assignment failed: %s\n", gsm0808_cause_name(cause));
resp = gsm0808_create_assignment_failure(cause, NULL);
sigtran_send(conn, resp, fi);
if (fi->state != ST_ACTIVE)
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 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
/* We're on an active subscriber connection, passing DTAP back and forth */
static void gscon_fsm_active(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct msgb *resp = NULL;
struct mgcp_conn_peer conn_peer;
int rc;
switch (event) {
case GSCON_EV_A_ASSIGNMENT_CMD:
/* MSC requests us to perform assignment, this code section is
* triggered via signal GSCON_EV_A_ASSIGNMENT_CMD from
* bssmap_handle_assignm_req() in osmo_bsc_bssap.c, which does
* the parsing of incoming assignment requests. */
LOGPFSML(fi, LOGL_NOTICE, "Channel assignment: chan_mode=%s, full_rate=%i\n",
get_value_string(gsm48_chan_mode_names, conn->user_plane.chan_mode),
conn->user_plane.full_rate);
/* FIXME: We need to check if current channel is sufficient. If
* yes, do MODIFY. If not, do assignment (see commented lines below) */
switch (conn->user_plane.chan_mode) {
case GSM48_CMODE_SPEECH_V1:
case GSM48_CMODE_SPEECH_EFR:
case GSM48_CMODE_SPEECH_AMR:
/* A voice channel is requested, so we run down the
* mgcp-ass-mgcp state-chain (see FIXME above) */
memset(&conn_peer, 0, sizeof(conn_peer));
bsc_subscr_pick_codec(&conn_peer, conn);
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
conn_peer.call_id = conn->sccp.conn_id;
conn_peer.ptime = 20;
osmo_strlcpy(conn_peer.endpoint, get_mgw_ep_name(conn), sizeof(conn_peer.endpoint));
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
/* (Pre)Change state and create the connection */
osmo_fsm_inst_state_chg(fi, ST_WAIT_CRCX_BTS, MGCP_MGW_TIMEOUT, MGCP_MGW_TIMEOUT_TIMER_NR);
conn->user_plane.fi_bts =
mgcp_conn_create(conn->network->mgw.client, fi, GSCON_EV_MGW_FAIL_BTS,
GSCON_EV_MGW_CRCX_RESP_BTS, &conn_peer);
if (!conn->user_plane.fi_bts) {
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
return;
}
break;
case GSM48_CMODE_SIGN:
/* A signalling channel is requested, so we perform the
* channel assignment directly without performing any
* MGCP actions. ST_WAIT_ASS_CMPL will see by the
* conn->user_plane.chan_mode parameter that this
* assignment is for a signalling channel and will then
* change back to ST_ACTIVE (here) immediately. */
rc = gsm0808_assign_req(conn, conn->user_plane.chan_mode,
conn->user_plane.full_rate);
if (rc == 1) {
send_ass_compl(conn->lchan, fi, false);
return;
} else if (rc != 0) {
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
return;
}
osmo_fsm_inst_state_chg(fi, ST_WAIT_ASS_CMPL, conn->network->T10, 10);
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
break;
default:
/* An unsupported channel is requested, so we have to
* reject this request by sending an assignment failure
* message immediately */
LOGPFSML(fi, LOGL_ERROR, "Requested channel mode is not supported! chan_mode=%s full_rate=%d\n",
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
get_value_string(gsm48_chan_mode_names, conn->user_plane.chan_mode),
conn->user_plane.full_rate);
/* The requested channel mode is not supported */
assignment_failed(fi, GSM0808_CAUSE_REQ_CODEC_TYPE_OR_CONFIG_NOT_SUPP);
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
break;
}
break;
case GSCON_EV_HO_START:
rc = bsc_handover_start_gscon(conn);
if (rc) {
resp = gsm0808_create_clear_rqst(GSM0808_CAUSE_EQUIPMENT_FAILURE);
sigtran_send(conn, resp, fi);
osmo_fsm_inst_state_chg(fi, ST_CLEARING, 0, 0);
return;
}
/* Note: No timeout is set here, T3103 in handover_logic.c
* will generate a GSCON_EV_HO_TIMEOUT event should the
* handover time out, so we do not need another timeout
* here (maybe its worth to think about giving GSCON
* more power over the actual handover process). */
osmo_fsm_inst_state_chg(fi, ST_WAIT_HO_COMPL, 0, 0);
break;
case GSCON_EV_A_HO_REQ:
/* FIXME: reject any handover requests with HO FAIL until implemented */
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* Before we may start the channel assignment we need to get an IP/Port for the
* RTP connection from the MGW */
static void gscon_fsm_wait_crcx_bts(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct mgcp_conn_peer *conn_peer = NULL;
int rc;
switch (event) {
case GSCON_EV_MGW_CRCX_RESP_BTS:
conn_peer = data;
/* Check if the MGW has assigned an enpoint to us, otherwise we
* can not proceed. */
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
if (strlen(conn_peer->endpoint) <= 0) {
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
return;
}
/* Memorize the endpoint name we got assigned from the MGW.
* When the BTS sided connection is done, we need to create
* a second connection on that same endpoint, so we need
* to know its ID */
if (!conn->user_plane.mgw_endpoint)
conn->user_plane.mgw_endpoint = talloc_zero_size(conn, MGCP_ENDPOINT_MAXLEN);
OSMO_ASSERT(conn->user_plane.mgw_endpoint);
osmo_strlcpy(conn->user_plane.mgw_endpoint, conn_peer->endpoint, MGCP_ENDPOINT_MAXLEN);
/* Store the IP-Address and the port the MGW assigned to us,
* then start the channel assignment. */
conn->user_plane.rtp_port = conn_peer->port;
conn->user_plane.rtp_ip = osmo_ntohl(inet_addr(conn_peer->addr));
rc = gsm0808_assign_req(conn, conn->user_plane.chan_mode, conn->user_plane.full_rate);
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
if (rc != 0) {
assignment_failed(fi, GSM0808_CAUSE_RQSTED_SPEECH_VERSION_UNAVAILABLE);
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
return;
}
osmo_fsm_inst_state_chg(fi, ST_WAIT_ASS_CMPL, conn->network->T10, 10);
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
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* We're waiting for an ASSIGNMENT COMPLETE from MS */
static void gscon_fsm_wait_ass_cmpl(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct gsm_lchan *lchan = conn->lchan;
struct mgcp_conn_peer conn_peer;
struct in_addr addr;
int rc;
switch (event) {
case GSCON_EV_RR_ASS_COMPL:
switch (conn->user_plane.chan_mode) {
case GSM48_CMODE_SPEECH_V1:
case GSM48_CMODE_SPEECH_EFR:
case GSM48_CMODE_SPEECH_AMR:
/* FIXME: What if we are using SCCP-Lite? */
/* We are dealing with a voice channel, so we can not
* confirm the assignment directly. We must first do
* some final steps on the MGCP side. */
/* Prepare parameters with the information we got during the assignment */
memset(&conn_peer, 0, sizeof(conn_peer));
bsc_subscr_pick_codec(&conn_peer, conn);
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
addr.s_addr = osmo_ntohl(lchan->abis_ip.bound_ip);
osmo_strlcpy(conn_peer.addr, inet_ntoa(addr), sizeof(conn_peer.addr));
conn_peer.port = lchan->abis_ip.bound_port;
conn_peer.ptime = 20;
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
/* (Pre)Change state and modify the connection */
osmo_fsm_inst_state_chg(fi, ST_WAIT_MDCX_BTS, MGCP_MGW_TIMEOUT, MGCP_MGW_TIMEOUT_TIMER_NR);
rc = mgcp_conn_modify(conn->user_plane.fi_bts, GSCON_EV_MGW_MDCX_RESP_BTS, &conn_peer);
if (rc != 0) {
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
return;
}
break;
case GSM48_CMODE_SIGN:
/* Confirm the successful assignment on BSSMAP and
* change back into active state */
send_ass_compl(lchan, fi, false);
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_state_chg(fi, ST_ACTIVE, 0, 0);
break;
default:
/* Unsupported modes should have been already filtered
* by gscon_fsm_active(). If we reach the default
* section here anyway than some unsupported mode must
* have made it into the FSM, this would be a bug, so
* we fire an assertion here */
OSMO_ASSERT(false);
break;
}
break;
case GSCON_EV_RR_ASS_FAIL:
{
enum gsm0808_cause cause = GSM0808_CAUSE_RQSTED_TERRESTRIAL_RESOURCE_UNAVAILABLE;
if (data)
cause = *((enum gsm0808_cause*)data);
assignment_failed(fi, cause);
}
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
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* We are waiting for the MGW response to the MDCX */
static void gscon_fsm_wait_mdcx_bts(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct mgcp_conn_peer conn_peer;
struct sockaddr_in *sin = NULL;
switch (event) {
case GSCON_EV_MGW_MDCX_RESP_BTS:
/* Prepare parameters with the connection information we got
* with the assignment command */
memset(&conn_peer, 0, sizeof(conn_peer));
bsc_subscr_pick_codec(&conn_peer, conn);
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
conn_peer.call_id = conn->sccp.conn_id;
sin = (struct sockaddr_in *)&conn->user_plane.aoip_rtp_addr_remote;
conn_peer.port = osmo_ntohs(sin->sin_port);
osmo_strlcpy(conn_peer.addr, inet_ntoa(sin->sin_addr), sizeof(conn_peer.addr));
conn_peer.ptime = 20;
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
/* Make sure we use the same endpoint where we created the
* BTS connection. */
osmo_strlcpy(conn_peer.endpoint, conn->user_plane.mgw_endpoint, sizeof(conn_peer.endpoint));
switch (conn->sccp.msc->a.asp_proto) {
case OSMO_SS7_ASP_PROT_IPA:
/* Send assignment complete message to the MSC */
send_ass_compl(conn->lchan, fi, true);
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_state_chg(fi, ST_ACTIVE, 0, 0);
break;
default:
/* (Pre)Change state and create the connection */
osmo_fsm_inst_state_chg(fi, ST_WAIT_CRCX_MSC, MGCP_MGW_TIMEOUT,
MGCP_MGW_TIMEOUT_TIMER_NR);
conn->user_plane.fi_msc = mgcp_conn_create(conn->network->mgw.client, fi,
GSCON_EV_MGW_FAIL_MSC,
GSCON_EV_MGW_CRCX_RESP_MSC, &conn_peer);
if (!conn->user_plane.fi_msc) {
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
return;
}
break;
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
}
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
static void gscon_fsm_wait_crcx_msc(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct mgcp_conn_peer *conn_peer = NULL;
struct gsm_lchan *lchan = conn->lchan;
struct sockaddr_in *sin = NULL;
switch (event) {
case GSCON_EV_MGW_CRCX_RESP_MSC:
conn_peer = data;
/* Store address information we got in response from the CRCX command. */
sin = (struct sockaddr_in *)&conn->user_plane.aoip_rtp_addr_local;
sin->sin_family = AF_INET;
sin->sin_addr.s_addr = inet_addr(conn_peer->addr);
sin->sin_port = osmo_ntohs(conn_peer->port);
/* Send assignment complete message to the MSC */
send_ass_compl(lchan, fi, true);
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_state_chg(fi, ST_ACTIVE, 0, 0);
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
static void gscon_fsm_clearing(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct msgb *resp;
switch (event) {
case GSCON_EV_RSL_CLEAR_COMPL:
resp = gsm0808_create_clear_complete();
sigtran_send(conn, resp, fi);
/* we cannot terminate the FSM here, as that would send N-DISCCONNET.req
* and 3GPP TS 48.006 Section 9.2 clearly states that SCCP connections must
* always be released from the MSC side*/
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
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* Wait for the handover logic to tell us whether the handover completed,
* failed or has timed out */
static void gscon_fsm_wait_ho_compl(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct mgcp_conn_peer conn_peer;
struct gsm_lchan *lchan = conn->lchan;
struct in_addr addr;
struct msgb *resp;
int rc;
switch (event) {
case GSCON_EV_HO_COMPL:
/* The handover logic informs us that the handover has been
* completet. Now we have to tell the MGW the IP/Port on the
* new BTS so that the uplink RTP traffic can be redirected
* there. */
/* Prepare parameters with the information we got during the
* handover procedure (via IPACC) */
memset(&conn_peer, 0, sizeof(conn_peer));
bsc_subscr_pick_codec(&conn_peer, conn);
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
addr.s_addr = osmo_ntohl(lchan->abis_ip.bound_ip);
osmo_strlcpy(conn_peer.addr, inet_ntoa(addr), sizeof(conn_peer.addr));
conn_peer.port = lchan->abis_ip.bound_port;
conn_peer.ptime = 20;
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
/* (Pre)Change state and modify the connection */
osmo_fsm_inst_state_chg(fi, ST_WAIT_MDCX_BTS_HO, MGCP_MGW_TIMEOUT, MGCP_MGW_HO_TIMEOUT_TIMER_NR);
rc = mgcp_conn_modify(conn->user_plane.fi_bts, GSCON_EV_MGW_MDCX_RESP_BTS, &conn_peer);
if (rc != 0) {
resp = gsm0808_create_clear_rqst(GSM0808_CAUSE_EQUIPMENT_FAILURE);
sigtran_send(conn, resp, fi);
osmo_fsm_inst_state_chg(fi, ST_CLEARING, 0, 0);
return;
}
break;
case GSCON_EV_HO_TIMEOUT:
case GSCON_EV_HO_FAIL:
/* The handover logic informs us that the handover failed for
* some reason. This means the phone stays on the TS/BTS on
* which it currently is. We will change back to the active
* state again as there are no further operations needed */
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 0);
break;
default:
OSMO_ASSERT(false);
break;
}
}
/* Wait for the MGW to confirm handover related modification of the connection
* parameters */
static void gscon_fsm_wait_mdcx_bts_ho(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
switch (event) {
case GSCON_EV_MGW_MDCX_RESP_BTS:
/* The MGW has confirmed the handover MDCX, and the handover
* is now also done on the RTP side. We may now change back
* to the active state. */
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 0);
break;
case GSCON_EV_MO_DTAP:
forward_dtap(conn, (struct msgb *)data, fi);
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
break;
case GSCON_EV_MT_DTAP:
submit_dtap(conn, (struct msgb *)data, fi);
break;
case GSCON_EV_TX_SCCP:
sigtran_send(conn, (struct msgb *)data, fi);
break;
default:
OSMO_ASSERT(false);
break;
}
}
#define EV_TRANSPARENT_SCCP S(GSCON_EV_TX_SCCP) | S(GSCON_EV_MO_DTAP) | S(GSCON_EV_MT_DTAP)
static const struct osmo_fsm_state gscon_fsm_states[] = {
[ST_INIT] = {
.name = "INIT",
.in_event_mask = S(GSCON_EV_A_CONN_REQ) | S(GSCON_EV_A_CONN_IND),
.out_state_mask = S(ST_WAIT_CC),
.action = gscon_fsm_init,
},
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
[ST_WAIT_CC] = {
.name = "WAIT_CC",
.in_event_mask = S(GSCON_EV_A_CONN_CFM),
.out_state_mask = S(ST_ACTIVE),
.action = gscon_fsm_wait_cc,
},
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
[ST_ACTIVE] = {
.name = "ACTIVE",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_A_ASSIGNMENT_CMD) |
S(GSCON_EV_A_HO_REQ) | S(GSCON_EV_HO_START),
.out_state_mask = S(ST_CLEARING) | S(ST_WAIT_CRCX_BTS) | S(ST_WAIT_ASS_CMPL) |
S(ST_WAIT_MO_HO_CMD) | S(ST_WAIT_HO_COMPL),
.action = gscon_fsm_active,
},
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
[ST_WAIT_CRCX_BTS] = {
.name = "WAIT_CRCX_BTS",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_MGW_CRCX_RESP_BTS),
.out_state_mask = S(ST_ACTIVE) | S(ST_WAIT_ASS_CMPL),
.action = gscon_fsm_wait_crcx_bts,
},
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
[ST_WAIT_ASS_CMPL] = {
.name = "WAIT_ASS_CMPL",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_RR_ASS_COMPL) | S(GSCON_EV_RR_ASS_FAIL),
.out_state_mask = S(ST_ACTIVE) | S(ST_WAIT_MDCX_BTS),
.action = gscon_fsm_wait_ass_cmpl,
},
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
[ST_WAIT_MDCX_BTS] = {
.name = "WAIT_MDCX_BTS",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_MGW_MDCX_RESP_BTS),
.out_state_mask = S(ST_ACTIVE) | S(ST_WAIT_CRCX_MSC),
.action = gscon_fsm_wait_mdcx_bts,
},
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
[ST_WAIT_CRCX_MSC] = {
.name = "WAIT_CRCX_MSC",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_MGW_CRCX_RESP_MSC),
.out_state_mask = S(ST_ACTIVE),
.action = gscon_fsm_wait_crcx_msc,
},
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
[ST_CLEARING] = {
.name = "CLEARING",
.in_event_mask = S(GSCON_EV_RSL_CLEAR_COMPL),
.action = gscon_fsm_clearing,
},
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
/* TODO: external handover, probably it makes sense to break up the
* program flow in handover_logic.c a bit and handle some of the logic
* here? */
[ST_WAIT_MT_HO_ACC] = {
.name = "WAIT_MT_HO_ACC",
},
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
[ST_WAIT_MT_HO_COMPL] = {
.name = "WAIT_MT_HO_COMPL",
},
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
[ST_WAIT_MO_HO_CMD] = {
.name = "WAIT_MO_HO_CMD",
},
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
[ST_MO_HO_PROCEEDING] = {
.name = "MO_HO_PROCEEDING",
},
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
/* Internal handover */
[ST_WAIT_HO_COMPL] = {
.name = "WAIT_HO_COMPL",
.in_event_mask = S(GSCON_EV_HO_COMPL) | S(GSCON_EV_HO_FAIL) | S(GSCON_EV_HO_TIMEOUT),
.out_state_mask = S(ST_ACTIVE) | S(ST_WAIT_MDCX_BTS_HO) | S(ST_CLEARING),
.action = gscon_fsm_wait_ho_compl,
},
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
[ST_WAIT_MDCX_BTS_HO] = {
.name = "WAIT_MDCX_BTS_HO",
.in_event_mask = EV_TRANSPARENT_SCCP | S(GSCON_EV_MGW_MDCX_RESP_BTS),
.action = gscon_fsm_wait_mdcx_bts_ho,
.out_state_mask = S(ST_ACTIVE),
},
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
};
static void gscon_fsm_allstate(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct gsm_subscriber_connection *conn = fi->priv;
struct msgb *resp = NULL;
/* When a connection on the MGW fails, make sure that the reference
* in our book-keeping is erased. */
switch (event) {
case GSCON_EV_MGW_FAIL_BTS:
conn->user_plane.fi_bts = NULL;
break;
case GSCON_EV_MGW_FAIL_MSC:
conn->user_plane.fi_msc = NULL;
break;
}
/* Regular allstate event processing */
switch (event) {
case GSCON_EV_MGW_FAIL_BTS:
case GSCON_EV_MGW_FAIL_MSC:
/* Note: An MGW connection die per definition at any time.
* However, if it dies during the assignment we must return
* with an assignment failure */
OSMO_ASSERT(fi->state != ST_INIT && fi->state != ST_WAIT_CC);
if (fi->state == ST_WAIT_CRCX_BTS || fi->state == ST_WAIT_ASS_CMPL || fi->state == ST_WAIT_MDCX_BTS
|| fi->state == ST_WAIT_CRCX_MSC)
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
break;
case GSCON_EV_A_CLEAR_CMD:
/* MSC tells us to cleanly shut down */
osmo_fsm_inst_state_chg(fi, ST_CLEARING, 0, 0);
gsm0808_clear(conn);
/* FIXME: Release all terestrial resources in ST_CLEARING */
/* According to 3GPP 48.008 3.1.9.1. "The BSS need not wait for the radio channel
* release to be completed or for the guard timer to expire before returning the
* CLEAR COMPLETE message" */
/* Close MGCP connections */
toss_mgcp_conn(conn, fi);
/* FIXME: Question: Is this a hack to force a clear complete from internel?
* nobody seems to send the event from outside? */
osmo_fsm_inst_dispatch(conn->fi, GSCON_EV_RSL_CLEAR_COMPL, NULL);
break;
case GSCON_EV_A_DISC_IND:
/* MSC or SIGTRAN network has hard-released SCCP connection,
* terminate the FSM now. */
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_REGULAR, data);
break;
case GSCON_EV_RLL_REL_IND:
/* BTS reports that one of the LAPDm data links was released */
/* send proper clear request to MSC */
LOGPFSML(fi, LOGL_DEBUG, "Tx BSSMAP CLEAR REQUEST to MSC\n");
resp = gsm0808_create_clear_rqst(GSM0808_CAUSE_RADIO_INTERFACE_MESSAGE_FAILURE);
sigtran_send(conn, resp, fi);
break;
case GSCON_EV_RSL_CONN_FAIL:
LOGPFSML(fi, LOGL_DEBUG, "Tx BSSMAP CLEAR REQUEST to MSC\n");
resp = gsm0808_create_clear_rqst(GSM0808_CAUSE_RADIO_INTERFACE_FAILURE);
sigtran_send(conn, resp, fi);
break;
case GSCON_EV_MGW_MDCX_RESP_MSC:
LOGPFSML(fi, LOGL_DEBUG, "Rx MDCX of MSC side (LCLS?)\n");
break;
case GSCON_EV_LCLS_FAIL:
break;
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
default:
OSMO_ASSERT(false);
break;
}
}
void ho_dtap_cache_flush(struct gsm_subscriber_connection *conn, int send);
static void gscon_cleanup(struct osmo_fsm_inst *fi, enum osmo_fsm_term_cause cause)
{
struct gsm_subscriber_connection *conn = fi->priv;
if (conn->ho) {
LOGPFSML(fi, LOGL_DEBUG, "Releasing handover state\n");
bsc_clear_handover(conn, 1);
conn->ho = NULL;
}
if (conn->secondary_lchan) {
LOGPFSML(fi, LOGL_DEBUG, "Releasing secondary_lchan\n");
lchan_release(conn->secondary_lchan, 0, RSL_REL_LOCAL_END);
conn->secondary_lchan = NULL;
}
if (conn->lchan) {
LOGPFSML(fi, LOGL_DEBUG, "Releasing lchan\n");
lchan_release(conn->lchan, 0, RSL_REL_LOCAL_END);
conn->lchan = NULL;
}
if (conn->sccp.state != SUBSCR_SCCP_ST_NONE) {
LOGPFSML(fi, LOGL_DEBUG, "Disconnecting SCCP\n");
struct bsc_msc_data *msc = conn->sccp.msc;
/* FIXME: include a proper cause value / error message? */
osmo_sccp_tx_disconn(msc->a.sccp_user, conn->sccp.conn_id, &msc->a.bsc_addr, 0);
conn->sccp.state = SUBSCR_SCCP_ST_NONE;
}
/* drop pending messages */
ho_dtap_cache_flush(conn, 0);
penalty_timers_free(&conn->hodec2.penalty_timers);
if (conn->bsub) {
LOGPFSML(fi, LOGL_DEBUG, "Putting bsc_subscr\n");
bsc_subscr_put(conn->bsub);
conn->bsub = 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
llist_del(&conn->entry);
talloc_free(conn);
fi->priv = NULL;
}
static void gscon_pre_term(struct osmo_fsm_inst *fi, enum osmo_fsm_term_cause cause)
{
struct gsm_subscriber_connection *conn = fi->priv;
/* Make sure all possibly still open MGCP connections get closed */
toss_mgcp_conn(conn, fi);
if (conn->lcls.fi) {
/* request termination of LCLS FSM */
osmo_fsm_inst_term(conn->lcls.fi, cause, NULL);
conn->lcls.fi = 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
}
static int gscon_timer_cb(struct osmo_fsm_inst *fi)
{
struct gsm_subscriber_connection *conn = fi->priv;
switch (fi->T) {
case 993210:
/* MSC has not responded/confirmed connection with CC, this
* could indicate a bad SCCP connection. We now inform the the
* FSM that controls the BSSMAP reset about the event. Maybe
* a BSSMAP reset is necessary. */
a_reset_conn_fail(conn->sccp.msc->a.reset_fsm);
/* Since we could not reach the MSC, we give up and terminate
* the FSM instance now (N-DISCONNET.req is sent in
* gscon_cleanup() above) */
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(fi, OSMO_FSM_TERM_REGULAR, NULL);
break;
case 10: /* Assignment Failed */
assignment_failed(fi, GSM0808_CAUSE_RADIO_INTERFACE_FAILURE);
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
break;
case MGCP_MGW_TIMEOUT_TIMER_NR: /* Assignment failed (no response from MGW) */
assignment_failed(fi, GSM0808_CAUSE_EQUIPMENT_FAILURE);
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
break;
case MGCP_MGW_HO_TIMEOUT_TIMER_NR: /* Handover failed (no response from MGW) */
osmo_fsm_inst_state_chg(fi, ST_ACTIVE, 0, 0);
break;
default:
OSMO_ASSERT(false);
}
return 0;
}
static struct osmo_fsm gscon_fsm = {
.name = "SUBSCR_CONN",
.states = gscon_fsm_states,
.num_states = ARRAY_SIZE(gscon_fsm_states),
.allstate_event_mask = S(GSCON_EV_A_DISC_IND) | S(GSCON_EV_A_CLEAR_CMD) | S(GSCON_EV_RSL_CONN_FAIL) |
S(GSCON_EV_RLL_REL_IND) | S(GSCON_EV_MGW_FAIL_BTS) | S(GSCON_EV_MGW_FAIL_MSC) |
S(GSCON_EV_MGW_MDCX_RESP_MSC) | S(GSCON_EV_LCLS_FAIL),
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
.allstate_action = gscon_fsm_allstate,
.cleanup = gscon_cleanup,
.pre_term = gscon_pre_term,
.timer_cb = gscon_timer_cb,
.log_subsys = DMSC,
.event_names = gscon_fsm_event_names,
};
/* Allocate a subscriber connection and its associated FSM */
struct gsm_subscriber_connection *bsc_subscr_con_allocate(struct gsm_network *net)
{
struct gsm_subscriber_connection *conn;
static bool g_initialized = false;
if (!g_initialized) {
osmo_fsm_register(&gscon_fsm);
osmo_fsm_register(&lcls_fsm);
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
g_initialized = true;
}
conn = talloc_zero(net, struct gsm_subscriber_connection);
if (!conn)
return NULL;
conn->network = net;
INIT_LLIST_HEAD(&conn->ho_dtap_cache);
/* BTW, penalty timers will be initialized on-demand. */
conn->sccp.conn_id = -1;
/* don't allocate from 'conn' context, as gscon_cleanup() will call talloc_free(conn) before
* libosmocore will call talloc_free(conn->fi), i.e. avoid use-after-free during cleanup */
conn->fi = osmo_fsm_inst_alloc(&gscon_fsm, net, conn, LOGL_NOTICE, NULL);
if (!conn->fi) {
talloc_free(conn);
return NULL;
}
/* initialize to some magic values that indicate "IE not [yet] received" */
conn->lcls.config = 0xff;
conn->lcls.control = 0xff;
conn->lcls.fi = osmo_fsm_inst_alloc_child(&lcls_fsm, conn->fi, GSCON_EV_LCLS_FAIL);
if (!conn->lcls.fi) {
osmo_fsm_inst_term(conn->fi, OSMO_FSM_TERM_ERROR, NULL);
return NULL;
}
conn->lcls.fi->priv = conn;
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
llist_add_tail(&conn->entry, &net->subscr_conns);
return conn;
}