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

448 lines
14 KiB
C
Raw Normal View History

large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* Select a suitable lchan from a given cell.
*
* (C) 2008 by Harald Welte <laforge@gnumonks.org>
* (C) 2008, 2009 by Holger Hans Peter Freyther <zecke@selfish.org>
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
* (C) 2018-2022 by sysmocom - s.f.m.c. GmbH <info@sysmocom.de>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
*
* 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/>.
*
*/
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
#include <stdlib.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/debug.h>
#include <osmocom/bsc/gsm_data.h>
#include <osmocom/bsc/timeslot_fsm.h>
#include <osmocom/bsc/lchan_fsm.h>
#include <osmocom/bsc/lchan_select.h>
Move struct gsm_bts: gsm_data.* => bts.* Place all code related to the object into the related file. Having all the data model in one file made sense in early stage of development to make progress quickly, but nowadays it hurts more than helps, due to constantly growing size and more and more bits being added to the model, gaining in complexity. Currently, having lots of different objects mixed up in gsm_data.h is a hole of despair, where nobody can make any sense were to properly put new stuff in, ending up with functions related to same object in different files or with wrong prefixes, declarations of non-existing functions, etc. because people cannot make up their mind on strict relation to objects in the data model. Splitting them in files really helps finding code operating on a specific object and helping with logically splitting in the future. Change-Id: I00c15f5285b5c1a0109279b7ab192d5467a04ece
2020-07-15 18:53:16 +00:00
#include <osmocom/bsc/bts.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
struct lchan_select_ts_list {
struct gsm_bts_trx_ts **list;
unsigned int num;
};
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
const struct value_string lchan_select_reason_names[] = {
OSMO_VALUE_STRING(SELECT_FOR_MS_CHAN_REQ),
OSMO_VALUE_STRING(SELECT_FOR_ASSIGNMENT),
OSMO_VALUE_STRING(SELECT_FOR_HANDOVER),
{0, NULL}
};
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
static struct gsm_lchan *pick_better_lchan(struct gsm_lchan *a, struct gsm_lchan *b)
{
if (!a)
return b;
if (!b)
return a;
/* comparing negative dBm values: smaller value means less interference. */
if (b->interf_dbm < a->interf_dbm)
return b;
return a;
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
static struct gsm_lchan *_lc_find(struct lchan_select_ts_list *ts_list,
enum gsm_phys_chan_config pchan,
enum gsm_phys_chan_config as_pchan,
bool allow_pchan_switch, bool log)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
struct gsm_lchan *lchan;
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
struct gsm_lchan *found_lchan = NULL;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
#define LOGPLCHANALLOC(fmt, args...) do { \
lchan_avail(): omit logging for handover decision 2 Add bool log argument to lchan_avail_by_type() and omit logging when passed as false. From handover_decision_2.c, pass 'log' as false, from all other callers pass true, i.e. for unchanged behavior. Rationale: Usually, we use lchan_avail_by_type() to select a new lchan to initiate actual service. For that, it is interesting to see how osmo-bsc decides which lchan will be used. For handover decision 2, we since recently call lchan_avail_by_type() for each and every handover candidate, to determine whether it will occupy a dynamic timeslot or not (to know whether we would congest the other TCH kind). So this happens for each permutation of source lchan and target cell. That produces a lot of logging, out of proportion of being useful to the maintainer. Change-Id: Ia403f8fc853ca9ea9e81f7a7395df6b23845ebed
2021-01-16 16:54:25 +00:00
if (log) \
fix TCH/H allocation: use half occupied dyn TS instead of switching more dyn TS Change-Id: I5a8d943f31774af00664d037550be14e767d312a
2020-11-13 03:19:24 +00:00
LOGP(DRLL, LOGL_DEBUG, "looking for lchan %s%s%s%s: " fmt, \
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
gsm_pchan_name(pchan), \
pchan == as_pchan ? "" : " as ", \
fix TCH/H allocation: use half occupied dyn TS instead of switching more dyn TS Change-Id: I5a8d943f31774af00664d037550be14e767d312a
2020-11-13 03:19:24 +00:00
pchan == as_pchan ? "" : gsm_pchan_name(as_pchan), \
((pchan != as_pchan) && !allow_pchan_switch) ? " without pchan switch" : "", \
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
## args); \
} while (0)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
for (unsigned int tn = 0; tn < ts_list->num; tn++) {
struct gsm_bts_trx_ts *ts = ts_list->list[tn];
replace ts_*_for_each_lchan() with ts_for_n_lchans() So far we have a couple of macros iterating a specific number of lchans, depending on dynamic timeslot state etc. With addition of VAMOS lchans, this would become more complex and bloated. Instead of separate iteration macros for each situation, only have one that takes a number of lchans as argument. That allows to more clearly pick the number of lchans, especially for non-trivial VAMOS scenarios. Related: SYS#5315 OS#4940 Change-Id: Ib2c6baf73a81ba371143ba5adc912aef6f79238d
2021-05-19 00:28:26 +00:00
int lchans_as_pchan;
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* The caller first selects what kind of TS to search in, e.g. looking for exact
Rename osmo dyn ts enums to contain SDCCH8 They will gain support to be activated as SDCCH/8 soon too. Related: OS#5309 Depends: libosmocore.git I56dcfe4d17899630b17f80145c3ced72f1e91e68 Change-Id: Id5b89fe589a52ff88486435ac43809edb4b80f98
2021-06-28 10:19:13 +00:00
* GSM_PCHAN_TCH_F, or maybe among dynamic GSM_PCHAN_OSMO_DYN... */
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (ts->pchan_on_init != pchan) {
LOGPLCHANALLOC("%s is != %s\n", gsm_ts_and_pchan_name(ts),
gsm_pchan_name(pchan));
continue;
}
/* Next, is this timeslot in or can it be switched to the pchan we want to use it for? */
fix TCH/H allocation: use half occupied dyn TS instead of switching more dyn TS Change-Id: I5a8d943f31774af00664d037550be14e767d312a
2020-11-13 03:19:24 +00:00
if (!ts_usable_as_pchan(ts, as_pchan, allow_pchan_switch)) {
LOGPLCHANALLOC("%s is not usable as %s%s\n", gsm_ts_and_pchan_name(ts),
gsm_pchan_name(as_pchan),
allow_pchan_switch ? "" : " without pchan switch");
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
continue;
}
/* TS is (going to be) in desired pchan mode. Go ahead and check for an available lchan. */
replace ts_*_for_each_lchan() with ts_for_n_lchans() So far we have a couple of macros iterating a specific number of lchans, depending on dynamic timeslot state etc. With addition of VAMOS lchans, this would become more complex and bloated. Instead of separate iteration macros for each situation, only have one that takes a number of lchans as argument. That allows to more clearly pick the number of lchans, especially for non-trivial VAMOS scenarios. Related: SYS#5315 OS#4940 Change-Id: Ib2c6baf73a81ba371143ba5adc912aef6f79238d
2021-05-19 00:28:26 +00:00
lchans_as_pchan = pchan_subslots(as_pchan);
ts_for_n_lchans(lchan, ts, lchans_as_pchan) {
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
struct gsm_lchan *was = found_lchan;
if (lchan->fi->state != LCHAN_ST_UNUSED) {
LOGPLCHANALLOC("%s ss=%d in type=%s,state=%s not suitable\n",
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
gsm_ts_and_pchan_name(ts), lchan->nr,
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
gsm_lchant_name(lchan->type),
osmo_fsm_inst_state_name(lchan->fi));
continue;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
found_lchan = pick_better_lchan(found_lchan, lchan);
if (found_lchan != was)
LOGPLCHANALLOC("%s ss=%d interf=%u=%ddBm is %s%s\n",
gsm_ts_and_pchan_name(ts), lchan->nr,
lchan->interf_band, lchan->interf_dbm,
was == NULL ? "available" : "better",
ts->pchan_is != as_pchan ? ", after dyn PCHAN change" : "");
else
LOGPLCHANALLOC("%s ss=%d interf=%u=%ddBm is also available but not better\n",
gsm_ts_and_pchan_name(ts), lchan->nr,
lchan->interf_band, lchan->interf_dbm);
/* When picking an lchan with least interference, continue to loop across all lchans. When
* ignoring interference levels, return the first match. */
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
if (found_lchan && !ts->trx->bts->chan_alloc_avoid_interf)
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
return found_lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
}
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
if (found_lchan)
LOGPLCHANALLOC("%s ss=%d interf=%ddBm%s is the best pick\n",
gsm_ts_and_pchan_name(found_lchan->ts), found_lchan->nr,
found_lchan->interf_dbm,
found_lchan->ts->pchan_is != as_pchan ? ", after dyn PCHAN change," : "");
else
log: fix missing newline in lchan_select.c Change-Id: Iea8bc73ba6da24f353ad422cdea3fa88173a9a1c
2021-07-19 17:07:52 +00:00
LOGPLCHANALLOC("Nothing found\n");
RES IND: pick lchan with least interference In lchan_select, do not return on the first available lchan, but iterate through all available lchans. Among those that would match, pick the one with the lowest interference level as indicated by earlier RES IND. Lchans that have no interference ratings are picked last. This feature is off by default, enable per BTS with network bts N channel allocator avoid-interference 1 osmo-bsc still does the same ascending/descending lchan allocation as before this patch with the default setting of: channel allocator avoid-interference 0 Related: SYS#5313 Change-Id: I844494092193811dfd9fa4d52983cbaed0fc9248
2021-07-08 02:20:59 +00:00
return found_lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#undef LOGPLCHANALLOC
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
static struct gsm_lchan *lc_dyn_find(struct lchan_select_ts_list *ts_list,
enum gsm_phys_chan_config pchan,
enum gsm_phys_chan_config dyn_as_pchan,
bool log)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
struct gsm_lchan *lchan;
fix TCH/H allocation: use half occupied dyn TS instead of switching more dyn TS Change-Id: I5a8d943f31774af00664d037550be14e767d312a
2020-11-13 03:19:24 +00:00
/* First find an lchan that needs no change in its timeslot pchan mode.
* In particular, this ensures that handover to a dynamic timeslot in TCH/H favors timeslots that are currently
* using only one of two TCH/H, so that we don't switch more dynamic timeslots to TCH/H than necessary.
* For non-dynamic timeslots, it is not necessary to do a second pass with allow_pchan_switch ==
* true, because they never switch anyway. */
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
if ((lchan = _lc_find(ts_list, pchan, dyn_as_pchan, false, log)))
return lchan;
if ((lchan = _lc_find(ts_list, pchan, dyn_as_pchan, true, log)))
return lchan;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return NULL;
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
static struct gsm_lchan *lc_find(struct lchan_select_ts_list *ts_list,
enum gsm_phys_chan_config pchan,
bool log)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
return _lc_find(ts_list, pchan, pchan, false, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
enum gsm_chan_t chan_mode_to_chan_type(enum gsm48_chan_mode chan_mode, enum channel_rate chan_rate)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
ensure chan_mode comparisons in non-VAMOS mode Both VAMOS- and non-VAMOS speech modes should result in indentical voice handling. So make sure that all chan_modes are converted to non-vamos before comparing / evaluating in switch statements. Change-Id: I791e7966b1f8eaa3299a8a46abeb313cf5136e0b
2021-05-28 17:18:11 +00:00
switch (gsm48_chan_mode_to_non_vamos(chan_mode)) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
case GSM48_CMODE_SIGN:
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
switch (chan_rate) {
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
case CH_RATE_SDCCH:
return GSM_LCHAN_SDCCH;
case CH_RATE_HALF:
return GSM_LCHAN_TCH_H;
case CH_RATE_FULL:
return GSM_LCHAN_TCH_F;
default:
return GSM_LCHAN_NONE;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
case GSM48_CMODE_SPEECH_EFR:
lchan_select: dont allow half rate EFR to be selected The function lchan_select_by_chan_mode() is prone to select an half rate lchan when EFR is used, even though EFR is not defined for half-rate. Lets protect against that. Change-Id: I961d9aaba81424053ab1dc04ce7799e716af4cd8 Related: OS#3503
2019-01-17 12:41:17 +00:00
/* EFR works over FR channels only */
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
if (chan_rate != CH_RATE_FULL)
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
return GSM_LCHAN_NONE;
lchan_select: dont allow half rate EFR to be selected The function lchan_select_by_chan_mode() is prone to select an half rate lchan when EFR is used, even though EFR is not defined for half-rate. Lets protect against that. Change-Id: I961d9aaba81424053ab1dc04ce7799e716af4cd8 Related: OS#3503
2019-01-17 12:41:17 +00:00
/* fall through */
case GSM48_CMODE_SPEECH_V1:
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
case GSM48_CMODE_SPEECH_AMR:
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
switch (chan_rate) {
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
case CH_RATE_HALF:
return GSM_LCHAN_TCH_H;
case CH_RATE_FULL:
return GSM_LCHAN_TCH_F;
default:
return GSM_LCHAN_NONE;
assignment_fsm: Properly support assigning signalling mode TCH/x To support the 3 possible preferences, the changes needed were: - Replace 'full_rate' bool with a 3 option enum to represent the channels types for signalling - Switch from _pref/_alt to using an array sorted in preference order Originally merged as Change-Id I4c7499c8c866ea3ff7b1327edb3615d003d927d3, reverted because the change broke voice calls. Re-submitting with the fix: don't forget to set conn->assignment.requires_voice_stream. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I7513d2cbe8b695ba6f031ad11560c63a6535cf2d
2019-03-07 15:32:02 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
default:
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
return GSM_LCHAN_NONE;
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
static int qsort_func(const void *_a, const void *_b)
{
const struct gsm_bts_trx *trx_a = *(const struct gsm_bts_trx **)_a;
const struct gsm_bts_trx *trx_b = *(const struct gsm_bts_trx **)_b;
int pwr_a = trx_a->nominal_power - trx_a->max_power_red;
int pwr_b = trx_b->nominal_power - trx_b->max_power_red;
/* Sort in descending order */
return pwr_b - pwr_a;
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
static void populate_ts_list(struct lchan_select_ts_list *ts_list,
struct gsm_bts *bts,
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
bool chan_alloc_reverse,
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
bool sort_by_trx_power,
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
bool log)
{
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
struct gsm_bts_trx **trx_list;
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
struct gsm_bts_trx *trx;
unsigned int num = 0;
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
/* Allocate an array with pointers to all TRX instances of a BTS */
trx_list = talloc_array_ptrtype(bts, trx_list, bts->num_trx);
OSMO_ASSERT(trx_list != NULL);
llist_for_each_entry(trx, &bts->trx_list, list)
trx_list[trx->nr] = trx;
/* Sort by TRX power in descending order (if needed) */
if (sort_by_trx_power)
qsort(&trx_list[0], bts->num_trx, sizeof(trx), &qsort_func);
for (unsigned int trxn = 0; trxn < bts->num_trx; trxn++) {
trx = trx_list[trxn];
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
for (unsigned int tn = 0; tn < ARRAY_SIZE(trx->ts); tn++) {
struct gsm_bts_trx_ts *ts = &trx->ts[tn];
if (ts_is_usable(ts))
ts_list->list[num++] = ts;
else if (log)
LOGP(DRLL, LOGL_DEBUG, "%s is not usable\n", gsm_ts_name(ts));
}
}
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
talloc_free(trx_list);
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
ts_list->num = num;
/* Reverse the timeslot list if required */
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
if (chan_alloc_reverse) {
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
for (unsigned int tn = 0; tn < num / 2; tn++) {
struct gsm_bts_trx_ts *temp = ts_list->list[tn];
ts_list->list[tn] = ts_list->list[num - tn - 1];
ts_list->list[num - tn - 1] = temp;
}
}
}
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
static bool chan_alloc_ass_dynamic_reverse(struct gsm_bts *bts,
void *ctx, bool log)
{
const struct load_counter *ll = &bts->c0->lchan_load;
const struct gsm_lchan *old_lchan = ctx;
unsigned int lchan_load;
int avg_ul_rxlev;
OSMO_ASSERT(old_lchan != NULL);
OSMO_ASSERT(old_lchan->ts->trx->bts == bts);
#define LOG_COND(fmt, args...) do { \
if (log) \
LOG_LCHAN(old_lchan, LOGL_DEBUG, fmt, ## args); \
} while (0)
/* Condition a) Channel load on the C0 (BCCH carrier) */
lchan_load = ll->total ? ll->used * 100 / ll->total : 0;
if (lchan_load < bts->chan_alloc_dyn_params.c0_chan_load_thresh) {
LOG_COND("C0 Channel Load %u%% < thresh %u%% => using ascending order\n",
lchan_load, bts->chan_alloc_dyn_params.c0_chan_load_thresh);
return false;
}
/* Condition b) average Uplink RxLev */
avg_ul_rxlev = get_meas_rep_avg(old_lchan, TDMA_MEAS_FIELD_RXLEV,
TDMA_MEAS_DIR_UL, TDMA_MEAS_SET_AUTO,
bts->chan_alloc_dyn_params.ul_rxlev_avg_num);
if (avg_ul_rxlev < 0) {
LOG_COND("Unknown AVG UL RxLev => using ascending order\n");
return false;
}
if (avg_ul_rxlev < bts->chan_alloc_dyn_params.ul_rxlev_thresh) {
LOG_COND("AVG UL RxLev %u < thresh %u => using ascending order\n",
avg_ul_rxlev, bts->chan_alloc_dyn_params.ul_rxlev_thresh);
return false;
}
LOG_COND("C0 Channel Load %u%% >= thresh %u%% and "
"AVG UL RxLev %u >= thresh %u => using descending order\n",
lchan_load, bts->chan_alloc_dyn_params.c0_chan_load_thresh,
avg_ul_rxlev, bts->chan_alloc_dyn_params.ul_rxlev_thresh);
#undef LOG_COND
return true;
}
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
struct gsm_lchan *lchan_select_by_chan_mode(struct gsm_bts *bts,
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
enum gsm48_chan_mode chan_mode,
enum channel_rate chan_rate,
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
enum lchan_select_reason reason,
void *ctx)
add chan_mode_to_chan_type() Move the conversion from chan_mode to lchan type out of the lchan_select_by_chan_mode() function, so that the conversion can be used by other code paths, coming up in VAMOS patches. Related: SYS#5315 OS#4940 Change-Id: I296651ebadba81f8b3db0d9bb5b5377877a43677
2021-05-19 00:32:33 +00:00
{
enum gsm_chan_t type = chan_mode_to_chan_type(chan_mode, chan_rate);
if (type == GSM_LCHAN_NONE)
return NULL;
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
return lchan_select_by_type(bts, type, reason, ctx);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
struct gsm_lchan *lchan_avail_by_type(struct gsm_bts *bts,
enum gsm_chan_t type,
enum lchan_select_reason reason,
void *ctx, bool log)
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
{
struct gsm_lchan *lchan = NULL;
enum gsm_phys_chan_config first, first_cbch, second, second_cbch;
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
struct lchan_select_ts_list ts_list;
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
bool sort_by_trx_power = false;
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
bool chan_alloc_reverse;
if (log) {
LOG_BTS(bts, DRLL, LOGL_DEBUG, "lchan_avail_by_type(type=%s, reason=%s)\n",
gsm_lchant_name(type), lchan_select_reason_name(reason));
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
switch (reason) {
case SELECT_FOR_MS_CHAN_REQ:
chan_alloc_reverse = bts->chan_alloc_chan_req_reverse;
break;
case SELECT_FOR_ASSIGNMENT:
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
if (bts->chan_alloc_assignment_dynamic) {
chan_alloc_reverse = chan_alloc_ass_dynamic_reverse(bts, ctx, log);
sort_by_trx_power = bts->chan_alloc_dyn_params.sort_by_trx_power;
} else {
chan_alloc_reverse = bts->chan_alloc_assignment_reverse;
}
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
break;
case SELECT_FOR_HANDOVER:
chan_alloc_reverse = bts->chan_alloc_handover_reverse;
break;
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
/* Allocate an array with pointers to all timeslots of a BTS */
ts_list.list = talloc_array_ptrtype(bts, ts_list.list, bts->num_trx * 8);
if (OSMO_UNLIKELY(ts_list.list == NULL))
return NULL;
/* Populate this array with the actual pointers */
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
populate_ts_list(&ts_list, bts, chan_alloc_reverse, sort_by_trx_power, log);
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
switch (type) {
case GSM_LCHAN_SDCCH:
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
if (chan_alloc_reverse) {
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
first = GSM_PCHAN_SDCCH8_SACCH8C;
first_cbch = GSM_PCHAN_SDCCH8_SACCH8C_CBCH;
second = GSM_PCHAN_CCCH_SDCCH4;
second_cbch = GSM_PCHAN_CCCH_SDCCH4_CBCH;
} else {
first = GSM_PCHAN_CCCH_SDCCH4;
first_cbch = GSM_PCHAN_CCCH_SDCCH4_CBCH;
second = GSM_PCHAN_SDCCH8_SACCH8C;
second_cbch = GSM_PCHAN_SDCCH8_SACCH8C_CBCH;
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, first, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (lchan == NULL)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, first_cbch, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (lchan == NULL)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, second, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
if (lchan == NULL)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, second_cbch, log);
Support SDCCH8 in osmo dyn ts This feature signals support to configure Osmocom Dynamic Timeslot type as SDCCH8, on top of historically supported TCH/H and TCH/F. The idea is that when unneeded, the TS is configured as PDCH, and as soon as there's need for an SDCCH and there's none available, the TS is dynamically reconfigured to SDCCH8. Once all logical channels in the dynamic TS are released and hence becomes free, the BSC will reconfigure it to PDCH. Related: SYS#5309 Depends: libosmocore.git Change-Id Ifc0ca8916bd3e93e5a60a7dd7391d2588fdb5532 Change-Id: I29ac8b90168dba3ac309daeb0b6cfdbbcb8e9172
2021-06-28 12:45:08 +00:00
/* No dedicated SDCCH available -- try fully dynamic
* TCH/F_TCH/H_SDCCH8_PDCH if BTS supports it: */
if (lchan == NULL && osmo_bts_has_feature(&bts->features, BTS_FEAT_DYN_TS_SDCCH8))
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_dyn_find(&ts_list, GSM_PCHAN_OSMO_DYN,
GSM_PCHAN_SDCCH8_SACCH8C, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
break;
case GSM_LCHAN_TCH_F:
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, GSM_PCHAN_TCH_F, log);
bsc: lchan_select: Avoid returning a TCH/H lchan when TCH/F is requested If BTS is configured to have only TCH/F_PDCH and TCH/H and a call is resolved to require TCH/F, don't return the TCH/H which have no use, but instead let it allocate the TCH/F_PDCH. 20181128185013783 DRLL <0000> lchan_select.c:159 (bts=0) lchan_select_by_type(TCH_F) 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=0,pchan=CCCH+SDCCH4,state=IN_USE) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=1,pchan=SDCCH8,state=UNUSED) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=2,pchan=TCH/H,state=UNUSED) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=3,pchan=TCH/H,state=UNUSED) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=4,pchan=TCH/H,state=UNUSED) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=5,pchan=TCH/H,state=UNUSED) is != TCH/F 20181128185013783 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=6,pchan_on_init=TCH/F_PDCH,pchan=PDCH,state=PDCH) is != TCH/F 20181128185013784 DRLL <0000> lchan_select.c:71 looking for lchan TCH/F: (bts=0,trx=0,ts=7,pchan_on_init=TCH/F_PDCH,pchan=PDCH,state=PDCH) is != TCH/F 20181128185013784 DRLL <0000> lchan_select.c:71 looking for lchan TCH/H: (bts=0,trx=0,ts=0,pchan=CCCH+SDCCH4,state=IN_USE) is != TCH/H 20181128185013784 DRLL <0000> lchan_select.c:71 looking for lchan TCH/H: (bts=0,trx=0,ts=1,pchan=SDCCH8,state=UNUSED) is != TCH/H 20181128185013784 DRLL <0000> lchan_select.c:86 looking for lchan TCH/H: (bts=0,trx=0,ts=2,pchan=TCH/H,state=UNUSED) ss=0 is available 20181128185013784 DCHAN <0010> lchan_select.c:253 lchan(0-0-2-TCH_H-0)[0x6120000066a0]{UNUSED}: (type=TCH_H) Selected Change-Id: I9a73beb0432fab16d5430e5b40d470694e09b189
2018-11-28 18:05:07 +00:00
/* If we don't have TCH/F available, try dynamic TCH/F_PDCH */
lchan-select: Avoid setting variable for no reason That variable is never used after being set. Furthermore, it is being set to the same value already stored, so there's no use in setting it and it creates confusion. Change-Id: Ib6ee28aa9a449992f5d3dea6df7dd2b7e30e73c9
2021-06-28 14:09:48 +00:00
if (!lchan)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_dyn_find(&ts_list, GSM_PCHAN_TCH_F_PDCH,
GSM_PCHAN_TCH_F, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* Try fully dynamic TCH/F_TCH/H_PDCH as TCH/F... */
lchan-select: Avoid setting variable for no reason That variable is never used after being set. Furthermore, it is being set to the same value already stored, so there's no use in setting it and it creates confusion. Change-Id: Ib6ee28aa9a449992f5d3dea6df7dd2b7e30e73c9
2021-06-28 14:09:48 +00:00
if (!lchan && bts->network->dyn_ts_allow_tch_f)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_dyn_find(&ts_list, GSM_PCHAN_OSMO_DYN,
GSM_PCHAN_TCH_F, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
break;
case GSM_LCHAN_TCH_H:
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_find(&ts_list, GSM_PCHAN_TCH_H, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
/* No dedicated TCH/x available -- try fully dynamic
* TCH/F_TCH/H_PDCH */
lchan-select: Avoid setting variable for no reason That variable is never used after being set. Furthermore, it is being set to the same value already stored, so there's no use in setting it and it creates confusion. Change-Id: Ib6ee28aa9a449992f5d3dea6df7dd2b7e30e73c9
2021-06-28 14:09:48 +00:00
if (!lchan)
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
lchan = lc_dyn_find(&ts_list, GSM_PCHAN_OSMO_DYN,
GSM_PCHAN_TCH_H, log);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
break;
default:
logging: introduce LOG_BTS and LOG_TRX and use it everywhere It's quite ugly to have manual "bts=%d" printf-statements all over the BSC code. Let's change this to use shared logging helper functions all over the place, whenever we need to log something related to one BTS or one TRX. This can also help us as the first step to later add alternative logging of BTS identities, e.g. by printing the Cell Global Identifier or LAC+CI, or even a human-readable/vty-defined 'name' of the BTS, rather than its numeric bts number. With this change in place, we can introduce such changes at a single location in the code. Change-Id: I4a7814d164384eecfb6913c31802cf2faead6e6c
2019-06-13 14:37:24 +00:00
LOG_BTS(bts, DRLL, LOGL_ERROR, "Unknown gsm_chan_t %u\n", type);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
}
lchan_select: prepare a list of timeslots once, iterate over it The lchan_avail_by_type() attempts to find an unused lchan for the given GSM_LCHAN_* value: TCH/F, TCH/H, or SDCCH. This is achieved by looking up timeslots with compatible GSM_PCHAN_* values. For instance, finding an unused SDCCH lchan may involve: * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4, * attempt to find a timeslot with pchan=GSM_PCHAN_CCCH_SDCCH4_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C, * attempt to find a timeslot with pchan=GSM_PCHAN_SDCCH8_SACCH8C_CBCH, * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (switched), * attempt to find a timeslot with pchan=GSM_PCHAN_OSMO_DYN (not switched). Each attempt involves iterating over all timeslots of each TRX, either in ascending or in descending order (see _lc_dyn_find_bts() and _lc_find_trx()). This patch simplifies the lookup logic by preparing a monolithic array of timeslot pointers once, and then using that array for each GSM_PCHAN_* lookup attempt. This change is required for the upcoming dynamic channel allocation mode, which is fa more complex than the existing ascending/descending ones. A side effect of this change is that the interference aware mode of allocation is not limited by the scope of a single TRX anymore. Interference levels are now compared within the scope of the whole BTS, so that lchans on the other TRXes may be picked if they are better according to the interference reports from the BTS. Change-Id: I7ccc56856bfd40fd7c63b7437736de60c2b516ff Related: SYS#5460
2022-06-15 18:22:33 +00:00
talloc_free(ts_list.list);
abis_rsl: prioritize emergency calls over regular calls when an emergency call arrives while all TCH are busy, the BSC should pick an arbitrary (preferably the longest lasting) call / lchan and release it in favor of the incoming emergancy call. The release of the existing call is a process that can not be done synchronously while the ChanRQD is handled sonce multiple messages are exchanged between BTS and MSC and multiple FSMs need to do their work. To be able to release one lchan while handling a ChanRQD a queue is implemented in which the incomming channel requests are collected. If an emergency call is established while all channels are busy, an arbitrary lchan is picked and freed. When freeing the lchan is done, the queue is checked again and the emergency call is put on the free lchan (TCH/H or TCH/F). Change-Id: If8651265928797dbda9f528b544931dcfa4a0b36 Related: OS#4549
2020-08-21 20:40:23 +00:00
return lchan;
}
/* Return a matching lchan from a specific BTS that is currently available. The next logical step is
* lchan_activate() on it, which would possibly cause dynamic timeslot pchan switching, taken care of by
* the lchan and timeslot FSMs. */
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
struct gsm_lchan *lchan_select_by_type(struct gsm_bts *bts,
enum gsm_chan_t type,
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
enum lchan_select_reason reason,
void *ctx)
abis_rsl: prioritize emergency calls over regular calls when an emergency call arrives while all TCH are busy, the BSC should pick an arbitrary (preferably the longest lasting) call / lchan and release it in favor of the incoming emergancy call. The release of the existing call is a process that can not be done synchronously while the ChanRQD is handled sonce multiple messages are exchanged between BTS and MSC and multiple FSMs need to do their work. To be able to release one lchan while handling a ChanRQD a queue is implemented in which the incomming channel requests are collected. If an emergency call is established while all channels are busy, an arbitrary lchan is picked and freed. When freeing the lchan is done, the queue is checked again and the emergency call is put on the free lchan (TCH/H or TCH/F). Change-Id: If8651265928797dbda9f528b544931dcfa4a0b36 Related: OS#4549
2020-08-21 20:40:23 +00:00
{
struct gsm_lchan *lchan = NULL;
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
LOG_BTS(bts, DRLL, LOGL_DEBUG, "lchan_select_by_type(type=%s, reason=%s)\n",
gsm_lchant_name(type), lchan_select_reason_name(reason));
abis_rsl: prioritize emergency calls over regular calls when an emergency call arrives while all TCH are busy, the BSC should pick an arbitrary (preferably the longest lasting) call / lchan and release it in favor of the incoming emergancy call. The release of the existing call is a process that can not be done synchronously while the ChanRQD is handled sonce multiple messages are exchanged between BTS and MSC and multiple FSMs need to do their work. To be able to release one lchan while handling a ChanRQD a queue is implemented in which the incomming channel requests are collected. If an emergency call is established while all channels are busy, an arbitrary lchan is picked and freed. When freeing the lchan is done, the queue is checked again and the emergency call is put on the free lchan (TCH/H or TCH/F). Change-Id: If8651265928797dbda9f528b544931dcfa4a0b36 Related: OS#4549
2020-08-21 20:40:23 +00:00
lchan_select: implement dynamic selection mode for assignment This change implements an additional channel allocation mode, which can be employed during a TCH channel allocation for assignment. Selection between ascending and descending order is performed depending on pre-configured parameters: * Uplink RxLev threshold and number of samples for averaging, * C0 (BCCH carrier) channel load threshold. This is useful in setups where Tx power of the RF carriers cannot be adjusted +dynamically at run-time and thus BS Power Control cannot be performed. In such setups the BCCH carrier is transmitting at relatively higher power than the other RF carriers. The key idea is to allocate channels in a smarter way, so that UEs with poor signal would get channels on carriers with high Tx power, while UEs with good signal could use carriers with lower Tx power. Change-Id: I1b7a0d706976b73cc5c30a8714b830811addfe8d Related: osmo-ttcn3-hacks.git Ia522f37c1c001b3a36f5145b8875fbb88311c2e5 Related: SYS#5460
2022-06-15 19:10:16 +00:00
lchan = lchan_avail_by_type(bts, type, reason, ctx, true);
Add new lchan_select_set_type() API helper It's an easy helper, but it helps in standarizing and finding similar places in code. Morevoer, it will be used in follow-up commit where we first use lchan_avail_by_type and finally we select it in a later stage. Change-Id: I025a40962a5e5d40543b297a0760e47618fb525c
2021-07-15 11:26:46 +00:00
if (!lchan) {
lchan_select: allow different alloc order for assignment and handover A follow-up patch implements a special channel allocation mode, which is only working for assignment (basically TCH selection for a voice call). This mode cannot be employed for initial CHANNEL REQUEST or handover due to the absence of an already established lchan. Adding this mode to the existing VTY command syntax would be confusing: channel allocator (ascending|desscending|dynamic) ^^^^^^^ so this patch extends the VTY syntax in a way that it becomes possible to configure different channel allocator modes for different cases: OsmoBSC(config-net-bts)# channel allocator mode ? set-all Set a single mode for all variants chan-req Channel allocation for CHANNEL REQUEST (RACH) assignment Channel allocation for assignment handover Channel allocation for handover The old command syntax, which is basically 'set-all', is kept for backwards compatibility, but marked as deprecated. Change-Id: I3ae73b36ee9433cc768376b56f0765e5f416162f Related: SYS#5460
2022-06-05 19:43:28 +00:00
LOG_BTS(bts, DRLL, LOGL_NOTICE, "Failed to select %s channel (%s)\n",
gsm_lchant_name(type), lchan_select_reason_name(reason));
Add new lchan_select_set_type() API helper It's an easy helper, but it helps in standarizing and finding similar places in code. Morevoer, it will be used in follow-up commit where we first use lchan_avail_by_type and finally we select it in a later stage. Change-Id: I025a40962a5e5d40543b297a0760e47618fb525c
2021-07-15 11:26:46 +00:00
return NULL;
}
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
Add new lchan_select_set_type() API helper It's an easy helper, but it helps in standarizing and finding similar places in code. Morevoer, it will be used in follow-up commit where we first use lchan_avail_by_type and finally we select it in a later stage. Change-Id: I025a40962a5e5d40543b297a0760e47618fb525c
2021-07-15 11:26:46 +00:00
lchan_select_set_type(lchan, type);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
return lchan;
}
Add new lchan_select_set_type() API helper It's an easy helper, but it helps in standarizing and finding similar places in code. Morevoer, it will be used in follow-up commit where we first use lchan_avail_by_type and finally we select it in a later stage. Change-Id: I025a40962a5e5d40543b297a0760e47618fb525c
2021-07-15 11:26:46 +00:00
/* Set available lchan to given type. Usually used on lchan obtained with
* lchan_avail_by_type. The next logical step is lchan_activate() on it, which
* would possibly cause dynamic timeslot pchan switching, taken care of by the
* lchan and timeslot FSMs. */
void lchan_select_set_type(struct gsm_lchan *lchan, enum gsm_chan_t type)
{
lchan->type = type;
LOG_LCHAN(lchan, LOGL_INFO, "Selected\n");
}