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/abis_om2000.c

3168 lines
92 KiB
C
Raw Normal View History

[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
/* Ericsson RBS 2xxx GSM O&M (OM2000) messages on the A-bis interface
* implemented based on protocol trace analysis, no formal documentation */
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* (C) 2010-2011,2016 by Harald Welte <laforge@gnumonks.org>
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +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/>.
*
*/
#include <errno.h>
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <stdint.h>
#include <arpa/inet.h>
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
#include <osmocom/core/byteswap.h>
src: use new library libosmogsm and new path to headers in libosmocore libosmogsm is a new library that is distributed in the libosmocore. Now, openbsc depends on it. This patch gets openbsc with this change. This patch also rewrites all include path to the new osmocom/[gsm|core] Signed-off-by: Pablo Neira Ayuso <pablo@gnumonks.org>
2011-03-22 15:47:59 +00:00
#include <osmocom/core/msgb.h>
#include <osmocom/gsm/tlv.h>
#include <osmocom/core/talloc.h>
#include <osmocom/core/utils.h>
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
#include <osmocom/core/fsm.h>
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
move include/openbsc to include/osmocom/bsc Change-Id: I39e7b882caa98334636d19ccd104fd83d07d5055
2017-09-04 13:15:32 +00:00
#include <osmocom/bsc/gsm_data.h>
#include <osmocom/bsc/debug.h>
#include <osmocom/bsc/abis_nm.h>
#include <osmocom/bsc/abis_rsl.h>
#include <osmocom/bsc/abis_om2000.h>
#include <osmocom/bsc/signal.h>
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
#include <osmocom/bsc/timeslot_fsm.h>
abis_om2000: fix missing signal The normal abis nm FSMs are sending S_NM_RUNNING_CHG signals that include an object class to notify other entities about state changes. This also includes paging.c, which only starts paging services when it sees NM_OC_RADIO_CARRIER becoming enabled. Change-Id: I305df5b2f962473e33e32484c42a79ff96e53e1a Fixes: I1b5b1a98115b4e9d821eb3330fc5b970a0e78a44 Related: OS#5634
2022-07-29 11:35:09 +00:00
#include <osmocom/bsc/nm_common_fsm.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>
src: port openBSC over libosmo-abis This is a big patch that ports openBSC over libosmo-abis. Sorry, the changes that are included here are all dependent of libosmo-abis, splitting them into smaller pieces would leave the repository in some intermediate state, which is not desired. The main changes are: - The directory libabis/ has been removed as it now lives in libosmo-abis. - new configuration file format for nanoBTS and HSL femto, we need to define the virtual e1_line and attach it to the OML link. - all the existing BTS drivers (nanoBTS, hsl femto, Nokia site, BS11 and rbs2000) now use the new libosmo-abis framework. - use r232 input driver available in libosmo-abis for bs11_config. - use ipa_msg_recv instead of old ipaccess_read_msg function. - delete definition of gsm_e1_subslot and input_signal_data. These structures now lives in libosmo-abis. Most of this patch are deletions of libabis/ which has been moved to libosmo-abis. This patch also modifies openBSC to use all the new definitions available in libosmocore and libosmo-abis. In order to do that, we have replaced the following: - DINP, DMI, DMIB and DMUX by their respective DL* correspondences. - SS_GLOBAL by SS_L_GLOBAL - SS_INPUT by SS_L_INPUT - S_GLOBAL_SHUTDOWN by S_L_GLOBAL_SHUTDOWN - SS_INPUT by SS_L_INPUT - S_INP_* by S_L_INP_* sub-signals - E1INP_NODE by L_E1INP_NODE vty node This patch has been tested with: - one nanoBTS - the HSL femto with the examples available under libosmo-abis - BS11 with both dahdi and misdn drivers.
2011-08-17 20:44:07 +00:00
#include <osmocom/abis/e1_input.h>
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
abis_om2000: fix missing signal The normal abis nm FSMs are sending S_NM_RUNNING_CHG signals that include an object class to notify other entities about state changes. This also includes paging.c, which only starts paging services when it sees NM_OC_RADIO_CARRIER becoming enabled. Change-Id: I305df5b2f962473e33e32484c42a79ff96e53e1a Fixes: I1b5b1a98115b4e9d821eb3330fc5b970a0e78a44 Related: OS#5634
2022-07-29 11:35:09 +00:00
static inline void abis_om2000_fsm_transc_becomes_enabled(struct gsm_bts_trx *trx)
{
nm_obj_fsm_becomes_enabled_disabled(trx->bts, trx, NM_OC_RADIO_CARRIER, true);
}
static inline void abis_om2000_fsm_transc_becomes_disabled(struct gsm_bts_trx *trx)
{
nm_obj_fsm_becomes_enabled_disabled(trx->bts, trx, NM_OC_RADIO_CARRIER, false);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* FIXME: move to libosmocore */
struct osmo_fsm_inst *osmo_fsm_inst_alloc_child_id(struct osmo_fsm *fsm,
struct osmo_fsm_inst *parent,
uint32_t parent_term_event,
const char *id)
{
struct osmo_fsm_inst *fi;
fi = osmo_fsm_inst_alloc(fsm, parent, NULL, parent->log_level,
id ? id : parent->id);
if (!fi) {
/* indicate immediate termination to caller */
osmo_fsm_inst_dispatch(parent, parent_term_event, NULL);
return NULL;
}
LOGPFSM(fi, "is child of %s\n", osmo_fsm_inst_name(parent));
fi->proc.parent = parent;
fi->proc.parent_term_event = parent_term_event;
llist_add(&fi->proc.child, &parent->proc.children);
return fi;
}
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
#define OM_ALLOC_SIZE 1024
#define OM_HEADROOM_SIZE 128
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
#define OM2K_TIMEOUT 10
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
#define TRX_LAPD_TIMEOUT 5
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
#define TRX_FSM_TIMEOUT 60
#define BTS_FSM_TIMEOUT 60
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
/* use following functions from abis_nm.c:
* om2k_msgb_alloc()
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
* abis_om2k_sendmsg()
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
*/
struct abis_om2k_hdr {
struct abis_om_hdr om;
uint16_t msg_type;
struct abis_om2k_mo mo;
uint8_t data[0];
} __attribute__ ((packed));
enum abis_om2k_msgtype {
OM2K_MSGT_ABORT_SP_CMD = 0x0000,
OM2K_MSGT_ABORT_SP_COMPL = 0x0002,
OM2K_MSGT_ALARM_REP_ACK = 0x0004,
OM2K_MSGT_ALARM_REP_NACK = 0x0005,
OM2K_MSGT_ALARM_REP = 0x0006,
OM2K_MSGT_ALARM_STATUS_REQ = 0x0008,
OM2K_MSGT_ALARM_STATUS_REQ_ACK = 0x000a,
OM2K_MSGT_ALARM_STATUS_REQ_REJ = 0x000b,
OM2K_MSGT_ALARM_STATUS_RES_ACK = 0x000c,
OM2K_MSGT_ALARM_STATUS_RES_NACK = 0x000d,
OM2K_MSGT_ALARM_STATUS_RES = 0x000e,
OM2K_MSGT_CAL_TIME_RESP = 0x0010,
OM2K_MSGT_CAL_TIME_REJ = 0x0011,
OM2K_MSGT_CAL_TIME_REQ = 0x0012,
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
OM2K_MSGT_CON_CONF_REQ = 0x0014,
OM2K_MSGT_CON_CONF_REQ_ACK = 0x0016,
OM2K_MSGT_CON_CONF_REQ_REJ = 0x0017,
OM2K_MSGT_CON_CONF_RES_ACK = 0x0018,
OM2K_MSGT_CON_CONF_RES_NACK = 0x0019,
OM2K_MSGT_CON_CONF_RES = 0x001a,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_MSGT_CONNECT_CMD = 0x001c,
OM2K_MSGT_CONNECT_COMPL = 0x001e,
OM2K_MSGT_CONNECT_REJ = 0x001f,
OM2000: Add enable/disable commands
2011-02-12 17:48:53 +00:00
OM2K_MSGT_DISABLE_REQ = 0x0028,
OM2K_MSGT_DISABLE_REQ_ACK = 0x002a,
OM2K_MSGT_DISABLE_REQ_REJ = 0x002b,
OM2K_MSGT_DISABLE_RES_ACK = 0x002c,
OM2K_MSGT_DISABLE_RES_NACK = 0x002d,
OM2K_MSGT_DISABLE_RES = 0x002e,
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
OM2K_MSGT_DISCONNECT_CMD = 0x0030,
OM2K_MSGT_DISCONNECT_COMPL = 0x0032,
OM2K_MSGT_DISCONNECT_REJ = 0x0033,
OM2000: Add enable/disable commands
2011-02-12 17:48:53 +00:00
OM2K_MSGT_ENABLE_REQ = 0x0034,
OM2K_MSGT_ENABLE_REQ_ACK = 0x0036,
OM2K_MSGT_ENABLE_REQ_REJ = 0x0037,
OM2K_MSGT_ENABLE_RES_ACK = 0x0038,
OM2K_MSGT_ENABLE_RES_NACK = 0x0039,
OM2K_MSGT_ENABLE_RES = 0x003a,
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_MSGT_FAULT_REP_ACK = 0x0040,
OM2K_MSGT_FAULT_REP_NACK = 0x0041,
OM2K_MSGT_FAULT_REP = 0x0042,
OM2K_MSGT_IS_CONF_REQ = 0x0060,
OM2K_MSGT_IS_CONF_REQ_ACK = 0x0062,
OM2K_MSGT_IS_CONF_REQ_REJ = 0x0063,
OM2K_MSGT_IS_CONF_RES_ACK = 0x0064,
OM2K_MSGT_IS_CONF_RES_NACK = 0x0065,
OM2K_MSGT_IS_CONF_RES = 0x0066,
OM2K_MSGT_OP_INFO = 0x0074,
OM2K_MSGT_OP_INFO_ACK = 0x0076,
OM2K_MSGT_OP_INFO_REJ = 0x0077,
OM2K_MSGT_RESET_CMD = 0x0078,
OM2K_MSGT_RESET_COMPL = 0x007a,
OM2K_MSGT_RESET_REJ = 0x007b,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_MSGT_RX_CONF_REQ = 0x007c,
OM2K_MSGT_RX_CONF_REQ_ACK = 0x007e,
OM2K_MSGT_RX_CONF_REQ_REJ = 0x007f,
OM2K_MSGT_RX_CONF_RES_ACK = 0x0080,
OM2K_MSGT_RX_CONF_RES_NACK = 0x0081,
OM2K_MSGT_RX_CONF_RES = 0x0082,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_MSGT_START_REQ = 0x0084,
OM2K_MSGT_START_REQ_ACK = 0x0086,
OM2K_MSGT_START_REQ_REJ = 0x0087,
OM2K_MSGT_START_RES_ACK = 0x0088,
OM2K_MSGT_START_RES_NACK = 0x0089,
OM2K_MSGT_START_RES = 0x008a,
OM2000: Introduce VTY commands to initiate OM2000 procedures
2011-02-12 13:42:59 +00:00
OM2K_MSGT_STATUS_REQ = 0x008c,
OM2K_MSGT_STATUS_RESP = 0x008e,
OM2K_MSGT_STATUS_REJ = 0x008f,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2000: Add support for sending the TEST REQUEST message
2011-02-12 14:07:30 +00:00
OM2K_MSGT_TEST_REQ = 0x0094,
OM2K_MSGT_TEST_REQ_ACK = 0x0096,
OM2K_MSGT_TEST_REQ_REJ = 0x0097,
OM2K_MSGT_TEST_RES_ACK = 0x0098,
OM2K_MSGT_TEST_RES_NACK = 0x0099,
OM2K_MSGT_TEST_RES = 0x009a,
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
OM2K_MSGT_TF_CONF_REQ = 0x00a0,
OM2K_MSGT_TF_CONF_REQ_ACK = 0x00a2,
OM2K_MSGT_TF_CONF_REQ_REJ = 0x00a3,
OM2K_MSGT_TF_CONF_RES_ACK = 0x00a4,
OM2K_MSGT_TF_CONF_RES_NACK = 0x00a5,
OM2K_MSGT_TF_CONF_RES = 0x00a6,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_MSGT_TS_CONF_REQ = 0x00a8,
OM2K_MSGT_TS_CONF_REQ_ACK = 0x00aa,
OM2K_MSGT_TS_CONF_REQ_REJ = 0x00ab,
OM2K_MSGT_TS_CONF_RES_ACK = 0x00ac,
OM2K_MSGT_TS_CONF_RES_NACK = 0x00ad,
OM2K_MSGT_TS_CONF_RES = 0x00ae,
OM2K_MSGT_TX_CONF_REQ = 0x00b0,
OM2K_MSGT_TX_CONF_REQ_ACK = 0x00b2,
OM2K_MSGT_TX_CONF_REQ_REJ = 0x00b3,
OM2K_MSGT_TX_CONF_RES_ACK = 0x00b4,
OM2K_MSGT_TX_CONF_RES_NACK = 0x00b5,
OM2K_MSGT_TX_CONF_RES = 0x00b6,
om2k: Acknowledge the HW Infos Reports Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ida92c9ca59a608854737ba9e330a5394cceff0fe
2020-05-05 18:16:16 +00:00
OM2K_MSGT_CAPA_HW_INFOS_REP_ACK = 0x00e4,
OM2K_MSGT_CAPA_HW_INFOS_REP_NACK = 0x00e5,
OM2K_MSGT_CAPA_HW_INFOS_REP = 0x00e6,
om2000: Add support for querying RBS capabilities Change-Id: Id1fbaf41286f3651ce8f210eb8da05fb51179c96
2016-10-15 19:24:57 +00:00
OM2K_MSGT_CAPA_REQ = 0x00e8,
OM2K_MSGT_CAPA_REQ_ACK = 0x00ea,
OM2K_MSGT_CAPA_REQ_REJ = 0x00eb,
OM2K_MSGT_CAPA_RES = 0x00ee,
OM2K_MSGT_CAPA_RES_ACK = 0x00ec,
OM2K_MSGT_CAPA_RES_NACK = 0x00ed,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_MSGT_NEGOT_REQ_ACK = 0x0104,
OM2K_MSGT_NEGOT_REQ_NACK = 0x0105,
OM2K_MSGT_NEGOT_REQ = 0x0106,
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
OM2K_MSGT_BTS_INITIATED_REQ_ACK = 0x0108,
OM2K_MSGT_BTS_INITIATED_REQ_NACK = 0x0109,
OM2K_MSGT_BTS_INITIATED_REQ = 0x010a,
OM2K_MSGT_RADIO_CHAN_REL_CMD = 0x010c,
OM2K_MSGT_RADIO_CHAN_REL_COMPL = 0x010e,
OM2K_MSGT_RADIO_CHAN_REL_REJ = 0x010f,
OM2K_MSGT_FEATURE_CTRL_CMD = 0x0118,
OM2K_MSGT_FEATURE_CTRL_COMPL = 0x011a,
OM2K_MSGT_FEATURE_CTRL_REJ = 0x011b,
om2k: Rename MCTR config request constants for consistency Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I776b0016837e018500ef69acb2f30a274008818e
2020-05-05 18:09:36 +00:00
OM2K_MSGT_MCTR_CONF_REQ = 0x012c,
OM2K_MSGT_MCTR_CONF_REQ_ACK = 0x012e,
OM2K_MSGT_MCTR_CONF_REQ_REJ = 0x012f,
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
om2k: Rename MCTR config request constants for consistency Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I776b0016837e018500ef69acb2f30a274008818e
2020-05-05 18:09:36 +00:00
OM2K_MSGT_MCTR_CONF_RES_ACK = 0x0130,
OM2K_MSGT_MCTR_CONF_RES_NACK = 0x0131,
OM2K_MSGT_MCTR_CONF_RES = 0x0132,
om2k: Properly name message 0x0136, found to be MCTR Statistics Report Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I4b4389fd2e7bf0b3078b8ff60f6ea109114a4475
2020-05-06 05:25:25 +00:00
OM2K_MSGT_MCTR_STATS_REP_ACK = 0x0134,
OM2K_MSGT_MCTR_STATS_REP_NACK = 0x0135,
OM2K_MSGT_MCTR_STATS_REP = 0x0136,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
};
enum abis_om2k_dei {
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
OM2K_DEI_ACCORDANCE_IND = 0x00,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_BCC = 0x06,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_BS_AG_BKS_RES = 0x07,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_BSIC = 0x09,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_BA_PA_MFRMS = 0x0a,
OM2K_DEI_CBCH_INDICATOR = 0x0b,
OM2K_DEI_CCCH_OPTIONS = 0x0c,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_DEI_CAL_TIME = 0x0d,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_COMBINATION = 0x0f,
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
OM2K_DEI_CON_CONN_LIST = 0x10,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_DRX_DEV_MAX = 0x12,
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
OM2K_DEI_END_LIST_NR = 0x13,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_EXT_COND_MAP_1 = 0x14,
OM2K_DEI_EXT_COND_MAP_2 = 0x15,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_FILLING_MARKER = 0x1c,
OM2K_DEI_FN_OFFSET = 0x1d,
OM2K_DEI_FREQ_LIST = 0x1e,
OM2K_DEI_FREQ_SPEC_RX = 0x1f,
OM2K_DEI_FREQ_SPEC_TX = 0x20,
OM2K_DEI_HSN = 0x21,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_ICM_INDICATOR = 0x22,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_INT_FAULT_MAP_1A = 0x23,
OM2K_DEI_INT_FAULT_MAP_1B = 0x24,
OM2K_DEI_INT_FAULT_MAP_2A = 0x25,
OM2K_DEI_INT_FAULT_MAP_2A_EXT = 0x26,
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
OM2K_DEI_IS_CONN_LIST = 0x27,
OM2K_DEI_LIST_NR = 0x28,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_LOCAL_ACCESS = 0x2a,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_MAIO = 0x2b,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_MO_STATE = 0x2c,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_NY1 = 0x2d,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2K_DEI_OP_INFO = 0x2e,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_POWER = 0x2f,
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
OM2K_DEI_REASON_CODE = 0x32,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_RX_DIVERSITY = 0x33,
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
OM2K_DEI_REPL_UNIT_MAP = 0x34,
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
OM2K_DEI_RESULT_CODE = 0x35,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_T3105 = 0x38,
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
OM2K_DEI_TF_MODE = 0x3a,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_TS_NR = 0x3c,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_TSC = 0x3d,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_BTS_VERSION = 0x40,
OM2K_DEI_OML_IWD_VERSION = 0x41,
OM2K_DEI_RSL_IWD_VERSION = 0x42,
OM2K_DEI_OML_FUNC_MAP_1 = 0x43,
OM2K_DEI_OML_FUNC_MAP_2 = 0x44,
OM2K_DEI_RSL_FUNC_MAP_1 = 0x45,
OM2K_DEI_RSL_FUNC_MAP_2 = 0x46,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
OM2K_DEI_EXT_RANGE = 0x47,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_REQ_IND = 0x48,
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
OM2K_DEI_REPL_UNIT_MAP_EXT = 0x50,
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
OM2K_DEI_ICM_BOUND_PARAMS = 0x74,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_LSC = 0x79,
OM2K_DEI_LSC_FILT_TIME = 0x7a,
OM2K_DEI_CALL_SUPV_TIME = 0x7b,
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
OM2K_DEI_ICM_CHAN_RATE = 0x7e,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_HW_INFO_SIG = 0x84,
OM2000: Fix configuration of Timing Function (missing timing src)
2011-03-06 18:01:16 +00:00
OM2K_DEI_TF_SYNC_SRC = 0x86,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_TTA = 0x87,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_CAPA_SIG = 0x8a,
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
OM2K_DEI_NEGOT_REC1 = 0x90,
OM2K_DEI_NEGOT_REC2 = 0x91,
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
OM2K_DEI_ENCR_ALG = 0x92,
OM2K_DEI_INTERF_REJ_COMB = 0x94,
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
OM2K_DEI_FS_OFFSET = 0x98,
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
OM2K_DEI_EXT_COND_MAP_2_EXT = 0x9c,
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
OM2K_DEI_TSS_MO_STATE = 0x9d,
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
OM2K_DEI_CONFIG_TYPE = 0x9e,
OM2K_DEI_JITTER_SIZE = 0x9f,
OM2K_DEI_PACKING_ALGO = 0xa0,
OM2K_DEI_TRXC_LIST = 0xa8,
OM2K_DEI_MAX_ALLOWED_POWER = 0xa9,
OM2K_DEI_MAX_ALLOWED_NUM_TRXCS = 0xaa,
OM2K_DEI_MCTR_FEAT_STATUS_BMAP = 0xab,
Improve parsing of om2000 fault reports Fault Reports are commonly oberved with a TLV id 0xd2 as are reports with up to 20 TLVs. Let's not have these cause logging at level ERROR. Closes: OS#5593 Change-Id: Ibe0b38835362c59d1576a206b2f64cea4427295f
2022-06-28 02:22:06 +00:00
OM2K_DEI_SEEN_UNKNOWN_D2 = 0xd2,
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
};
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
enum abis_om2k_mostate {
OM2K_MOSTATE_RESET = 0x00,
OM2K_MOSTATE_STARTED = 0x01,
OM2K_MOSTATE_ENABLED = 0x02,
OM2K_MOSTATE_DISABLED = 0x03,
};
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
const struct tlv_definition om2k_att_tlvdef = {
.def = {
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
[OM2K_DEI_ACCORDANCE_IND] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_BCC] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_BS_AG_BKS_RES] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_BSIC] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_BA_PA_MFRMS] = { TLV_TYPE_TV },
[OM2K_DEI_CBCH_INDICATOR] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_INT_FAULT_MAP_1A] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_INT_FAULT_MAP_1B] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_INT_FAULT_MAP_2A] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_INT_FAULT_MAP_2A_EXT]={ TLV_TYPE_FIXED, 6 },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_CCCH_OPTIONS] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_CAL_TIME] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_COMBINATION] = { TLV_TYPE_TV },
[OM2K_DEI_CON_CONN_LIST] = { TLV_TYPE_TLV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_DRX_DEV_MAX] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_END_LIST_NR] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_EXT_COND_MAP_1] = { TLV_TYPE_FIXED, 2 },
[OM2K_DEI_EXT_COND_MAP_2] = { TLV_TYPE_FIXED, 2 },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_FILLING_MARKER] = { TLV_TYPE_TV },
[OM2K_DEI_FN_OFFSET] = { TLV_TYPE_FIXED, 2 },
[OM2K_DEI_FREQ_LIST] = { TLV_TYPE_TLV },
[OM2K_DEI_FREQ_SPEC_RX] = { TLV_TYPE_FIXED, 2 },
[OM2K_DEI_FREQ_SPEC_TX] = { TLV_TYPE_FIXED, 2 },
[OM2K_DEI_HSN] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_ICM_INDICATOR] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_IS_CONN_LIST] = { TLV_TYPE_TLV },
[OM2K_DEI_LIST_NR] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_LOCAL_ACCESS] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_MAIO] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_MO_STATE] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_NY1] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_OP_INFO] = { TLV_TYPE_TV },
[OM2K_DEI_POWER] = { TLV_TYPE_TV },
[OM2K_DEI_REASON_CODE] = { TLV_TYPE_TV },
[OM2K_DEI_RX_DIVERSITY] = { TLV_TYPE_TV },
[OM2K_DEI_RESULT_CODE] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_T3105] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_TF_MODE] = { TLV_TYPE_TV },
[OM2K_DEI_TS_NR] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_TSC] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_BTS_VERSION] = { TLV_TYPE_FIXED, 12 },
[OM2K_DEI_OML_IWD_VERSION] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_RSL_IWD_VERSION] = { TLV_TYPE_FIXED, 6 },
[OM2K_DEI_OML_FUNC_MAP_1] = { TLV_TYPE_TLV },
[OM2K_DEI_OML_FUNC_MAP_2] = { TLV_TYPE_TLV },
[OM2K_DEI_RSL_FUNC_MAP_1] = { TLV_TYPE_TLV },
[OM2K_DEI_RSL_FUNC_MAP_2] = { TLV_TYPE_TLV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_EXT_RANGE] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_REQ_IND] = { TLV_TYPE_TV },
[OM2K_DEI_REPL_UNIT_MAP] = { TLV_TYPE_FIXED, 6 },
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
[OM2K_DEI_REPL_UNIT_MAP_EXT] = {TLV_TYPE_FIXED, 6},
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
[OM2K_DEI_ICM_BOUND_PARAMS] = { TLV_TYPE_FIXED, 5 },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_LSC] = { TLV_TYPE_TV },
[OM2K_DEI_LSC_FILT_TIME] = { TLV_TYPE_TV },
[OM2K_DEI_CALL_SUPV_TIME] = { TLV_TYPE_TV },
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
[OM2K_DEI_ICM_CHAN_RATE] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_HW_INFO_SIG] = { TLV_TYPE_FIXED, 2 },
OM2000: Fix configuration of Timing Function (missing timing src)
2011-03-06 18:01:16 +00:00
[OM2K_DEI_TF_SYNC_SRC] = { TLV_TYPE_TV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_TTA] = { TLV_TYPE_TV },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_CAPA_SIG] = { TLV_TYPE_FIXED, 2 },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_NEGOT_REC1] = { TLV_TYPE_TLV },
[OM2K_DEI_NEGOT_REC2] = { TLV_TYPE_TLV },
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
[OM2K_DEI_ENCR_ALG] = { TLV_TYPE_TV },
[OM2K_DEI_INTERF_REJ_COMB] = { TLV_TYPE_TV },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
[OM2K_DEI_FS_OFFSET] = { TLV_TYPE_FIXED, 5 },
OM2000: TLV parser definitions for more DEI's
2011-03-05 19:13:52 +00:00
[OM2K_DEI_EXT_COND_MAP_2_EXT] = { TLV_TYPE_FIXED, 4 },
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
[OM2K_DEI_TSS_MO_STATE] = { TLV_TYPE_FIXED, 4 },
Improve parsing of om2000 fault reports Fault Reports are commonly oberved with a TLV id 0xd2 as are reports with up to 20 TLVs. Let's not have these cause logging at level ERROR. Closes: OS#5593 Change-Id: Ibe0b38835362c59d1576a206b2f64cea4427295f
2022-06-28 02:22:06 +00:00
[OM2K_DEI_SEEN_UNKNOWN_D2] = { TLV_TYPE_FIXED, 6 },
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
},
};
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
static const struct value_string om2k_msgcode_vals[] = {
{ 0x0000, "Abort SP Command" },
{ 0x0002, "Abort SP Complete" },
{ 0x0004, "Alarm Report ACK" },
{ 0x0005, "Alarm Report NACK" },
{ 0x0006, "Alarm Report" },
{ 0x0008, "Alarm Status Request" },
{ 0x000a, "Alarm Status Request Accept" },
{ 0x000b, "Alarm Status Request Reject" },
{ 0x000c, "Alarm Status Result ACK" },
{ 0x000d, "Alarm Status Result NACK" },
{ 0x000e, "Alarm Status Result" },
{ 0x0010, "Calendar Time Response" },
{ 0x0011, "Calendar Time Reject" },
{ 0x0012, "Calendar Time Request" },
{ 0x0014, "CON Configuration Request" },
{ 0x0016, "CON Configuration Request Accept" },
{ 0x0017, "CON Configuration Request Reject" },
{ 0x0018, "CON Configuration Result ACK" },
{ 0x0019, "CON Configuration Result NACK" },
{ 0x001a, "CON Configuration Result" },
{ 0x001c, "Connect Command" },
{ 0x001e, "Connect Complete" },
OM2000: Handle (Dis)Connect Reject in NACK handler
2011-03-05 16:58:13 +00:00
{ 0x001f, "Connect Reject" },
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0x0028, "Disable Request" },
{ 0x002a, "Disable Request Accept" },
{ 0x002b, "Disable Request Reject" },
{ 0x002c, "Disable Result ACK" },
{ 0x002d, "Disable Result NACK" },
{ 0x002e, "Disable Result" },
{ 0x0030, "Disconnect Command" },
{ 0x0032, "Disconnect Complete" },
{ 0x0033, "Disconnect Reject" },
{ 0x0034, "Enable Request" },
{ 0x0036, "Enable Request Accept" },
{ 0x0037, "Enable Request Reject" },
{ 0x0038, "Enable Result ACK" },
{ 0x0039, "Enable Result NACK" },
{ 0x003a, "Enable Result" },
{ 0x003c, "Escape Downlink Normal" },
{ 0x003d, "Escape Downlink NACK" },
{ 0x003e, "Escape Uplink Normal" },
{ 0x003f, "Escape Uplink NACK" },
{ 0x0040, "Fault Report ACK" },
{ 0x0041, "Fault Report NACK" },
{ 0x0042, "Fault Report" },
{ 0x0044, "File Package End Command" },
{ 0x0046, "File Package End Result" },
{ 0x0047, "File Package End Reject" },
{ 0x0048, "File Relation Request" },
{ 0x004a, "File Relation Response" },
{ 0x004b, "File Relation Request Reject" },
{ 0x004c, "File Segment Transfer" },
{ 0x004e, "File Segment Transfer Complete" },
{ 0x004f, "File Segment Transfer Reject" },
{ 0x0050, "HW Information Request" },
{ 0x0052, "HW Information Request Accept" },
{ 0x0053, "HW Information Request Reject" },
{ 0x0054, "HW Information Result ACK" },
{ 0x0055, "HW Information Result NACK" },
{ 0x0056, "HW Information Result" },
{ 0x0060, "IS Configuration Request" },
{ 0x0062, "IS Configuration Request Accept" },
{ 0x0063, "IS Configuration Request Reject" },
{ 0x0064, "IS Configuration Result ACK" },
{ 0x0065, "IS Configuration Result NACK" },
{ 0x0066, "IS Configuration Result" },
{ 0x0068, "Load Data End" },
{ 0x006a, "Load Data End Result" },
{ 0x006b, "Load Data End Reject" },
{ 0x006c, "Load Data Init" },
{ 0x006e, "Load Data Init Accept" },
{ 0x006f, "Load Data Init Reject" },
{ 0x0070, "Loop Control Command" },
{ 0x0072, "Loop Control Complete" },
{ 0x0073, "Loop Control Reject" },
{ 0x0074, "Operational Information" },
{ 0x0076, "Operational Information Accept" },
{ 0x0077, "Operational Information Reject" },
{ 0x0078, "Reset Command" },
{ 0x007a, "Reset Complete" },
{ 0x007b, "Reset Reject" },
{ 0x007c, "RX Configuration Request" },
{ 0x007e, "RX Configuration Request Accept" },
{ 0x007f, "RX Configuration Request Reject" },
{ 0x0080, "RX Configuration Result ACK" },
{ 0x0081, "RX Configuration Result NACK" },
{ 0x0082, "RX Configuration Result" },
{ 0x0084, "Start Request" },
{ 0x0086, "Start Request Accept" },
{ 0x0087, "Start Request Reject" },
{ 0x0088, "Start Result ACK" },
{ 0x0089, "Start Result NACK" },
{ 0x008a, "Start Result" },
{ 0x008c, "Status Request" },
{ 0x008e, "Status Response" },
{ 0x008f, "Status Reject" },
{ 0x0094, "Test Request" },
{ 0x0096, "Test Request Accept" },
{ 0x0097, "Test Request Reject" },
{ 0x0098, "Test Result ACK" },
{ 0x0099, "Test Result NACK" },
{ 0x009a, "Test Result" },
{ 0x00a0, "TF Configuration Request" },
{ 0x00a2, "TF Configuration Request Accept" },
{ 0x00a3, "TF Configuration Request Reject" },
{ 0x00a4, "TF Configuration Result ACK" },
{ 0x00a5, "TF Configuration Result NACK" },
{ 0x00a6, "TF Configuration Result" },
{ 0x00a8, "TS Configuration Request" },
{ 0x00aa, "TS Configuration Request Accept" },
{ 0x00ab, "TS Configuration Request Reject" },
{ 0x00ac, "TS Configuration Result ACK" },
{ 0x00ad, "TS Configuration Result NACK" },
{ 0x00ae, "TS Configuration Result" },
{ 0x00b0, "TX Configuration Request" },
{ 0x00b2, "TX Configuration Request Accept" },
{ 0x00b3, "TX Configuration Request Reject" },
{ 0x00b4, "TX Configuration Result ACK" },
{ 0x00b5, "TX Configuration Result NACK" },
{ 0x00b6, "TX Configuration Result" },
{ 0x00bc, "DIP Alarm Report ACK" },
{ 0x00bd, "DIP Alarm Report NACK" },
{ 0x00be, "DIP Alarm Report" },
{ 0x00c0, "DIP Alarm Status Request" },
{ 0x00c2, "DIP Alarm Status Response" },
{ 0x00c3, "DIP Alarm Status Reject" },
{ 0x00c4, "DIP Quality Report I ACK" },
{ 0x00c5, "DIP Quality Report I NACK" },
{ 0x00c6, "DIP Quality Report I" },
{ 0x00c8, "DIP Quality Report II ACK" },
{ 0x00c9, "DIP Quality Report II NACK" },
{ 0x00ca, "DIP Quality Report II" },
{ 0x00dc, "DP Configuration Request" },
{ 0x00de, "DP Configuration Request Accept" },
{ 0x00df, "DP Configuration Request Reject" },
{ 0x00e0, "DP Configuration Result ACK" },
{ 0x00e1, "DP Configuration Result NACK" },
{ 0x00e2, "DP Configuration Result" },
{ 0x00e4, "Capabilities HW Info Report ACK" },
{ 0x00e5, "Capabilities HW Info Report NACK" },
{ 0x00e6, "Capabilities HW Info Report" },
{ 0x00e8, "Capabilities Request" },
{ 0x00ea, "Capabilities Request Accept" },
{ 0x00eb, "Capabilities Request Reject" },
{ 0x00ec, "Capabilities Result ACK" },
{ 0x00ed, "Capabilities Result NACK" },
{ 0x00ee, "Capabilities Result" },
{ 0x00f0, "FM Configuration Request" },
{ 0x00f2, "FM Configuration Request Accept" },
{ 0x00f3, "FM Configuration Request Reject" },
{ 0x00f4, "FM Configuration Result ACK" },
{ 0x00f5, "FM Configuration Result NACK" },
{ 0x00f6, "FM Configuration Result" },
{ 0x00f8, "FM Report Request" },
{ 0x00fa, "FM Report Response" },
{ 0x00fb, "FM Report Reject" },
{ 0x00fc, "FM Start Command" },
{ 0x00fe, "FM Start Complete" },
{ 0x00ff, "FM Start Reject" },
{ 0x0100, "FM Stop Command" },
{ 0x0102, "FM Stop Complete" },
{ 0x0103, "FM Stop Reject" },
{ 0x0104, "Negotiation Request ACK" },
{ 0x0105, "Negotiation Request NACK" },
{ 0x0106, "Negotiation Request" },
{ 0x0108, "BTS Initiated Request ACK" },
{ 0x0109, "BTS Initiated Request NACK" },
{ 0x010a, "BTS Initiated Request" },
{ 0x010c, "Radio Channels Release Command" },
{ 0x010e, "Radio Channels Release Complete" },
{ 0x010f, "Radio Channels Release Reject" },
{ 0x0118, "Feature Control Command" },
{ 0x011a, "Feature Control Complete" },
{ 0x011b, "Feature Control Reject" },
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
{ 0x012c, "MCTR Configuration Request" },
{ 0x012e, "MCTR Configuration Request Accept" },
{ 0x012f, "MCTR Configuration Request Reject" },
{ 0x0130, "MCTR Configuration Result ACK" },
{ 0x0131, "MCTR Configuration Result NACK" },
{ 0x0132, "MCTR Configuration Result" },
om2k: Properly name message 0x0136, found to be MCTR Statistics Report Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I4b4389fd2e7bf0b3078b8ff60f6ea109114a4475
2020-05-06 05:25:25 +00:00
{ 0x0134, "MCTR Statistics report ACK" },
{ 0x0135, "MCTR Statistics report NACK" },
{ 0x0136, "MCTR Statistics report" },
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0, NULL }
};
/* TS 12.21 Section 9.4: Attributes */
static const struct value_string om2k_attr_vals[] = {
{ 0x00, "Accordance indication" },
{ 0x01, "Alarm Id" },
{ 0x02, "Alarm Data" },
{ 0x03, "Alarm Severity" },
{ 0x04, "Alarm Status" },
{ 0x05, "Alarm Status Type" },
{ 0x06, "BCC" },
{ 0x07, "BS_AG_BKS_RES" },
{ 0x09, "BSIC" },
{ 0x0a, "BA_PA_MFRMS" },
{ 0x0b, "CBCH Indicator" },
{ 0x0c, "CCCH Options" },
{ 0x0d, "Calendar Time" },
{ 0x0f, "Channel Combination" },
{ 0x10, "CON Connection List" },
{ 0x11, "Data End Indication" },
{ 0x12, "DRX_DEV_MAX" },
{ 0x13, "End List Number" },
{ 0x14, "External Condition Map Class 1" },
{ 0x15, "External Condition Map Class 2" },
{ 0x16, "File Relation Indication" },
{ 0x17, "File Revision" },
{ 0x18, "File Segment Data" },
{ 0x19, "File Segment Length" },
{ 0x1a, "File Segment Sequence Number" },
{ 0x1b, "File Size" },
{ 0x1c, "Filling Marker" },
{ 0x1d, "FN Offset" },
{ 0x1e, "Frequency List" },
{ 0x1f, "Frequency Specifier RX" },
{ 0x20, "Frequency Specifier TX" },
{ 0x21, "HSN" },
{ 0x22, "ICM Indicator" },
{ 0x23, "Internal Fault Map Class 1A" },
{ 0x24, "Internal Fault Map Class 1B" },
{ 0x25, "Internal Fault Map Class 2A" },
{ 0x26, "Internal Fault Map Class 2A Extension" },
{ 0x27, "IS Connection List" },
{ 0x28, "List Number" },
{ 0x29, "File Package State Indication" },
{ 0x2a, "Local Access State" },
{ 0x2b, "MAIO" },
{ 0x2c, "MO State" },
{ 0x2d, "Ny1" },
{ 0x2e, "Operational Information" },
{ 0x2f, "Power" },
{ 0x30, "RU Position Data" },
{ 0x31, "Protocol Error" },
{ 0x32, "Reason Code" },
{ 0x33, "Receiver Diversity" },
{ 0x34, "Replacement Unit Map" },
{ 0x35, "Result Code" },
{ 0x36, "RU Revision Data" },
{ 0x38, "T3105" },
{ 0x39, "Test Loop Setting" },
{ 0x3a, "TF Mode" },
{ 0x3b, "TF Compensation Value" },
{ 0x3c, "Time Slot Number" },
{ 0x3d, "TSC" },
{ 0x3e, "RU Logical Id" },
{ 0x3f, "RU Serial Number Data" },
{ 0x40, "BTS Version" },
{ 0x41, "OML IWD Version" },
{ 0x42, "RWL IWD Version" },
{ 0x43, "OML Function Map 1" },
{ 0x44, "OML Function Map 2" },
{ 0x45, "RSL Function Map 1" },
{ 0x46, "RSL Function Map 2" },
{ 0x47, "Extended Range Indicator" },
{ 0x48, "Request Indicators" },
{ 0x49, "DIP Alarm Condition Map" },
{ 0x4a, "ES Incoming" },
{ 0x4b, "ES Outgoing" },
{ 0x4e, "SES Incoming" },
{ 0x4f, "SES Outgoing" },
{ 0x50, "Replacement Unit Map Extension" },
{ 0x52, "UAS Incoming" },
{ 0x53, "UAS Outgoing" },
{ 0x58, "DF Incoming" },
{ 0x5a, "DF Outgoing" },
{ 0x5c, "SF" },
{ 0x60, "S Bits Setting" },
{ 0x61, "CRC-4 Use Option" },
{ 0x62, "T Parameter" },
{ 0x63, "N Parameter" },
{ 0x64, "N1 Parameter" },
{ 0x65, "N3 Parameter" },
{ 0x66, "N4 Parameter" },
{ 0x67, "P Parameter" },
{ 0x68, "Q Parameter" },
{ 0x69, "BI_Q1" },
{ 0x6a, "BI_Q2" },
{ 0x74, "ICM Boundary Parameters" },
{ 0x77, "AFT" },
{ 0x78, "AFT RAI" },
{ 0x79, "Link Supervision Control" },
{ 0x7a, "Link Supervision Filtering Time" },
{ 0x7b, "Call Supervision Time" },
{ 0x7c, "Interval Length UAS Incoming" },
{ 0x7d, "Interval Length UAS Outgoing" },
{ 0x7e, "ICM Channel Rate" },
{ 0x7f, "Attribute Identifier" },
{ 0x80, "FM Frequency List" },
{ 0x81, "FM Frequency Report" },
{ 0x82, "FM Percentile" },
{ 0x83, "FM Clear Indication" },
{ 0x84, "HW Info Signature" },
{ 0x85, "MO Record" },
{ 0x86, "TF Synchronisation Source" },
{ 0x87, "TTA" },
{ 0x88, "End Segment Number" },
{ 0x89, "Segment Number" },
{ 0x8a, "Capabilities Signature" },
{ 0x8c, "File Relation List" },
{ 0x90, "Negotiation Record I" },
{ 0x91, "Negotiation Record II" },
{ 0x92, "Encryption Algorithm" },
{ 0x94, "Interference Rejection Combining" },
{ 0x95, "Dedication Information" },
{ 0x97, "Feature Code" },
{ 0x98, "FS Offset" },
{ 0x99, "ESB Timeslot" },
{ 0x9a, "Master TG Instance" },
{ 0x9b, "Master TX Chain Delay" },
{ 0x9c, "External Condition Class 2 Extension" },
{ 0x9d, "TSs MO State" },
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
{ 0x9e, "Configuration Type" },
{ 0x9f, "Jitter Size" },
{ 0xa0, "Packing Algorithm" },
{ 0xa8, "TRXC List" },
{ 0xa9, "Maximum Allowed Power" },
{ 0xaa, "Maximum Allowed Number of TRXCs" },
{ 0xab, "MCTR Feature Status Bitmap" },
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0, NULL }
};
OM2000: Introduce VTY commands to initiate OM2000 procedures
2011-02-12 13:42:59 +00:00
const struct value_string om2k_mo_class_short_vals[] = {
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0x01, "TRXC" },
om2k: Add definitions for the TG object Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I6d90554a54baa78f454281a486e4b5e95784fdee
2020-04-28 07:29:14 +00:00
{ 0x02, "TG" },
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0x03, "TS" },
{ 0x04, "TF" },
{ 0x05, "IS" },
{ 0x06, "CON" },
{ 0x07, "DP" },
OM2000: Add some more message types and IEs we now understand OM2000 is not only used for the venerable RBS2000 family, but also for the more modern RBS6000 family, specifically the DUG 20 GSM baseband unit. In RBS6000, there are some protocol extensions which are not yet fully understood. However, we are understanding some bits around the MCTR (multi carrier transceiver?), a new MO that appears to be present for every physical RUS (Radio Unit) attached to the DUG 20. Let's add what we have learned so far. Thanks to Sylvain Munaut for his help with this. Change-Id: Ib868358eca12b94c4fcca58e94ec8ab1a4edfda2
2020-01-29 08:19:45 +00:00
{ 0x08, "MCTR" },
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{ 0x0a, "CF" },
{ 0x0b, "TX" },
{ 0x0c, "RX" },
{ 0, NULL }
};
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
const struct value_string om2k_result_strings[] = {
{ 0x02, "Wrong state or out of sequence" },
{ 0x03, "File error" },
{ 0x04, "Fault, unspecified" },
{ 0x05, "Tuning fault" },
{ 0x06, "Protocol error" },
{ 0x07, "MO not connected" },
{ 0x08, "Parameter error" },
{ 0x09, "Optional function not supported" },
{ 0x0a, "Local access state LOCALLY DISCONNECTED" },
{ 0, NULL }
};
const struct value_string om2k_accordance_strings[] = {
{ 0x00, "Data according to request" },
{ 0x01, "Data not according to request" },
{ 0x02, "Inconsistent MO data" },
{ 0x03, "Capability constraint violation" },
{ 0, NULL }
};
const struct value_string om2k_mostate_vals[] = {
{ 0x00, "RESET" },
{ 0x01, "STARTED" },
{ 0x02, "ENABLED" },
{ 0x03, "DISABLED" },
{ 0, NULL }
};
/* entire decoded OM2K message (header + parsed TLV) */
struct om2k_decoded_msg {
struct abis_om2k_hdr o2h;
uint16_t msg_type;
struct tlv_parsed tp;
};
/* resolve the OM2000 Managed Object by BTS + MO Address */
static struct om2k_mo *
get_om2k_mo(struct gsm_bts *bts, const struct abis_om2k_mo *abis_mo)
{
struct om2k_mo *mo = NULL;
struct gsm_bts_trx *trx;
switch (abis_mo->class) {
abis_om2000: Handle DP object in get_om2k_mo() This was somehow forgotten for the DP MO... Change-Id: Ie2ef736eb712e01a3c657ae30c8422c9aa32062a
2020-12-21 17:44:32 +00:00
case OM2K_MO_CLS_DP:
mo = &bts->rbs2000.dp.om2k_mo;
break;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MO_CLS_CF:
mo = &bts->rbs2000.cf.om2k_mo;
break;
case OM2K_MO_CLS_CON:
mo = &bts->rbs2000.con.om2k_mo;
break;
case OM2K_MO_CLS_IS:
mo = &bts->rbs2000.is.om2k_mo;
break;
case OM2K_MO_CLS_TF:
mo = &bts->rbs2000.tf.om2k_mo;
break;
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MO_CLS_MCTR:
mo = &bts->rbs2000.mctr.om2k_mo;
break;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MO_CLS_TRXC:
trx = gsm_bts_trx_num(bts, abis_mo->inst);
if (!trx)
return NULL;
mo = &trx->rbs2000.trxc.om2k_mo;
break;
case OM2K_MO_CLS_TX:
trx = gsm_bts_trx_num(bts, abis_mo->inst);
if (!trx)
return NULL;
mo = &trx->rbs2000.tx.om2k_mo;
break;
case OM2K_MO_CLS_RX:
trx = gsm_bts_trx_num(bts, abis_mo->inst);
if (!trx)
return NULL;
mo = &trx->rbs2000.rx.om2k_mo;
break;
case OM2K_MO_CLS_TS:
trx = gsm_bts_trx_num(bts, abis_mo->assoc_so);
if (!trx)
return NULL;
if (abis_mo->inst >= ARRAY_SIZE(trx->ts))
return NULL;
mo = &trx->ts[abis_mo->inst].rbs2000.om2k_mo;
break;
default:
return NULL;
};
return mo;
}
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
static struct msgb *om2k_msgb_alloc(void)
{
return msgb_alloc_headroom(OM_ALLOC_SIZE, OM_HEADROOM_SIZE,
"OM2000");
}
misc: Move from u_int to uint types of stdint.h This was done with sed on the files.
2011-04-18 15:04:00 +00:00
static int abis_om2k_tlv_parse(struct tlv_parsed *tp, const uint8_t *buf, int len)
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
{
return tlv_parse(tp, &om2k_att_tlvdef, buf, len, 0, 0);
}
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
static int abis_om2k_msg_tlv_parse(struct tlv_parsed *tp, struct abis_om2k_hdr *oh)
{
return abis_om2k_tlv_parse(tp, oh->data, oh->om.length - 6);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* decode/parse the message */
static int om2k_decode_msg(struct om2k_decoded_msg *odm, struct msgb *msg)
{
struct abis_om2k_hdr *o2h = msgb_l2(msg);
odm->msg_type = ntohs(o2h->msg_type);
odm->o2h = *o2h;
return abis_om2k_msg_tlv_parse(&odm->tp, o2h);
}
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
const char *abis_om2k_mo_name(const struct abis_om2k_mo *mo)
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
{
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
static char mo_buf[64];
memset(mo_buf, 0, sizeof(mo_buf));
snprintf(mo_buf, sizeof(mo_buf), "%s/%02x/%02x/%02x",
get_value_string(om2k_mo_class_short_vals, mo->class),
mo->bts, mo->assoc_so, mo->inst);
return mo_buf;
}
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
/* resolve the gsm_nm_state data structure for a given MO */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static struct gsm_nm_state *mo2nm_state(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
{
struct gsm_bts_trx *trx;
struct gsm_nm_state *nm_state = NULL;
switch (mo->class) {
case OM2K_MO_CLS_TRXC:
OM2000: use assoc_so *only* for TS objects all other objects always use the MO instance. The existing code likely is due to copy+paste mistakes. Change-Id: Ie0a31cd93993da10f31eecf530a5a05773c11eb1
2016-11-15 21:09:08 +00:00
trx = gsm_bts_trx_num(bts, mo->inst);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
if (!trx)
return NULL;
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &trx->mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_TS:
trx = gsm_bts_trx_num(bts, mo->assoc_so);
if (!trx)
return NULL;
if (mo->inst >= ARRAY_SIZE(trx->ts))
return NULL;
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &trx->ts[mo->inst].mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
case OM2K_MO_CLS_MCTR:
nm_state = &bts->rbs2000.mctr.mo.nm_state;
break;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
case OM2K_MO_CLS_TF:
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &bts->rbs2000.tf.mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_IS:
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &bts->rbs2000.is.mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_CON:
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &bts->rbs2000.con.mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_DP:
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &bts->rbs2000.con.mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_CF:
gsm_data_shared: introduce 'struct gsm_abis_mo' ... as a common wrapper around nm_attr and nm_state
2011-05-30 10:07:53 +00:00
nm_state = &bts->mo.nm_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_TX:
OM2000: use assoc_so *only* for TS objects all other objects always use the MO instance. The existing code likely is due to copy+paste mistakes. Change-Id: Ie0a31cd93993da10f31eecf530a5a05773c11eb1
2016-11-15 21:09:08 +00:00
trx = gsm_bts_trx_num(bts, mo->inst);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
if (!trx)
return NULL;
OM2000: Add FIXME comments for missing resolving of RX/TX MO! Change-Id: I45708df724c5fc3316eca6bd2ac2c0738b19a45b
2016-11-15 21:58:49 +00:00
/* FIXME */
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
case OM2K_MO_CLS_RX:
OM2000: use assoc_so *only* for TS objects all other objects always use the MO instance. The existing code likely is due to copy+paste mistakes. Change-Id: Ie0a31cd93993da10f31eecf530a5a05773c11eb1
2016-11-15 21:09:08 +00:00
trx = gsm_bts_trx_num(bts, mo->inst);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
if (!trx)
return NULL;
OM2000: Add FIXME comments for missing resolving of RX/TX MO! Change-Id: I45708df724c5fc3316eca6bd2ac2c0738b19a45b
2016-11-15 21:58:49 +00:00
/* FIXME */
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
break;
}
return nm_state;
}
abis_om2000: constify mo in mo2obj The function mo2obj does only read only access to mo, so mo can be const. Change-Id: Ia09d0c96c8938e287fdbad343b9605cbfd6ff6a1 Related: OS#5101
2022-08-05 09:54:15 +00:00
static void *mo2obj(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
{
struct gsm_bts_trx *trx;
switch (mo->class) {
case OM2K_MO_CLS_TX:
case OM2K_MO_CLS_RX:
case OM2K_MO_CLS_TRXC:
OM2000: use assoc_so *only* for TS objects all other objects always use the MO instance. The existing code likely is due to copy+paste mistakes. Change-Id: Ie0a31cd93993da10f31eecf530a5a05773c11eb1
2016-11-15 21:09:08 +00:00
return gsm_bts_trx_num(bts, mo->inst);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
case OM2K_MO_CLS_TS:
trx = gsm_bts_trx_num(bts, mo->assoc_so);
if (!trx)
return NULL;
if (mo->inst >= ARRAY_SIZE(trx->ts))
return NULL;
return &trx->ts[mo->inst];
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
case OM2K_MO_CLS_MCTR:
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
case OM2K_MO_CLS_TF:
case OM2K_MO_CLS_IS:
case OM2K_MO_CLS_CON:
case OM2K_MO_CLS_DP:
case OM2K_MO_CLS_CF:
return bts;
}
return NULL;
}
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
/* Derive an OML Availability state from an OM2000 MO state */
static enum abis_nm_avail_state abis_nm_av_state_from_om2k_av_state(struct abis_om2k_mo *mo, uint8_t mo_state)
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
{
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
bool has_enabled_state;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
switch (mo->class) {
case OM2K_MO_CLS_CF:
case OM2K_MO_CLS_TRXC:
has_enabled_state = false;
break;
default:
has_enabled_state = true;
break;
}
switch (mo_state) {
case OM2K_MOSTATE_RESET:
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
return NM_AVSTATE_POWER_OFF;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
case OM2K_MOSTATE_STARTED:
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
return has_enabled_state ? NM_AVSTATE_OFF_LINE : NM_AVSTATE_OK;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
case OM2K_MOSTATE_ENABLED:
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
return NM_AVSTATE_OK;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
case OM2K_MOSTATE_DISABLED:
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
return NM_AVSTATE_POWER_OFF;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
default:
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
return NM_AVSTATE_DEGRADED;
om2k: Properly update the 'fake' 12.21 states using OM2000 status Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I1caafa2e87c6198bf3b1a77f0fa1edc774deeef9
2020-05-08 07:59:23 +00:00
}
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
}
static void update_mo_state(struct gsm_bts *bts, struct abis_om2k_mo *mo, uint8_t mo_state)
{
struct gsm_nm_state *nm_state = mo2nm_state(bts, mo);
struct nm_statechg_signal_data nsd;
if (!nm_state)
return;
memset(&nsd, 0, sizeof(nsd));
nsd.bts = bts;
nsd.obj = mo2obj(bts, mo);
nsd.old_state = *nm_state;
nsd.new_state = *nm_state;
nsd.om2k_mo = mo;
nsd.new_state.availability = abis_nm_av_state_from_om2k_av_state(mo, mo_state);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
Update current NM object state before signalling S_NM_STATECHG This way code triggered through signal has an updated view of the object tree when running generic code which queries the current state of objects. This way for instance one can use APIs like trx_is_usable() or alike. Change-Id: Ib46234e3f3e446e866d27b0dfee65edf4af4d2ba
2022-05-04 14:23:24 +00:00
/* Update current state before emitting signal: */
nm_statechg_signal_data: Convert state ptr to data There's no need to use pointers there, it is only asking for errors from code handling the data structe from the signal by attempting to change them. Even for mem size point of view it doesn't make sense, since it's 3 byte vs a 4 byte pointer. Furthermore, this is a preparation for new commit, where the NM object current state will be updated before emitting the signal. This patch eases a lot the follow up mentioned patch. Change-Id: I9b648dfd8392b7b40bfe2b38f3345017481f5129
2022-05-04 14:17:04 +00:00
nm_state->availability = nsd.new_state.availability;
Update current NM object state before signalling S_NM_STATECHG This way code triggered through signal has an updated view of the object tree when running generic code which queries the current state of objects. This way for instance one can use APIs like trx_is_usable() or alike. Change-Id: Ib46234e3f3e446e866d27b0dfee65edf4af4d2ba
2022-05-04 14:23:24 +00:00
osmo_signal_dispatch(SS_NM, S_NM_STATECHG, &nsd);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
}
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
/* Derive an OML Operational state from an OM2000 OP state */
static enum abis_nm_op_state abis_nm_op_state_from_om2k_op_state(uint8_t op_state)
{
switch (op_state) {
case 1:
return NM_OPSTATE_ENABLED;
case 0:
return NM_OPSTATE_DISABLED;
default:
return NM_OPSTATE_NULL;
}
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static void update_op_state(struct gsm_bts *bts, const struct abis_om2k_mo *mo, uint8_t op_state)
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
{
struct gsm_nm_state *nm_state = mo2nm_state(bts, mo);
abis_om2000: update_op_state() does not send signal The function update_op_state() updates the OML MO, but it does not notify the update to other entities using S_NM_STATECHG Change-Id: Id19c6beb2bc79c4db0ec07ef593aacb57fff8b75 Related: OS#5634
2022-08-03 13:15:48 +00:00
struct nm_statechg_signal_data nsd;
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
if (!nm_state)
return;
abis_om2000: update_op_state() does not send signal The function update_op_state() updates the OML MO, but it does not notify the update to other entities using S_NM_STATECHG Change-Id: Id19c6beb2bc79c4db0ec07ef593aacb57fff8b75 Related: OS#5634
2022-08-03 13:15:48 +00:00
memset(&nsd, 0, sizeof(nsd));
nsd.bts = bts;
nsd.obj = mo2obj(bts, mo);
nsd.old_state = *nm_state;
nsd.new_state = *nm_state;
nsd.om2k_mo = mo;
abis_om2000: move switch-case to function The switch-case in update_op_state() and update_mo_state() can be split off into functions. This makes to code better readable. Change-Id: I41f0d9d0d498f6f698c2c959baac50424f5ac317 Related: OS#5634
2022-08-03 09:50:58 +00:00
abis_om2000: update_op_state() does not send signal The function update_op_state() updates the OML MO, but it does not notify the update to other entities using S_NM_STATECHG Change-Id: Id19c6beb2bc79c4db0ec07ef593aacb57fff8b75 Related: OS#5634
2022-08-03 13:15:48 +00:00
nsd.new_state.operational = abis_nm_op_state_from_om2k_op_state(op_state);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
abis_om2000: update_op_state() does not send signal The function update_op_state() updates the OML MO, but it does not notify the update to other entities using S_NM_STATECHG Change-Id: Id19c6beb2bc79c4db0ec07ef593aacb57fff8b75 Related: OS#5634
2022-08-03 13:15:48 +00:00
/* Update current state before emitting signal: */
nm_state->operational = nsd.new_state.operational;
osmo_signal_dispatch(SS_NM, S_NM_STATECHG, &nsd);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
}
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
static int abis_om2k_sendmsg(struct gsm_bts *bts, struct msgb *msg)
{
struct abis_om2k_hdr *o2h;
Fix Ericsson RBS2000 support after libosmo-abis merge The libosmo-abis merge broke Ericsson RBS support, as it didn't get the part right where the per-TRX OML sign_link is determined while transmitting OM2000 messages. As a result of this fix, we can remove the 'to_trx_oml' parameter to _abis_nm_sendmsg(), which is a nice cleanup.
2011-09-26 21:43:23 +00:00
struct gsm_bts_trx *trx;
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
msg->l2h = msg->data;
o2h = (struct abis_om2k_hdr *) msg->l2h;
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
/* Compute the length in the OML header */
o2h->om.length = 6 + msgb_l2len(msg)-sizeof(*o2h);
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
switch (o2h->mo.class) {
case OM2K_MO_CLS_TRXC:
case OM2K_MO_CLS_TX:
case OM2K_MO_CLS_RX:
/* Route through per-TRX OML Link to the appropriate TRX */
bts: Drop duplicated function to get trx by number In the big mess of gsm_data we reached a point where we have multiple functions doing the same thing, most probably because it's hard finding stuff in there. Let's drop one of them (the one which less callers) and move it to bts.*, where it belongs. Change-Id: I9071a0ab250844619280fbe2be63ed99f2c87eb1
2020-07-16 12:20:32 +00:00
trx = gsm_bts_trx_num(bts, o2h->mo.inst);
Fix Ericsson RBS2000 support after libosmo-abis merge The libosmo-abis merge broke Ericsson RBS support, as it didn't get the part right where the per-TRX OML sign_link is determined while transmitting OM2000 messages. As a result of this fix, we can remove the 'to_trx_oml' parameter to _abis_nm_sendmsg(), which is a nice cleanup.
2011-09-26 21:43:23 +00:00
if (!trx) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "MO=%s Tx Dropping msg to non-existing TRX\n",
abis_om2k_mo_name(&o2h->mo));
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
return -ENODEV;
}
OM2000: make sure to route messages to IXU/DXU over BTS oml link Messages to CF and other core objects need to go to the BTS (IXU/DXU) OML link, not to the OML link of the primary TRX!
2012-09-07 17:09:46 +00:00
msg->dst = trx->oml_link;
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
break;
OM2000: Route TS config requests to the right TRX The TRX number is encoded in mo.assoc_so, whereas the TS number is in mo.inst!
2011-03-05 19:15:09 +00:00
case OM2K_MO_CLS_TS:
/* Route through per-TRX OML Link to the appropriate TRX */
bts: Drop duplicated function to get trx by number In the big mess of gsm_data we reached a point where we have multiple functions doing the same thing, most probably because it's hard finding stuff in there. Let's drop one of them (the one which less callers) and move it to bts.*, where it belongs. Change-Id: I9071a0ab250844619280fbe2be63ed99f2c87eb1
2020-07-16 12:20:32 +00:00
trx = gsm_bts_trx_num(bts, o2h->mo.assoc_so);
Fix Ericsson RBS2000 support after libosmo-abis merge The libosmo-abis merge broke Ericsson RBS support, as it didn't get the part right where the per-TRX OML sign_link is determined while transmitting OM2000 messages. As a result of this fix, we can remove the 'to_trx_oml' parameter to _abis_nm_sendmsg(), which is a nice cleanup.
2011-09-26 21:43:23 +00:00
if (!trx) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "MO=%s Tx Dropping msg to non-existing TRX\n",
abis_om2k_mo_name(&o2h->mo));
OM2000: Route TS config requests to the right TRX The TRX number is encoded in mo.assoc_so, whereas the TS number is in mo.inst!
2011-03-05 19:15:09 +00:00
return -ENODEV;
}
OM2000: make sure to route messages to IXU/DXU over BTS oml link Messages to CF and other core objects need to go to the BTS (IXU/DXU) OML link, not to the OML link of the primary TRX!
2012-09-07 17:09:46 +00:00
msg->dst = trx->oml_link;
OM2000: Route TS config requests to the right TRX The TRX number is encoded in mo.assoc_so, whereas the TS number is in mo.inst!
2011-03-05 19:15:09 +00:00
break;
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
default:
/* Route through the IXU/DXU OML Link */
OM2000: make sure to route messages to IXU/DXU over BTS oml link Messages to CF and other core objects need to go to the BTS (IXU/DXU) OML link, not to the OML link of the primary TRX!
2012-09-07 17:09:46 +00:00
msg->dst = bts->oml_link;
OM2000: Route outgoing NM message depending on MO Depending on the MO we adress, select the proper OML link
2011-02-14 14:26:13 +00:00
break;
}
Fix Ericsson RBS2000 support after libosmo-abis merge The libosmo-abis merge broke Ericsson RBS support, as it didn't get the part right where the per-TRX OML sign_link is determined while transmitting OM2000 messages. As a result of this fix, we can remove the 'to_trx_oml' parameter to _abis_nm_sendmsg(), which is a nice cleanup.
2011-09-26 21:43:23 +00:00
return _abis_nm_sendmsg(msg);
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static void fill_om2k_hdr(struct abis_om2k_hdr *o2h, const struct abis_om2k_mo *mo, uint16_t msg_type)
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{
o2h->om.mdisc = ABIS_OM_MDISC_FOM;
o2h->om.placement = ABIS_OM_PLACEMENT_ONLY;
o2h->om.sequence = 0;
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
/* We fill o2h->om.length later during om2k_sendmsg() */
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
o2h->msg_type = htons(msg_type);
memcpy(&o2h->mo, mo, sizeof(o2h->mo));
}
static int abis_om2k_cal_time_resp(struct gsm_bts *bts)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
time_t tm_t;
struct tm *tm;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fill_om2k_hdr(o2k, &bts->rbs2000.cf.om2k_mo.addr, OM2K_MSGT_CAL_TIME_RESP);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
tm_t = time(NULL);
tm = localtime(&tm_t);
msgb_put_u8(msg, OM2K_DEI_CAL_TIME);
msgb_put_u8(msg, tm->tm_year % 100);
msgb_put_u8(msg, tm->tm_mon + 1);
msgb_put_u8(msg, tm->tm_mday);
msgb_put_u8(msg, tm->tm_hour);
msgb_put_u8(msg, tm->tm_min);
msgb_put_u8(msg, tm->tm_sec);
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
return abis_om2k_sendmsg(bts, msg);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static int abis_om2k_tx_simple(struct gsm_bts *bts, const struct abis_om2k_mo *mo, uint16_t msg_type)
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, mo, msg_type);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(mo), get_value_string(om2k_msgcode_vals, msg_type));
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
return abis_om2k_sendmsg(bts, msg);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
}
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
int abis_om2k_tx_reset_cmd(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_RESET_CMD);
}
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
int abis_om2k_tx_start_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_START_REQ);
}
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
int abis_om2k_tx_status_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
OM2000: Introduce VTY commands to initiate OM2000 procedures
2011-02-12 13:42:59 +00:00
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_STATUS_REQ);
}
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
int abis_om2k_tx_connect_cmd(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_CONNECT_CMD);
}
int abis_om2k_tx_disconnect_cmd(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_DISCONNECT_CMD);
}
OM2000: Add support for sending the TEST REQUEST message
2011-02-12 14:07:30 +00:00
int abis_om2k_tx_test_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_TEST_REQ);
}
OM2000: Add enable/disable commands
2011-02-12 17:48:53 +00:00
int abis_om2k_tx_enable_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_ENABLE_REQ);
}
int abis_om2k_tx_disable_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_DISABLE_REQ);
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
int abis_om2k_tx_op_info(struct gsm_bts *bts, const struct abis_om2k_mo *mo, uint8_t operational)
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, mo, OM2K_MSGT_OP_INFO);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
msgb_tv_put(msg, OM2K_DEI_OP_INFO, operational);
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(mo),
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_OP_INFO));
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
/* we update the state here... and send the signal at ACK */
update_op_state(bts, mo, operational);
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
return abis_om2k_sendmsg(bts, msg);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
}
om2000: Add support for querying RBS capabilities Change-Id: Id1fbaf41286f3651ce8f210eb8da05fb51179c96
2016-10-15 19:24:57 +00:00
int abis_om2k_tx_cap_req(struct gsm_bts *bts, const struct abis_om2k_mo *mo)
{
return abis_om2k_tx_simple(bts, mo, OM2K_MSGT_CAPA_REQ);
}
om2k: Add VTY command to allow TX of arbitrary message for testing Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I5a385614fc670946f83ea79cc176c2d448675672
2020-04-28 07:30:52 +00:00
int abis_om2k_tx_arb(struct gsm_bts *bts, struct abis_om2k_mo *mo,
uint16_t req, uint8_t *buf, int buf_len)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
fill_om2k_hdr(o2k, mo, req);
if (buf_len)
memcpy(msgb_put(msg, buf_len), buf, buf_len);
return abis_om2k_sendmsg(bts, msg);
}
OM2000: Move IS conn group list generation out of VTY code
2011-03-06 20:49:21 +00:00
static void om2k_fill_is_conn_grp(struct om2k_is_conn_grp *grp, uint16_t icp1,
uint16_t icp2, uint8_t cont_idx)
{
grp->icp1 = htons(icp1);
grp->icp2 = htons(icp2);
grp->cont_idx = cont_idx;
}
int abis_om2k_tx_is_conf_req(struct gsm_bts *bts)
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
OM2000: Move IS conn group list generation out of VTY code
2011-03-06 20:49:21 +00:00
struct is_conn_group *grp;
unsigned int num_grps = 0, i = 0;
struct om2k_is_conn_grp *cg;
/* count number of groups in linked list */
llist_for_each_entry(grp, &bts->rbs2000.is.conn_groups, list)
num_grps++;
if (!num_grps)
return -EINVAL;
/* allocate buffer for oml group array */
cg = talloc_zero_array(bts, struct om2k_is_conn_grp, num_grps);
/* fill array with data from linked list */
llist_for_each_entry(grp, &bts->rbs2000.is.conn_groups, list)
om2k_fill_is_conn_grp(&cg[i++], grp->icp1, grp->icp2, grp->ci);
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fill_om2k_hdr(o2k, &bts->rbs2000.is.om2k_mo.addr, OM2K_MSGT_IS_CONF_REQ);
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
msgb_tv_put(msg, OM2K_DEI_LIST_NR, 1);
msgb_tv_put(msg, OM2K_DEI_END_LIST_NR, 1);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
msgb_tlv_put(msg, OM2K_DEI_IS_CONN_LIST, num_grps * sizeof(*cg), (uint8_t *)cg);
OM2000: Move IS conn group list generation out of VTY code
2011-03-06 20:49:21 +00:00
talloc_free(cg);
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(&bts->rbs2000.is.om2k_mo.addr),
OM2000: don't silently send CON/IS/TF configuration request
2012-09-07 18:15:50 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_IS_CONF_REQ));
OM2000: Add IS Conf Req
2011-02-12 19:23:40 +00:00
return abis_om2k_sendmsg(bts, msg);
}
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
int abis_om2k_tx_con_conf_req(struct gsm_bts *bts)
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
struct con_group *grp;
unsigned int num_grps = 0;
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
/* count number of groups in linked list */
llist_for_each_entry(grp, &bts->rbs2000.con.conn_groups, list)
num_grps++;
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
if (!num_grps)
return -EINVAL;
/* first build the value part of the OM2K_DEI_CON_CONN_LIST DEI */
msgb_put_u8(msg, num_grps);
llist_for_each_entry(grp, &bts->rbs2000.con.conn_groups, list) {
struct con_path *cp;
unsigned int num_paths = 0;
llist_for_each_entry(cp, &grp->paths, list)
num_paths++;
msgb_put_u8(msg, num_paths);
llist_for_each_entry(cp, &grp->paths, list) {
struct om2k_con_path *om2k_cp;
om2k_cp = (struct om2k_con_path *) msgb_put(msg, sizeof(*om2k_cp));
om2k_cp->ccp = htons(cp->ccp);
om2k_cp->ci = cp->ci;
om2k_cp->tag = cp->tag;
om2k_cp->tei = cp->tei;
}
}
msgb_push_u8(msg, msgb_length(msg));
msgb_push_u8(msg, OM2K_DEI_CON_CONN_LIST);
/* pre-pend the list number DEIs */
msgb_tv_push(msg, OM2K_DEI_END_LIST_NR, 1);
msgb_tv_push(msg, OM2K_DEI_LIST_NR, 1);
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
/* pre-pend the OM2K header */
o2k = (struct abis_om2k_hdr *) msgb_push(msg, sizeof(*o2k));
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fill_om2k_hdr(o2k, &bts->rbs2000.con.om2k_mo.addr, OM2K_MSGT_CON_CONF_REQ);
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(&bts->rbs2000.con.om2k_mo.addr),
OM2000: don't silently send CON/IS/TF configuration request
2012-09-07 18:15:50 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_CON_CONF_REQ));
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
return abis_om2k_sendmsg(bts, msg);
}
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
int abis_om2k_tx_mctr_conf_req(struct gsm_bts *bts)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
struct gsm_bts_trx *trx;
uint8_t trxc_list = 0;
const uint8_t features[] = { 0x00 };
/* build trxc list */
llist_for_each_entry(trx, &bts->trx_list, list)
trxc_list |= (1 << trx->nr);
/* fill message */
msgb_tv16_put(msg, OM2K_DEI_TRXC_LIST, osmo_swab16(trxc_list)); /* Read as LE by the BTS ... */
msgb_tv_put (msg, OM2K_DEI_MAX_ALLOWED_POWER, 0x31);
msgb_tv_put (msg, OM2K_DEI_MAX_ALLOWED_NUM_TRXCS, 0x08);
msgb_tlv_put (msg, OM2K_DEI_MCTR_FEAT_STATUS_BMAP, 1, features);
/* pre-pend the OM2K header */
o2k = (struct abis_om2k_hdr *) msgb_push(msg, sizeof(*o2k));
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fill_om2k_hdr(o2k, &bts->rbs2000.mctr.om2k_mo.addr, OM2K_MSGT_MCTR_CONF_REQ);
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(&bts->rbs2000.mctr.om2k_mo.addr),
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_MCTR_CONF_REQ));
return abis_om2k_sendmsg(bts, msg);
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static void om2k_trx_to_mo(struct abis_om2k_mo *mo, const struct gsm_bts_trx *trx,
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
enum abis_om2k_mo_cls cls)
{
mo->class = cls;
mo->bts = 0;
mo->inst = trx->nr;
OM2000: Make sure to set the Associated SO to 255 in case of RX/TX
2011-03-05 15:21:01 +00:00
mo->assoc_so = 255;
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static void om2k_ts_to_mo(struct abis_om2k_mo *mo, const struct gsm_bts_trx_ts *ts)
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
{
mo->class = OM2K_MO_CLS_TS;
mo->bts = 0;
mo->inst = ts->nr;
mo->assoc_so = ts->trx->nr;
}
/* Configure a Receiver MO */
int abis_om2k_tx_rx_conf_req(struct gsm_bts_trx *trx)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
struct abis_om2k_mo mo;
om2k_trx_to_mo(&mo, trx, OM2K_MO_CLS_RX);
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, &mo, OM2K_MSGT_RX_CONF_REQ);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
/* OM2K_DEI_FREQ_SPEC_RX: Using trx_nr as "RX address" only works for single MCTR case */
msgb_tv16_put(msg, OM2K_DEI_FREQ_SPEC_RX, 0x8000 | ((uint16_t)trx->nr << 10));
Add command to enable RX diversity to RBS2000 Allow selection of RX diversity from VTY Options are a,ab,b Default is 'a' so there is no change from previous behavior Change-Id: I430762b8cfa51060841d90ba4446de73bd557c6c
2021-03-11 00:00:25 +00:00
msgb_tv_put(msg, OM2K_DEI_RX_DIVERSITY, trx->rbs2000.rx_diversity);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
return abis_om2k_sendmsg(trx->bts, msg);
}
/* Configure a Transmitter MO */
int abis_om2k_tx_tx_conf_req(struct gsm_bts_trx *trx)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
struct abis_om2k_mo mo;
om2k_trx_to_mo(&mo, trx, OM2K_MO_CLS_TX);
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, &mo, OM2K_MSGT_TX_CONF_REQ);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
/* OM2K_DEI_FREQ_SPEC_TX: Using trx_nr as "TX address" only works for single MCTR case */
msgb_tv16_put(msg, OM2K_DEI_FREQ_SPEC_TX, trx->arfcn | ((uint16_t)trx->nr << 10));
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
msgb_tv_put(msg, OM2K_DEI_POWER, trx->nominal_power-trx->max_power_red);
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
msgb_tv_put(msg, OM2K_DEI_FILLING_MARKER, trx != trx->bts->c0); /* Filling enabled for C0 only */
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
msgb_tv_put(msg, OM2K_DEI_BCC, trx->bts->bsic & 0x7);
/* Dedication Information is optional */
return abis_om2k_sendmsg(trx->bts, msg);
}
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
enum abis_om2k_tf_mode {
OM2K_TF_MODE_MASTER = 0x00,
OM2K_TF_MODE_STANDALONE = 0x01,
OM2K_TF_MODE_SLAVE = 0x02,
OM2K_TF_MODE_UNDEFINED = 0xff,
};
static const uint8_t fs_offset_undef[5] = { 0xff, 0xff, 0xff, 0xff, 0xff };
int abis_om2k_tx_tf_conf_req(struct gsm_bts *bts)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fill_om2k_hdr(o2k, &bts->rbs2000.tf.om2k_mo.addr, OM2K_MSGT_TF_CONF_REQ);
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
msgb_tv_put(msg, OM2K_DEI_TF_MODE, OM2K_TF_MODE_STANDALONE);
Add vty command for Ericsson RBS2000 sync This commit adds support for Selection of syncronization source. Options are internal for E1 and external for GPS Change-Id: Ia3d1acd6b3442238b35fc911092e12a6ac989adb
2021-03-10 23:33:46 +00:00
msgb_tv_put(msg, OM2K_DEI_TF_SYNC_SRC, bts->rbs2000.sync_src);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
msgb_tv_fixed_put(msg, OM2K_DEI_FS_OFFSET, sizeof(fs_offset_undef), fs_offset_undef);
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(&bts->rbs2000.tf.om2k_mo.addr),
OM2000: don't silently send CON/IS/TF configuration request
2012-09-07 18:15:50 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_TF_CONF_REQ));
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
return abis_om2k_sendmsg(bts, msg);
}
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
static uint8_t pchan2comb(enum gsm_phys_chan_config pchan)
{
switch (pchan) {
case GSM_PCHAN_CCCH:
return 4;
case GSM_PCHAN_CCCH_SDCCH4:
return 5;
case GSM_PCHAN_SDCCH8_SACCH8C:
return 3;
case GSM_PCHAN_TCH_F:
case GSM_PCHAN_TCH_H:
case GSM_PCHAN_PDCH:
OM2000: Fix missing dynamic TCH initialization When OM2000 has confirmed that a TS is started, call dyn_ts_init() on the timeslot to start the processing for fully dynamic (osmocom style) TCH/F_TCH/H_PDCH. This should in turn trigger the activation of idle timeslots as PDCH until we want to allocate any of them for TCH/F or TCH/H. Change-Id: I1a1fd61d6afd85449cacad4bacfb830252dab6b1
2016-11-15 22:23:42 +00:00
case GSM_PCHAN_TCH_F_PDCH:
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
case GSM_PCHAN_OSMO_DYN:
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
return 8;
default:
return 0;
}
}
dyn TS: fix: abis_om2000: also handle dyn TS as TCH Add ts2comb() to switch on dyn TS so that dyn TS in TCH mode are also treated like normal TCH/H or TCH/F pchans. Use ts2comb() instead of pchan2comb(). Change-Id: Iddc51a4409488d91db59228ca66aaab73ce3f1df
2016-09-25 13:15:59 +00:00
static uint8_t ts2comb(struct gsm_bts_trx_ts *ts)
{
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 == GSM_PCHAN_TCH_F_PDCH) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "%s pchan %s not intended for use with OM2000, use %s instead\n",
gsm_ts_and_pchan_name(ts), gsm_pchan_name(GSM_PCHAN_TCH_F_PDCH),
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_name(GSM_PCHAN_OSMO_DYN));
OM2000: disallow ip.access style TCH/F_PDCH pchan type For TCH/F_PDCH, return an invalid chan comb (0) and print an error message that hints at the proper pchan type to use instead: TCH/F_TCH/H_PDCH Change-Id: Ibe0f944573f0a6d1be4bf7cf4986c4b2b3bd6d0d
2016-11-10 01:18:00 +00:00
/* If we allowed initialization of TCH/F_PDCH, it would fail
* when we try to send the ip.access specific RSL PDCH Act
* message for it. Rather fail completely right now: */
return 0;
dyn TS: fix: abis_om2000: also handle dyn TS as TCH Add ts2comb() to switch on dyn TS so that dyn TS in TCH mode are also treated like normal TCH/H or TCH/F pchans. Use ts2comb() instead of pchan2comb(). Change-Id: Iddc51a4409488d91db59228ca66aaab73ce3f1df
2016-09-25 13:15:59 +00:00
}
om2k: Use the "from config" TS config to setup OM objects During the configuration of the TS object through OML we must use pchan_from_config since it's too early to use anything else. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iecdc911a79b66d8f3d746347710ad697cb288174
2020-05-05 18:40:02 +00:00
return pchan2comb(ts->pchan_from_config);
dyn TS: fix: abis_om2000: also handle dyn TS as TCH Add ts2comb() to switch on dyn TS so that dyn TS in TCH mode are also treated like normal TCH/H or TCH/F pchans. Use ts2comb() instead of pchan2comb(). Change-Id: Iddc51a4409488d91db59228ca66aaab73ce3f1df
2016-09-25 13:15:59 +00:00
}
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
static int put_freq_list(uint8_t *buf, struct gsm_bts_trx_ts *ts, uint16_t arfcn)
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
{
om2k: Fix invalid use of linked list when building hopping freq list I originally assumed that after a complete scan with llist_for_each_entry the loop counter would be either NULL or a valid entry. That's just not the case. Fixes: CID#210256 Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iad647b74771c4ac406a88effd371ed7748c8847e
2020-05-09 06:04:47 +00:00
struct gsm_bts_trx *t, *trx = NULL;
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
/* Find the TRX that's configured for that ARFCN */
om2k: Fix invalid use of linked list when building hopping freq list I originally assumed that after a complete scan with llist_for_each_entry the loop counter would be either NULL or a valid entry. That's just not the case. Fixes: CID#210256 Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iad647b74771c4ac406a88effd371ed7748c8847e
2020-05-09 06:04:47 +00:00
llist_for_each_entry(t, &ts->trx->bts->trx_list, list)
if (t->arfcn == arfcn) {
trx = t;
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
break;
om2k: Fix invalid use of linked list when building hopping freq list I originally assumed that after a complete scan with llist_for_each_entry the loop counter would be either NULL or a valid entry. That's just not the case. Fixes: CID#210256 Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iad647b74771c4ac406a88effd371ed7748c8847e
2020-05-09 06:04:47 +00:00
}
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
om2k: Fix invalid use of linked list when building hopping freq list I originally assumed that after a complete scan with llist_for_each_entry the loop counter would be either NULL or a valid entry. That's just not the case. Fixes: CID#210256 Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iad647b74771c4ac406a88effd371ed7748c8847e
2020-05-09 06:04:47 +00:00
if (!trx) {
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
LOGP(DNM, LOGL_ERROR, "Trying to use ARFCN %d for hopping with no TRX configured for it", arfcn);
return 0;
}
/*
* [7:4] - TX address
* This must be the same number that was used when configuring the TX
* MO object with that target arfcn
*
* [3:0] - RX address
* The logical TRX number we're configuring the hopping sequence for
* This must basically match the MO object instance number
*
* ATM since we only support 1 MCTR, we use trx->nr
*/
buf[0] = (trx->nr << 4) | ts->trx->nr;
/* ARFCN Number */
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
buf[1] = (arfcn >> 8);
buf[2] = (arfcn & 0xff);
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
/* C0 marker */
if (trx == trx->bts->c0)
buf[1] |= 0x04;
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
return 3;
}
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
/* Compute a frequency list in OM2000 fomrmat */
static int om2k_gen_freq_list(uint8_t *list, struct gsm_bts_trx_ts *ts)
{
uint8_t *cur = list;
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
int len;
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
if (ts->hopping.enabled) {
unsigned int i;
for (i = 0; i < ts->hopping.arfcns.data_len*8; i++) {
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
if (bitvec_get_bit_pos(&ts->hopping.arfcns, i))
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
cur += put_freq_list(cur, ts, i);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
}
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
} else
om2k: Fix the frequency specifier for TX/RX/TS conf requests * OM2K_DEI_FREQ_SPEC_RX: (hop << 15) | (rx_addr << 10) | arfcn . hop Hopping marker . rx_addr is not really the TRX number, it's just a sequential number different for all TRX, doesn't need to be 'in-order' of TRX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_SPEC_TX: (hop << 15) | (tx_id << 10) | arfcn tx_id Pretty much same as rx_id except here we know that the order doesn't have to match TRX order. It seems to even vary from one reboot to another on some real-world capture we got. . hop Hopping marker . tx_addr Same as 'rx_addr' above but for TX . arfcn The ARFCN number (0 for hopping and 1023 for 'no frequency') * OM2K_DEI_FREQ_LIST: Groups of 3 bytes (24 bits), 1 per frequency used. (tx_addr << 20) | (rx_addr << 16) | (is_c0 << 10) | arfcn . tx_addr See above . rx_addr See above . is_c0 Must be 1 if that ARFCN hosts the C0. (first TRX of a MCTR) . arfcn The ARFCN number (Note MAIO must also be set properly on the different TRX/TS sharing a frequency ... ) The way we generate theses here is what we gathered from real-world traces: - Each 'TX' of each TRX is set to the ARFCN set in that TRX config - Each 'RX' of each TRX is configures as 'hopping' (which I assume means it will just pick the appropriate freq ?) - For each TS, we use : . tx_addr of the TRX that has the ARFCN we want to TX on . rx_addr of the TRX where the TS we're configuring is . arfcn The actual ARFCN we want to add to the list This is incomplete but will work for the 1 MCTR case. Config for multiple MCTR or multiple virtual-bts still need to be handled but it's not yet known exactly how those need to be configured. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ie39a857543adaa11d1822346d8563ce3718412c8
2020-05-05 19:06:33 +00:00
cur += put_freq_list(cur, ts, ts->trx->arfcn);
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
len = cur - list;
return len;
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
}
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
const uint8_t icm_bound_params[] = { 0x02, 0x06, 0x0c, 0x16, 0x06 };
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
int abis_om2k_tx_ts_conf_req(struct gsm_bts_trx_ts *ts)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
struct abis_om2k_mo mo;
uint8_t freq_list[64*3]; /* BA max size: 64 ARFCN */
int freq_list_len;
om2k_ts_to_mo(&mo, ts);
OM2000: fix generation of frequency list DIE
2011-03-05 16:30:07 +00:00
memset(freq_list, 0, sizeof(freq_list));
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
freq_list_len = om2k_gen_freq_list(freq_list, ts);
if (freq_list_len < 0)
return freq_list_len;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, &mo, OM2K_MSGT_TS_CONF_REQ);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
dyn TS: fix: abis_om2000: also handle dyn TS as TCH Add ts2comb() to switch on dyn TS so that dyn TS in TCH mode are also treated like normal TCH/H or TCH/F pchans. Use ts2comb() instead of pchan2comb(). Change-Id: Iddc51a4409488d91db59228ca66aaab73ce3f1df
2016-09-25 13:15:59 +00:00
msgb_tv_put(msg, OM2K_DEI_COMBINATION, ts2comb(ts));
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
msgb_tv_put(msg, OM2K_DEI_TS_NR, ts->nr);
msgb_tlv_put(msg, OM2K_DEI_FREQ_LIST, freq_list_len, freq_list);
msgb_tv_put(msg, OM2K_DEI_HSN, ts->hopping.hsn);
msgb_tv_put(msg, OM2K_DEI_MAIO, ts->hopping.maio);
msgb_tv_put(msg, OM2K_DEI_BSIC, ts->trx->bts->bsic);
Add command to enable RX diversity to RBS2000 Allow selection of RX diversity from VTY Options are a,ab,b Default is 'a' so there is no change from previous behavior Change-Id: I430762b8cfa51060841d90ba4446de73bd557c6c
2021-03-11 00:00:25 +00:00
msgb_tv_put(msg, OM2K_DEI_RX_DIVERSITY, ts->trx->rbs2000.rx_diversity);
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
msgb_tv16_put(msg, OM2K_DEI_FN_OFFSET, 0);
msgb_tv_put(msg, OM2K_DEI_EXT_RANGE, 0); /* Off */
/* Optional: Interference Rejection Combining */
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_INTERF_REJ_COMB, 0x00);
om2k: Use the "from config" TS config to setup OM objects During the configuration of the TS object through OML we must use pchan_from_config since it's too early to use anything else. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iecdc911a79b66d8f3d746347710ad697cb288174
2020-05-05 18:40:02 +00:00
switch (ts->pchan_from_config) {
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
case GSM_PCHAN_CCCH:
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_put(msg, OM2K_DEI_BA_PA_MFRMS, 0x06);
msgb_tv_put(msg, OM2K_DEI_BS_AG_BKS_RES, 0x01);
msgb_tv_put(msg, OM2K_DEI_DRX_DEV_MAX, 0x05);
/* Repeat Paging/IMM.ASS: True, Allow Paging Type 3: Yes, Page for 5 seconds (default) */
msgb_tv_put(msg, OM2K_DEI_CCCH_OPTIONS, 0x01);
break;
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
case GSM_PCHAN_CCCH_SDCCH4:
add gsm_timers, for Tnnn definitions usable by FSMs Change-Id: If212fcd042051b6fa53484254223614c5b93a9c6
2018-06-16 17:29:50 +00:00
msgb_tv_put(msg, OM2K_DEI_T3105,
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
osmo_tdef_get(ts->trx->bts->network->T_defs, 3105, OSMO_TDEF_MS, -1) / 10);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_put(msg, OM2K_DEI_NY1, 35);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_BA_PA_MFRMS, 0x06);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_put(msg, OM2K_DEI_CBCH_INDICATOR, 0);
Fix more hidden TSC bugs for Ericsson BTS It seems that also the Ericsson RBS2000 code was assuming that we always use the bts-global TSC, rather than the possibly different TS-specific TSC.
2014-01-19 16:19:10 +00:00
msgb_tv_put(msg, OM2K_DEI_TSC, gsm_ts_tsc(ts));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_BS_AG_BKS_RES, 0x01);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_put(msg, OM2K_DEI_ICM_INDICATOR, 0);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_DRX_DEV_MAX, 0x05);
/* Repeat Paging/IMM.ASS: True, Allow Paging Type 3: Yes, Page for 5 seconds (default) */
msgb_tv_put(msg, OM2K_DEI_CCCH_OPTIONS, 0x01);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_fixed_put(msg, OM2K_DEI_ICM_BOUND_PARAMS,
sizeof(icm_bound_params), icm_bound_params);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
break;
case GSM_PCHAN_SDCCH8_SACCH8C:
add gsm_timers, for Tnnn definitions usable by FSMs Change-Id: If212fcd042051b6fa53484254223614c5b93a9c6
2018-06-16 17:29:50 +00:00
msgb_tv_put(msg, OM2K_DEI_T3105,
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
osmo_tdef_get(ts->trx->bts->network->T_defs, 3105, OSMO_TDEF_MS, -1) / 10);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_NY1, 35);
msgb_tv_put(msg, OM2K_DEI_CBCH_INDICATOR, 0);
Fix more hidden TSC bugs for Ericsson BTS It seems that also the Ericsson RBS2000 code was assuming that we always use the bts-global TSC, rather than the possibly different TS-specific TSC.
2014-01-19 16:19:10 +00:00
msgb_tv_put(msg, OM2K_DEI_TSC, gsm_ts_tsc(ts));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
/* Disable RF RESOURCE INDICATION on idle channels */
msgb_tv_put(msg, OM2K_DEI_ICM_INDICATOR, 0);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_fixed_put(msg, OM2K_DEI_ICM_BOUND_PARAMS,
sizeof(icm_bound_params), icm_bound_params);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
break;
default:
add gsm_timers, for Tnnn definitions usable by FSMs Change-Id: If212fcd042051b6fa53484254223614c5b93a9c6
2018-06-16 17:29:50 +00:00
msgb_tv_put(msg, OM2K_DEI_T3105,
use libosmocore osmo_tdef Move the T_defs API to libosmocore as osmo_tdefs: remove the local T_defs API and use libosmocore's osmo_tdef* API instead. The root reason is moving the mgw_endpoint_fsm to libosmo-mgcp-client to be able to use it in osmo-msc for inter-MSC handover. When adding osmo_tdef, the new concept of timer groups was added to the API. It would make sense to apply group names here as well, but do not modify the VTY configuration for timers. The future might bring separate groups (or not). Depends: Ibd6b1ed7f1bd6e1f2e0fde53352055a4468f23e5 (libosmocore) Change-Id: I66674a5d8403d820038762888c846bae10ceac58
2019-01-28 02:52:14 +00:00
osmo_tdef_get(ts->trx->bts->network->T_defs, 3105, OSMO_TDEF_MS, -1) / 10);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_NY1, 35);
Fix more hidden TSC bugs for Ericsson BTS It seems that also the Ericsson RBS2000 code was assuming that we always use the bts-global TSC, rather than the possibly different TS-specific TSC.
2014-01-19 16:19:10 +00:00
msgb_tv_put(msg, OM2K_DEI_TSC, gsm_ts_tsc(ts));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
/* Disable RF RESOURCE INDICATION on idle channels */
msgb_tv_put(msg, OM2K_DEI_ICM_INDICATOR, 0);
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_fixed_put(msg, OM2K_DEI_ICM_BOUND_PARAMS,
sizeof(icm_bound_params), icm_bound_params);
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_TTA, 10); /* Timer for Time Alignment */
om2k: Use the "from config" TS config to setup OM objects During the configuration of the TS object through OML we must use pchan_from_config since it's too early to use anything else. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Iecdc911a79b66d8f3d746347710ad697cb288174
2020-05-05 18:40:02 +00:00
if (ts->pchan_from_config == GSM_PCHAN_TCH_H)
OM2000: More work on the TS configuration request It seems the differences for the various channel combinations are big
2011-03-05 19:38:35 +00:00
msgb_tv_put(msg, OM2K_DEI_ICM_CHAN_RATE, 1); /* TCH/H */
else
msgb_tv_put(msg, OM2K_DEI_ICM_CHAN_RATE, 0); /* TCH/F */
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_LSC, 1); /* enabled */
OM2000: Fix configuration of TCH timeslots
2011-03-06 18:01:54 +00:00
msgb_tv_put(msg, OM2K_DEI_LSC_FILT_TIME, 10); /* units of 100ms */
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
msgb_tv_put(msg, OM2K_DEI_CALL_SUPV_TIME, 8);
msgb_tv_put(msg, OM2K_DEI_ENCR_ALG, 0x00);
om2k: Fix TS channel config payload for non-superchannel case * In superchannel mode, those 3 are required. * In normal mode, "Config Type" is optional and the two others are forbidden. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: If02d02c067ae8af8ce693ddfb8747212f3f4e441
2020-05-05 18:32:50 +00:00
/* Those are only use for superchannel */
if (ts->trx->bts->rbs2000.use_superchannel) {
msgb_tv_put(msg, OM2K_DEI_CONFIG_TYPE, 0x00); /* 1-bit, lsb */
msgb_tv_put(msg, OM2K_DEI_JITTER_SIZE, 0x37);
msgb_tv_put(msg, OM2K_DEI_PACKING_ALGO, 0x01);
}
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
break;
}
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n",
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
abis_om2k_mo_name(&mo),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_TS_CONF_REQ));
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
return abis_om2k_sendmsg(ts->trx->bts, msg);
}
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/***********************************************************************
* OM2000 Managed Object (MO) FSM
***********************************************************************/
#define S(x) (1 << (x))
enum om2k_event_name {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OM2K_MO_EVT_RESET,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_MO_EVT_START,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OM2K_MO_EVT_CHILD_TERM,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_MO_EVT_RX_CONN_COMPL,
OM2K_MO_EVT_RX_RESET_COMPL,
OM2K_MO_EVT_RX_START_REQ_ACCEPT,
OM2K_MO_EVT_RX_START_RES,
OM2K_MO_EVT_RX_CFG_REQ_ACCEPT,
OM2K_MO_EVT_RX_CFG_RES,
OM2K_MO_EVT_RX_ENA_REQ_ACCEPT,
OM2K_MO_EVT_RX_ENA_RES,
OM2K_MO_EVT_RX_OPINFO_ACC,
};
static const struct value_string om2k_event_names[] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_MO_EVT_RESET, "RESET" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_MO_EVT_START, "START" },
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_MO_EVT_CHILD_TERM, "CHILD-TERM" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_MO_EVT_RX_CONN_COMPL, "RX-CONN-COMPL" },
{ OM2K_MO_EVT_RX_RESET_COMPL, "RX-RESET-COMPL" },
{ OM2K_MO_EVT_RX_START_REQ_ACCEPT, "RX-RESET-REQ-ACCEPT" },
{ OM2K_MO_EVT_RX_START_RES, "RX-START-RESULT" },
{ OM2K_MO_EVT_RX_CFG_REQ_ACCEPT, "RX-CFG-REQ-ACCEPT" },
{ OM2K_MO_EVT_RX_CFG_RES, "RX-CFG-RESULT" },
{ OM2K_MO_EVT_RX_ENA_REQ_ACCEPT, "RX-ENABLE-REQ-ACCEPT" },
{ OM2K_MO_EVT_RX_ENA_RES, "RX-ENABLE-RESULT" },
{ OM2K_MO_EVT_RX_OPINFO_ACC, "RX-OPINFO-ACCEPT" },
{ 0, NULL }
};
enum om2k_mo_fsm_state {
OM2K_ST_INIT,
OM2K_ST_WAIT_CONN_COMPL,
OM2K_ST_WAIT_RES_COMPL,
OM2K_ST_WAIT_START_ACCEPT,
OM2K_ST_WAIT_START_RES,
OM2K_ST_WAIT_CFG_ACCEPT,
OM2K_ST_WAIT_CFG_RES,
OM2K_ST_WAIT_ENABLE_ACCEPT,
OM2K_ST_WAIT_ENABLE_RES,
OM2K_ST_WAIT_OPINFO_ACCEPT,
OM2K_ST_DONE,
OM2K_ST_ERROR,
};
struct om2k_mo_fsm_priv {
struct gsm_bts_trx *trx;
struct om2k_mo *mo;
uint8_t ts_nr;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
uint32_t done_event;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
};
static void om2k_mo_st_init(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
OSMO_ASSERT(event == OM2K_MO_EVT_START);
switch (omfp->mo->addr.class) {
case OM2K_MO_CLS_CF:
/* no Connect required, is always connected */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_START_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_start_req(omfp->trx->bts, &omfp->mo->addr);
break;
case OM2K_MO_CLS_TRXC:
/* no Connect required, start with Reset */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_RES_COMPL, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_reset_cmd(omfp->trx->bts, &omfp->mo->addr);
break;
default:
/* start with Connect */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_CONN_COMPL, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_connect_cmd(omfp->trx->bts, &omfp->mo->addr);
break;
}
}
static void om2k_mo_st_wait_conn_compl(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
switch (omfp->mo->addr.class) {
#if 0
case OM2K_MO_CLS_TF:
/* skip the reset, hope that helps */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_START_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_start_req(omfp->trx->bts, &omfp->mo->addr);
break;
#endif
default:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_RES_COMPL, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_reset_cmd(omfp->trx->bts, &omfp->mo->addr);
break;
}
}
static void om2k_mo_st_wait_res_compl(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_START_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_start_req(omfp->trx->bts, &omfp->mo->addr);
}
static void om2k_mo_st_wait_start_accept(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_decoded_msg *omd = data;
switch (omd->msg_type) {
case OM2K_MSGT_START_REQ_ACK:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_START_RES, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_START_REQ_REJ:
osmo_fsm_inst_state_chg(fi, OM2K_ST_ERROR, 0, 0);
break;
}
}
static void om2k_mo_st_wait_start_res(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
struct gsm_bts_trx_ts *ts;
switch (omfp->mo->addr.class) {
case OM2K_MO_CLS_CF:
case OM2K_MO_CLS_TRXC:
/* Transition directly to Operational Info */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_OPINFO_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_op_info(omfp->trx->bts, &omfp->mo->addr, 1);
OM2000: Fix state machien for CF/TRXC on START Result When receiving the 'Start Result' message, for CF and TRXC MO we directly transition to performing the Operational Info. In that case, we need to return after sending the Operational Info and skip the usual processing for the default case below. Change-Id: I99860d198b337ffe461b240bda20dc10e1b5b2cb
2016-10-15 14:30:35 +00:00
return;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MO_CLS_DP:
Fix some typos Fix typos and common misspellings in code comments and in the manual. Change-Id: I46fc9d424620c77ae9ccf78b58081bd303386d7c
2019-11-13 21:10:41 +00:00
/* Transition directory to WAIT_ENABLE_ACCEPT */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_ENABLE_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_enable_req(omfp->trx->bts, &omfp->mo->addr);
return;
#if 0
case OM2K_MO_CLS_TF:
/* skip the config, hope that helps speeding things up */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_ENABLE_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_enable_req(omfp->trx->bts, &omfp->mo->addr);
return;
#endif
}
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_CFG_ACCEPT,
OM2K_TIMEOUT, 0);
switch (omfp->mo->addr.class) {
case OM2K_MO_CLS_TF:
abis_om2k_tx_tf_conf_req(omfp->trx->bts);
break;
case OM2K_MO_CLS_IS:
abis_om2k_tx_is_conf_req(omfp->trx->bts);
break;
case OM2K_MO_CLS_CON:
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
abis_om2k_tx_con_conf_req(omfp->trx->bts);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MO_CLS_MCTR:
abis_om2k_tx_mctr_conf_req(omfp->trx->bts);
break;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MO_CLS_TX:
abis_om2k_tx_tx_conf_req(omfp->trx);
break;
case OM2K_MO_CLS_RX:
abis_om2k_tx_rx_conf_req(omfp->trx);
break;
case OM2K_MO_CLS_TS:
ts = mo2obj(omfp->trx->bts, &omfp->mo->addr);
abis_om2k_tx_ts_conf_req(ts);
break;
}
}
static void om2k_mo_st_wait_cfg_accept(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
uint32_t timeout = OM2K_TIMEOUT;
if (omfp->mo->addr.class == OM2K_MO_CLS_TF)
timeout = 600;
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_CFG_RES, timeout, 0);
}
static void om2k_mo_st_wait_cfg_res(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
struct om2k_decoded_msg *omd = data;
uint8_t accordance;
if (!TLVP_PRESENT(&omd->tp, OM2K_DEI_ACCORDANCE_IND)) {
osmo_fsm_inst_state_chg(fi, OM2K_ST_ERROR, 0, 0);
return;
}
accordance = *TLVP_VAL(&omd->tp, OM2K_DEI_ACCORDANCE_IND);
if (accordance != 0) {
/* accordance not OK */
osmo_fsm_inst_state_chg(fi, OM2K_ST_ERROR, 0, 0);
return;
}
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_ENABLE_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_enable_req(omfp->trx->bts, &omfp->mo->addr);
}
static void om2k_mo_st_wait_enable_accept(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
OM2000: Send ALTCRQ for SuperChannel after receiving IS Enable Req Ack When the BTS is configured to use a SuperChannel and it is using a unix domain socket based transport towards the L2TP daemon, then we must instruct the L2TP daemon to instruct the SIU to change the Abis Lower Transport Mode using the ALTCRQ / ALTCRP L2TP signalling. Change-Id: I672bfaa09c42fbeb0c8459f24b2222b952de954b
2016-11-15 21:59:24 +00:00
struct om2k_mo_fsm_priv *omfp = fi->priv;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
struct om2k_decoded_msg *omd = data;
switch (omd->msg_type) {
case OM2K_MSGT_ENABLE_REQ_REJ:
osmo_fsm_inst_state_chg(fi, OM2K_ST_ERROR, 0, 0);
break;
case OM2K_MSGT_ENABLE_REQ_ACK:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
if (omfp->mo->addr.class == OM2K_MO_CLS_IS && omfp->trx->bts->rbs2000.use_superchannel)
OM2000: Send ALTCRQ for SuperChannel after receiving IS Enable Req Ack When the BTS is configured to use a SuperChannel and it is using a unix domain socket based transport towards the L2TP daemon, then we must instruct the L2TP daemon to instruct the SIU to change the Abis Lower Transport Mode using the ALTCRQ / ALTCRP L2TP signalling. Change-Id: I672bfaa09c42fbeb0c8459f24b2222b952de954b
2016-11-15 21:59:24 +00:00
e1inp_ericsson_set_altc(omfp->trx->bts->oml_link->ts->line, 1);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_ENABLE_RES, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
}
static void om2k_mo_st_wait_enable_res(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
//struct om2k_decoded_msg *omd = data;
/* TODO: check if state is actually enabled now? */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_ST_WAIT_OPINFO_ACCEPT, OM2K_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2k_tx_op_info(omfp->trx->bts, &omfp->mo->addr, 1);
}
static void om2k_mo_st_wait_opinfo_accept(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
osmo_fsm_inst_state_chg(fi, OM2K_ST_DONE, 0, 0);
}
static void om2k_mo_s_done_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
abis_om2k: protect MO FSMs by NULL check Also set MO FSMs to NULL after freeing them. Change-Id: I30df0b9ab8bc47ba9756c8388e977deed0e40200
2016-11-02 01:41:41 +00:00
struct om2k_mo_fsm_priv *omfp = fi->priv;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
if (fi->proc.parent)
osmo_fsm_inst_dispatch(fi->proc.parent, omfp->done_event, NULL);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
abis_om2k: protect MO FSMs by NULL check Also set MO FSMs to NULL after freeing them. Change-Id: I30df0b9ab8bc47ba9756c8388e977deed0e40200
2016-11-02 01:41:41 +00:00
static void om2k_mo_s_error_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
struct om2k_mo_fsm_priv *omfp = fi->priv;
omfp->mo->fsm = NULL;
osmo_fsm_inst_term(fi, OSMO_FSM_TERM_ERROR, NULL);
}
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
static void om2k_mo_allstate(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
switch (event) {
case OM2K_MO_EVT_RESET:
osmo_fsm_inst_broadcast_children(fi, event, data);
osmo_fsm_inst_state_chg(fi, OM2K_ST_INIT, 0, 0);
break;
default:
OSMO_ASSERT(0);
}
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
static const struct osmo_fsm_state om2k_is_states[] = {
[OM2K_ST_INIT] = {
.name = "INIT",
.in_event_mask = S(OM2K_MO_EVT_START),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_CONN_COMPL) |
S(OM2K_ST_WAIT_START_ACCEPT) |
S(OM2K_ST_WAIT_RES_COMPL),
.action = om2k_mo_st_init,
},
[OM2K_ST_WAIT_CONN_COMPL] = {
.name = "WAIT-CONN-COMPL",
.in_event_mask = S(OM2K_MO_EVT_RX_CONN_COMPL),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_START_ACCEPT) |
S(OM2K_ST_WAIT_RES_COMPL),
.action = om2k_mo_st_wait_conn_compl,
},
[OM2K_ST_WAIT_RES_COMPL] = {
.name = "WAIT-RES-COMPL",
.in_event_mask = S(OM2K_MO_EVT_RX_RESET_COMPL),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_START_ACCEPT),
.action = om2k_mo_st_wait_res_compl,
},
[OM2K_ST_WAIT_START_ACCEPT] = {
.name = "WAIT-START-ACCEPT",
.in_event_mask = S(OM2K_MO_EVT_RX_START_REQ_ACCEPT),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_START_RES),
.action =om2k_mo_st_wait_start_accept,
},
[OM2K_ST_WAIT_START_RES] = {
.name = "WAIT-START-RES",
.in_event_mask = S(OM2K_MO_EVT_RX_START_RES),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_CFG_ACCEPT) |
OM2K: Permit transition from WAIT_CONF_RES -> WAIT_ENABLE_ACCEPT This is how the DP MO FSM is doing, so let's permit it. Change-Id: I9df44f840a2818ebc877453c74bde127dce9b3e9
2020-12-20 23:10:05 +00:00
S(OM2K_ST_WAIT_OPINFO_ACCEPT) |
S(OM2K_ST_WAIT_ENABLE_ACCEPT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.action = om2k_mo_st_wait_start_res,
},
[OM2K_ST_WAIT_CFG_ACCEPT] = {
.name = "WAIT-CFG-ACCEPT",
.in_event_mask = S(OM2K_MO_EVT_RX_CFG_REQ_ACCEPT),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_CFG_RES),
.action = om2k_mo_st_wait_cfg_accept,
},
[OM2K_ST_WAIT_CFG_RES] = {
.name = "WAIT-CFG-RES",
.in_event_mask = S(OM2K_MO_EVT_RX_CFG_RES),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_ENABLE_ACCEPT),
.action = om2k_mo_st_wait_cfg_res,
},
[OM2K_ST_WAIT_ENABLE_ACCEPT] = {
.name = "WAIT-ENABLE-ACCEPT",
.in_event_mask = S(OM2K_MO_EVT_RX_ENA_REQ_ACCEPT),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_ENABLE_RES),
.action = om2k_mo_st_wait_enable_accept,
},
[OM2K_ST_WAIT_ENABLE_RES] = {
.name = "WAIT-ENABLE-RES",
.in_event_mask = S(OM2K_MO_EVT_RX_ENA_RES),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR) |
S(OM2K_ST_WAIT_OPINFO_ACCEPT),
.action = om2k_mo_st_wait_enable_res,
},
[OM2K_ST_WAIT_OPINFO_ACCEPT] = {
.name = "WAIT-OPINFO-ACCEPT",
.in_event_mask = S(OM2K_MO_EVT_RX_OPINFO_ACC),
.out_state_mask = S(OM2K_ST_DONE) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_ST_INIT) |
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
S(OM2K_ST_ERROR),
.action = om2k_mo_st_wait_opinfo_accept,
},
[OM2K_ST_DONE] = {
.name = "DONE",
.in_event_mask = 0,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_ST_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.onenter = om2k_mo_s_done_onenter,
},
[OM2K_ST_ERROR] = {
.name = "ERROR",
.in_event_mask = 0,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_ST_INIT),
abis_om2k: protect MO FSMs by NULL check Also set MO FSMs to NULL after freeing them. Change-Id: I30df0b9ab8bc47ba9756c8388e977deed0e40200
2016-11-02 01:41:41 +00:00
.onenter = om2k_mo_s_error_onenter,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
},
};
static int om2k_mo_timer_cb(struct osmo_fsm_inst *fi)
{
osmo_fsm_inst_state_chg(fi, OM2K_ST_ERROR, 0, 0);
return 0;
}
static struct osmo_fsm om2k_mo_fsm = {
.name = "OM2000-MO",
.states = om2k_is_states,
.num_states = ARRAY_SIZE(om2k_is_states),
.log_subsys = DNM,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.allstate_event_mask = S(OM2K_MO_EVT_RESET),
.allstate_action = om2k_mo_allstate,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.event_names = om2k_event_names,
.timer_cb = om2k_mo_timer_cb,
};
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
static struct osmo_fsm_inst *om2k_mo_fsm_alloc(struct osmo_fsm_inst *parent, uint32_t done_event,
abis_om2000: Mark om2k_mo_fsm_start() as static It's not used anywhere outside of the source file Change-Id: If9ebf07ac662607826270df0b0966c8dd5026f1c
2020-12-21 10:16:04 +00:00
struct gsm_bts_trx *trx, struct om2k_mo *mo)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct osmo_fsm_inst *fi;
struct om2k_mo_fsm_priv *omfp;
char idbuf[64];
om2k: Don't use slashes in FSM IDs and use dashes instead slashes are invalid so we can't use om2k_mo_name() directly, so we just build it manually with dashes. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I3cfc19670e6d7bb607d796cabcee5e86a15d1985
2020-05-05 18:03:22 +00:00
snprintf(idbuf, sizeof(idbuf), "%s-%s-%02x-%02x-%02x", parent->id,
get_value_string(om2k_mo_class_short_vals, mo->addr.class),
mo->addr.bts, mo->addr.assoc_so, mo->addr.inst);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
fi = osmo_fsm_inst_alloc_child_id(&om2k_mo_fsm, parent, OM2K_MO_EVT_CHILD_TERM, idbuf);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
if (!fi)
return NULL;
mo->fsm = fi;
omfp = talloc_zero(fi, struct om2k_mo_fsm_priv);
omfp->mo = mo;
omfp->trx = trx;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
omfp->done_event = done_event;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
fi->priv = omfp;
return fi;
}
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
static void om2k_mo_fsm_start(struct om2k_mo *mo)
{
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_START, NULL);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
int om2k_mo_fsm_recvmsg(struct gsm_bts *bts, struct om2k_mo *mo,
struct om2k_decoded_msg *odm)
{
switch (odm->msg_type) {
case OM2K_MSGT_CONNECT_COMPL:
case OM2K_MSGT_CONNECT_REJ:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_CONN_COMPL, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_RESET_COMPL:
case OM2K_MSGT_RESET_REJ:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_RESET_COMPL, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_START_REQ_ACK:
case OM2K_MSGT_START_REQ_REJ:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_START_REQ_ACCEPT, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_START_RES:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_START_RES, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_CON_CONF_REQ_ACK:
case OM2K_MSGT_IS_CONF_REQ_ACK:
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MSGT_MCTR_CONF_REQ_ACK:
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MSGT_RX_CONF_REQ_ACK:
case OM2K_MSGT_TF_CONF_REQ_ACK:
case OM2K_MSGT_TS_CONF_REQ_ACK:
case OM2K_MSGT_TX_CONF_REQ_ACK:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_CFG_REQ_ACCEPT, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_CON_CONF_RES:
case OM2K_MSGT_IS_CONF_RES:
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MSGT_MCTR_CONF_RES:
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
case OM2K_MSGT_RX_CONF_RES:
case OM2K_MSGT_TF_CONF_RES:
case OM2K_MSGT_TS_CONF_RES:
case OM2K_MSGT_TX_CONF_RES:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_CFG_RES, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_ENABLE_REQ_ACK:
case OM2K_MSGT_ENABLE_REQ_REJ:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_ENA_REQ_ACCEPT, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_ENABLE_RES:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_ENA_RES, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
case OM2K_MSGT_OP_INFO_ACK:
case OM2K_MSGT_OP_INFO_REJ:
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_dispatch(mo->fsm, OM2K_MO_EVT_RX_OPINFO_ACC, odm);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
break;
default:
return -1;
}
return 0;
}
/***********************************************************************
* OM2000 TRX Finite State Machine, initializes TRXC and all siblings
***********************************************************************/
enum om2k_trx_event {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OM2K_TRX_EVT_RESET = OM2K_MO_EVT_RESET,
OM2K_TRX_EVT_START = OM2K_MO_EVT_START,
OM2K_TRX_EVT_CHILD_TERM = OM2K_MO_EVT_CHILD_TERM,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_TRX_EVT_TRXC_DONE,
OM2K_TRX_EVT_TX_DONE,
OM2K_TRX_EVT_RX_DONE,
OM2K_TRX_EVT_TS_DONE,
OM2K_TRX_EVT_STOP,
};
static struct value_string om2k_trx_events[] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_TRX_EVT_RESET, "RESET" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_TRX_EVT_START, "START" },
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_TRX_EVT_CHILD_TERM, "CHILD-TERM" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_TRX_EVT_TRXC_DONE, "TRXC-DONE" },
{ OM2K_TRX_EVT_TX_DONE, "TX-DONE" },
{ OM2K_TRX_EVT_RX_DONE, "RX-DONE" },
{ OM2K_TRX_EVT_TS_DONE, "TS-DONE" },
{ OM2K_TRX_EVT_STOP, "STOP" },
{ 0, NULL }
};
enum om2k_trx_state {
OM2K_TRX_S_INIT,
OM2K_TRX_S_WAIT_TRXC,
OM2K_TRX_S_WAIT_TX,
OM2K_TRX_S_WAIT_RX,
OM2K_TRX_S_WAIT_TS,
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
OM2K_TRX_S_SEND_SI,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_TRX_S_DONE,
OM2K_TRX_S_ERROR
};
struct om2k_trx_fsm_priv {
struct gsm_bts_trx *trx;
om2k: Dispatch TS_EV_OML_READY to TS FSM only when it's actually ready Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ic2a84ea406e9a39332313d63b27d73f334d4e0a0
2020-05-06 09:49:08 +00:00
uint8_t cur_ts_nr;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
uint32_t done_event;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
};
static void om2k_trx_s_init(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
/* First initialize TRXC */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_WAIT_TRXC, TRX_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&otfp->trx->rbs2000.trxc.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_trx_s_wait_trxc(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
/* Initialize TX after TRXC */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_WAIT_TX, TRX_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&otfp->trx->rbs2000.tx.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_trx_s_wait_tx(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
/* Initialize RX after TX */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_WAIT_RX, TRX_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&otfp->trx->rbs2000.rx.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_trx_s_wait_rx(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
struct gsm_bts_trx_ts *ts;
/* Initialize Timeslots after TX */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_WAIT_TS, TRX_FSM_TIMEOUT, 0);
om2k: Dispatch TS_EV_OML_READY to TS FSM only when it's actually ready Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ic2a84ea406e9a39332313d63b27d73f334d4e0a0
2020-05-06 09:49:08 +00:00
otfp->cur_ts_nr = 0;
ts = &otfp->trx->ts[otfp->cur_ts_nr];
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&ts->rbs2000.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_trx_s_wait_ts(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
struct gsm_bts_trx_ts *ts;
om2k: Dispatch TS_EV_OML_READY to TS FSM only when it's actually ready Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ic2a84ea406e9a39332313d63b27d73f334d4e0a0
2020-05-06 09:49:08 +00:00
/* notify TS is ready */
ts = &otfp->trx->ts[otfp->cur_ts_nr];
osmo_fsm_inst_dispatch(ts->fi, TS_EV_OML_READY, NULL);
/* next ? */
if (++otfp->cur_ts_nr < 8) {
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* iterate to the next timeslot */
om2k: Dispatch TS_EV_OML_READY to TS FSM only when it's actually ready Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ic2a84ea406e9a39332313d63b27d73f334d4e0a0
2020-05-06 09:49:08 +00:00
ts = &otfp->trx->ts[otfp->cur_ts_nr];
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&ts->rbs2000.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
} else {
/* only after all 8 TS */
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_SEND_SI, 0, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
}
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
static void om2k_trx_s_send_si(struct osmo_fsm_inst *fi, uint32_t prev_state)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct om2k_trx_fsm_priv *otfp = fi->priv;
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
if (gsm_bts_trx_set_system_infos(otfp->trx) == 0)
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_DONE, 0, 0);
else
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_ERROR, 0, 0);
}
static void om2k_trx_s_done_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
abis_om2000: Force TRX NM state to UNLOCKED once TRX is up In I68ae0bc51a565f903b47cf72f3e3dd6f1a2d2651 we started to initialize all (TS 12.21) NM MO state to LOCKED by default, which means BTSs not using TS 12.21 are stumbling across a general check in bts_isdn_sign_link() and will fail to process and uplink RSL messages. Change-Id: I4450cc47ede4005bf4bbd4af8d5a22c78377d4cd Related: OSE4914
2020-12-20 22:19:19 +00:00
struct om2k_trx_fsm_priv *otfp = fi->priv;
/* See e1_config:bts_isdn_sign_link() / OS#4914 */
otfp->trx->mo.nm_state.administrative = NM_STATE_UNLOCKED;
abis_om2000: fix missing signal The normal abis nm FSMs are sending S_NM_RUNNING_CHG signals that include an object class to notify other entities about state changes. This also includes paging.c, which only starts paging services when it sees NM_OC_RADIO_CARRIER becoming enabled. Change-Id: I305df5b2f962473e33e32484c42a79ff96e53e1a Fixes: I1b5b1a98115b4e9d821eb3330fc5b970a0e78a44 Related: OS#5634
2022-07-29 11:35:09 +00:00
abis_om2000_fsm_transc_becomes_enabled(otfp->trx);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
if (fi->proc.parent)
osmo_fsm_inst_dispatch(fi->proc.parent, otfp->done_event, NULL);
}
static void om2k_trx_allstate(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
abis_om2000: fix missing signal The normal abis nm FSMs are sending S_NM_RUNNING_CHG signals that include an object class to notify other entities about state changes. This also includes paging.c, which only starts paging services when it sees NM_OC_RADIO_CARRIER becoming enabled. Change-Id: I305df5b2f962473e33e32484c42a79ff96e53e1a Fixes: I1b5b1a98115b4e9d821eb3330fc5b970a0e78a44 Related: OS#5634
2022-07-29 11:35:09 +00:00
struct om2k_trx_fsm_priv *otfp = fi->priv;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
switch (event) {
case OM2K_TRX_EVT_RESET:
abis_om2000: fix missing signal The normal abis nm FSMs are sending S_NM_RUNNING_CHG signals that include an object class to notify other entities about state changes. This also includes paging.c, which only starts paging services when it sees NM_OC_RADIO_CARRIER becoming enabled. Change-Id: I305df5b2f962473e33e32484c42a79ff96e53e1a Fixes: I1b5b1a98115b4e9d821eb3330fc5b970a0e78a44 Related: OS#5634
2022-07-29 11:35:09 +00:00
abis_om2000_fsm_transc_becomes_disabled(otfp->trx);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
osmo_fsm_inst_broadcast_children(fi, event, data);
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_INIT, 0, 0);
break;
default:
OSMO_ASSERT(0);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static const struct osmo_fsm_state om2k_trx_states[] = {
[OM2K_TRX_S_INIT] = {
.in_event_mask = S(OM2K_TRX_EVT_START),
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_TRX_S_WAIT_TRXC) |
S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "INIT",
.action = om2k_trx_s_init,
},
[OM2K_TRX_S_WAIT_TRXC] = {
.in_event_mask = S(OM2K_TRX_EVT_TRXC_DONE),
.out_state_mask = S(OM2K_TRX_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_TRX_S_WAIT_TX) |
S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-TRXC",
.action = om2k_trx_s_wait_trxc,
},
[OM2K_TRX_S_WAIT_TX] = {
.in_event_mask = S(OM2K_TRX_EVT_TX_DONE),
.out_state_mask = S(OM2K_TRX_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_TRX_S_WAIT_RX) |
S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-TX",
.action = om2k_trx_s_wait_tx,
},
[OM2K_TRX_S_WAIT_RX] = {
.in_event_mask = S(OM2K_TRX_EVT_RX_DONE),
.out_state_mask = S(OM2K_TRX_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_TRX_S_WAIT_TS) |
S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-RX",
.action = om2k_trx_s_wait_rx,
},
[OM2K_TRX_S_WAIT_TS] = {
.in_event_mask = S(OM2K_TRX_EVT_TS_DONE),
.out_state_mask = S(OM2K_TRX_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_TRX_S_SEND_SI) |
S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-TS",
.action = om2k_trx_s_wait_ts,
},
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
[OM2K_TRX_S_SEND_SI] = {
.out_state_mask = S(OM2K_TRX_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_TRX_S_DONE) |
S(OM2K_TRX_S_INIT),
abis_om2000: check result of gsm_bts_trx_set_system_infos() Ensure that osmo-bsc would not continue to work as usual, if for some reason we cannot encode or send System Information messages. Introduce transitional state OM2K_TRX_S_SEND_SI, from where we can generate and send System Information messages. Otherwise it's confusing if we fail to do something when we're already in state OM2K_TRX_S_DONE. Change-Id: Ia6df539d0914c57ea80fdb29882832678b47f267
2020-09-03 12:11:36 +00:00
.name = "SEND-SI",
.onenter = om2k_trx_s_send_si,
},
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
[OM2K_TRX_S_DONE] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "DONE",
.onenter = om2k_trx_s_done_onenter,
},
[OM2K_TRX_S_ERROR] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_TRX_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "ERROR",
},
};
static int om2k_trx_timer_cb(struct osmo_fsm_inst *fi)
{
osmo_fsm_inst_state_chg(fi, OM2K_TRX_S_ERROR, 0, 0);
return 0;
}
static struct osmo_fsm om2k_trx_fsm = {
.name = "OM2000-TRX",
.states = om2k_trx_states,
.num_states = ARRAY_SIZE(om2k_trx_states),
.log_subsys = DNM,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.allstate_event_mask = S(OM2K_TRX_EVT_RESET),
.allstate_action = om2k_trx_allstate,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.event_names = om2k_trx_events,
.timer_cb = om2k_trx_timer_cb,
};
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
static struct osmo_fsm_inst *om2k_trx_fsm_alloc(struct osmo_fsm_inst *parent,
struct gsm_bts_trx *trx, uint32_t done_event)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct osmo_fsm_inst *fi;
struct om2k_trx_fsm_priv *otfp;
char idbuf[32];
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OSMO_ASSERT(!trx->rbs2000.trx_fi);
om2k: Don't use slashes in FSM IDs and use dashes instead slashes are invalid so we can't use om2k_mo_name() directly, so we just build it manually with dashes. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I3cfc19670e6d7bb607d796cabcee5e86a15d1985
2020-05-05 18:03:22 +00:00
snprintf(idbuf, sizeof(idbuf), "%u-%u", trx->bts->nr, trx->nr);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
fi = osmo_fsm_inst_alloc_child_id(&om2k_trx_fsm, parent, OM2K_MO_EVT_CHILD_TERM, idbuf);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
if (!fi)
return NULL;
otfp = talloc_zero(fi, struct om2k_trx_fsm_priv);
otfp->trx = trx;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
otfp->done_event = done_event;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
fi->priv = otfp;
return fi;
}
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
void om2k_trx_fsm_start(struct gsm_bts_trx *trx)
{
struct osmo_fsm_inst *bts_fi = trx->bts->rbs2000.bts_fi;
OSMO_ASSERT(trx->rbs2000.trx_fi);
/* suppress if BTS is not yet brought up */
if (bts_fi->state == OM2K_BTS_S_DONE || bts_fi->state == OM2K_BTS_S_WAIT_TRX)
return;
osmo_fsm_inst_dispatch(trx->rbs2000.trx_fi, OM2K_TRX_EVT_START, NULL);
}
void om2k_trx_fsm_reset(struct gsm_bts_trx *trx)
{
struct osmo_fsm_inst *bts_fi = trx->bts->rbs2000.bts_fi;
OSMO_ASSERT(trx->rbs2000.trx_fi);
OSMO_ASSERT(trx->rbs2000.trx_fi);
/* suppress if BTS is not yet brought up */
if (bts_fi->state == OM2K_BTS_S_DONE || bts_fi->state == OM2K_BTS_S_WAIT_TRX)
return;
osmo_fsm_inst_dispatch(trx->rbs2000.trx_fi, OM2K_TRX_EVT_RESET, NULL);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/***********************************************************************
* OM2000 BTS Finite State Machine, initializes CF and all siblings
***********************************************************************/
enum om2k_bts_event {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OM2K_BTS_EVT_RESET = OM2K_MO_EVT_RESET,
OM2K_BTS_EVT_START = OM2K_MO_EVT_START,
OM2K_BTS_EVT_CHILD_TERM = OM2K_MO_EVT_CHILD_TERM,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_BTS_EVT_CF_DONE,
OM2K_BTS_EVT_IS_DONE,
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
OM2K_BTS_EVT_CON_DONE,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_BTS_EVT_TF_DONE,
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
OM2K_BTS_EVT_MCTR_DONE,
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
OM2K_BTS_EVT_TRX_LAPD_UP,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_BTS_EVT_TRX_DONE,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OM2K_BTS_EVT_TRX_TERM,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2K_BTS_EVT_STOP,
};
static const struct value_string om2k_bts_events[] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_BTS_EVT_RESET, "RESET" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_BTS_EVT_START, "START" },
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
{ OM2K_BTS_EVT_CHILD_TERM, "CHILD-TERM" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_BTS_EVT_CF_DONE, "CF-DONE" },
{ OM2K_BTS_EVT_IS_DONE, "IS-DONE" },
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
{ OM2K_BTS_EVT_CON_DONE, "CON-DONE" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_BTS_EVT_TF_DONE, "TF-DONE" },
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
{ OM2K_BTS_EVT_MCTR_DONE, "MCTR-DONE" },
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
{ OM2K_BTS_EVT_TRX_LAPD_UP, "TRX-LAPD-UP" },
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{ OM2K_BTS_EVT_TRX_DONE, "TRX-DONE" },
{ OM2K_BTS_EVT_STOP, "STOP" },
{ 0, NULL }
};
struct om2k_bts_fsm_priv {
struct gsm_bts *bts;
uint8_t next_trx_nr;
};
static void om2k_bts_s_init(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
struct gsm_bts *bts = obfp->bts;
OSMO_ASSERT(event == OM2K_BTS_EVT_START);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_CF, BTS_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.cf.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_bts_s_wait_cf(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
struct gsm_bts *bts = obfp->bts;
OSMO_ASSERT(event == OM2K_BTS_EVT_CF_DONE);
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
/* TF can take a long time to initialize, wait for 10min */
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_TF, 600, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.tf.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
static void om2k_bts_s_wait_tf(struct osmo_fsm_inst *fi, uint32_t event, void *data)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
struct gsm_bts *bts = obfp->bts;
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
OSMO_ASSERT(event == OM2K_BTS_EVT_TF_DONE);
OM2K: Skip the entire CON MO if there are no connection groups If the user doesn't specify any CON connection groups in the config file, then the CON is not used. The current code runs into an error condition, as abis_om2k_tx_con_conf_req() never sends the CON CONF REQ if the groups list is empty, but we still wait in the FSM for the arrival of a CFG REQ ACCEPT. The CON FSM eventually times out in T10 and we proceed with the IS, ignoring the error. With this patch, we simply skip the entire CON MO in case there is no related configuration. Change-Id: Ia4d5bd96734686381f04aa3b380b17a161a31174
2020-12-20 23:02:54 +00:00
if (!llist_count(&bts->rbs2000.con.conn_groups)) {
/* skip CON object if we have no configuration for it */
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_IS, BTS_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.is.om2k_mo);
OM2K: Skip the entire CON MO if there are no connection groups If the user doesn't specify any CON connection groups in the config file, then the CON is not used. The current code runs into an error condition, as abis_om2k_tx_con_conf_req() never sends the CON CONF REQ if the groups list is empty, but we still wait in the FSM for the arrival of a CFG REQ ACCEPT. The CON FSM eventually times out in T10 and we proceed with the IS, ignoring the error. With this patch, we simply skip the entire CON MO in case there is no related configuration. Change-Id: Ia4d5bd96734686381f04aa3b380b17a161a31174
2020-12-20 23:02:54 +00:00
} else {
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_CON, BTS_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.con.om2k_mo);
OM2K: Skip the entire CON MO if there are no connection groups If the user doesn't specify any CON connection groups in the config file, then the CON is not used. The current code runs into an error condition, as abis_om2k_tx_con_conf_req() never sends the CON CONF REQ if the groups list is empty, but we still wait in the FSM for the arrival of a CFG REQ ACCEPT. The CON FSM eventually times out in T10 and we proceed with the IS, ignoring the error. With this patch, we simply skip the entire CON MO in case there is no related configuration. Change-Id: Ia4d5bd96734686381f04aa3b380b17a161a31174
2020-12-20 23:02:54 +00:00
}
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
}
static void om2k_bts_s_wait_con(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
struct gsm_bts *bts = obfp->bts;
OSMO_ASSERT(event == OM2K_BTS_EVT_CON_DONE);
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_IS, BTS_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.is.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
static void om2k_bts_s_wait_is(struct osmo_fsm_inst *fi, uint32_t event, void *data)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
struct gsm_bts *bts = obfp->bts;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
OSMO_ASSERT(event == OM2K_BTS_EVT_IS_DONE);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
/* If we're running OML >= G12R13, start MCTR, else skip directly to TRX */
if (bts->rbs2000.om2k_version[0].active >= 0x0c0d) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_MCTR, BTS_FSM_TIMEOUT, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_start(&bts->rbs2000.mctr.om2k_mo);
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
} else {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_TRX_LAPD, TRX_LAPD_TIMEOUT, 0);
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
}
}
static void om2k_bts_s_wait_mctr(struct osmo_fsm_inst *fi, uint32_t event, void *data)
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
{
OSMO_ASSERT(event == OM2K_BTS_EVT_MCTR_DONE);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_TRX_LAPD, TRX_LAPD_TIMEOUT, 0);
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
}
static void om2k_bts_s_wait_trx_lapd(struct osmo_fsm_inst *fi, uint32_t event, void *data)
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
struct gsm_bts_trx *trx;
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
OSMO_ASSERT(event == OM2K_BTS_EVT_TRX_LAPD_UP);
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_WAIT_TRX, BTS_FSM_TIMEOUT, 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
obfp->next_trx_nr = 0;
trx = gsm_bts_trx_num(obfp->bts, obfp->next_trx_nr++);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_trx_fsm_start(trx);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static void om2k_bts_s_wait_trx(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
struct om2k_bts_fsm_priv *obfp = fi->priv;
OSMO_ASSERT(event == OM2K_BTS_EVT_TRX_DONE);
if (obfp->next_trx_nr < obfp->bts->num_trx) {
struct gsm_bts_trx *trx;
trx = gsm_bts_trx_num(obfp->bts, obfp->next_trx_nr++);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_trx_fsm_start(trx);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
} else {
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_DONE, 0, 0);
}
}
static void om2k_bts_s_done_onenter(struct osmo_fsm_inst *fi, uint32_t prev_state)
{
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
}
static void om2k_bts_allstate(struct osmo_fsm_inst *fi, uint32_t event, void *data)
{
switch (event) {
case OM2K_BTS_EVT_RESET:
osmo_fsm_inst_broadcast_children(fi, event, data);
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_INIT, 0, 0);
break;
default:
OSMO_ASSERT(0);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static const struct osmo_fsm_state om2k_bts_states[] = {
[OM2K_BTS_S_INIT] = {
.in_event_mask = S(OM2K_BTS_EVT_START),
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_BTS_S_WAIT_CF) |
S(OM2K_BTS_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "INIT",
.action = om2k_bts_s_init,
},
[OM2K_BTS_S_WAIT_CF] = {
.in_event_mask = S(OM2K_BTS_EVT_CF_DONE),
.out_state_mask = S(OM2K_BTS_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_WAIT_TF) |
S(OM2K_BTS_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-CF",
.action = om2k_bts_s_wait_cf,
},
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
[OM2K_BTS_S_WAIT_TF] = {
.in_event_mask = S(OM2K_BTS_EVT_TF_DONE),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.out_state_mask = S(OM2K_BTS_S_ERROR) |
OM2K: Skip the entire CON MO if there are no connection groups If the user doesn't specify any CON connection groups in the config file, then the CON is not used. The current code runs into an error condition, as abis_om2k_tx_con_conf_req() never sends the CON CONF REQ if the groups list is empty, but we still wait in the FSM for the arrival of a CFG REQ ACCEPT. The CON FSM eventually times out in T10 and we proceed with the IS, ignoring the error. With this patch, we simply skip the entire CON MO in case there is no related configuration. Change-Id: Ia4d5bd96734686381f04aa3b380b17a161a31174
2020-12-20 23:02:54 +00:00
S(OM2K_BTS_S_WAIT_CON) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_WAIT_IS) |
S(OM2K_BTS_S_INIT),
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
.name = "WAIT-TF",
.action = om2k_bts_s_wait_tf,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
},
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
[OM2K_BTS_S_WAIT_CON] = {
.in_event_mask = S(OM2K_BTS_EVT_CON_DONE),
.out_state_mask = S(OM2K_BTS_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_WAIT_IS) |
S(OM2K_BTS_S_INIT),
Support configuration of CON MO Groups/Paths from VTY The code for supporting the configuration of the OM2000 CON (LAPD Concentrator) MO was so far incomplete and not used from the OM2000 FSM initialization. This patch adds * VTY commands for configuration of CON Groups and Paths * The FSM integration to actually configure the CON MO Change-Id: I56dc1b5e35adef3a2078bcf9536537eb0f454192
2016-11-11 18:41:59 +00:00
.name = "WAIT-CON",
.action = om2k_bts_s_wait_con,
},
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
[OM2K_BTS_S_WAIT_IS] = {
.in_event_mask = S(OM2K_BTS_EVT_IS_DONE),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.out_state_mask = S(OM2K_BTS_S_ERROR) |
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
S(OM2K_BTS_S_WAIT_MCTR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_WAIT_TRX_LAPD) |
S(OM2K_BTS_S_INIT),
OM2000: Change the order of MO initialization So far: CF-IS-CON-TF Now: CF-TF-CON-IS Change-Id: I8efd9bafdcf9504d2e5fc85c44c708fa53f4dff8
2017-03-13 08:39:08 +00:00
.name = "WAIT-IS",
.action = om2k_bts_s_wait_is,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
},
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
[OM2K_BTS_S_WAIT_MCTR] = {
.in_event_mask = S(OM2K_BTS_EVT_MCTR_DONE),
.out_state_mask = S(OM2K_BTS_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_WAIT_TRX_LAPD) |
S(OM2K_BTS_S_INIT),
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
.name = "WAIT-MCTR",
.action = om2k_bts_s_wait_mctr,
},
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
[OM2K_BTS_S_WAIT_TRX_LAPD] = {
.in_event_mask = S(OM2K_BTS_EVT_TRX_LAPD_UP),
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_BTS_S_WAIT_TRX) |
S(OM2K_BTS_S_INIT),
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
.name = "WAIT-TRX-LAPD",
.action = om2k_bts_s_wait_trx_lapd,
},
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
[OM2K_BTS_S_WAIT_TRX] = {
.in_event_mask = S(OM2K_BTS_EVT_TRX_DONE),
.out_state_mask = S(OM2K_BTS_S_ERROR) |
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
S(OM2K_BTS_S_DONE) |
S(OM2K_BTS_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "WAIT-TRX",
.action = om2k_bts_s_wait_trx,
},
[OM2K_BTS_S_DONE] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_BTS_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "DONE",
.onenter = om2k_bts_s_done_onenter,
},
[OM2K_BTS_S_ERROR] = {
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.out_state_mask = S(OM2K_BTS_S_INIT),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.name = "ERROR",
},
};
static int om2k_bts_timer_cb(struct osmo_fsm_inst *fi)
{
om2k: Wait for OM TRX links to stabilize before trying to bring up TRX The OM2K link "per-trx" only comes up after MCTR is setup. So that means we need to wait for it before trying to boot the TRX itself. He we simply apply a "dumb" 5 sec timeout as this is the most reliable way I found to get this working reliably. Tracking the link state proved difficult and unreliable: - Multiple TRX can be present with their link coming up in random order. - They can already be up at the start (BTS already initialized from a previous boot) and so the link may actually come up, down, and up again. - All of theses transitions might happens before/after we get to the OM2K_BTS_S_WAIT_TRX state depending on how the LAPD timeout expire, if the BTS config was actually changed or not and how much time it takes to apply the new config. So all in all, what we must do is wait for the link to stabilize ... hence just waiting 5 second. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I55a06e08b9c52ff6e97e8c72f2d55770809eba51
2020-05-08 09:50:46 +00:00
switch (fi->state) {
case OM2K_BTS_S_WAIT_TRX_LAPD:
osmo_fsm_inst_dispatch(fi, OM2K_BTS_EVT_TRX_LAPD_UP, NULL);
break;
default:
osmo_fsm_inst_state_chg(fi, OM2K_BTS_S_ERROR, 0, 0);
break;
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
return 0;
}
static struct osmo_fsm om2k_bts_fsm = {
.name = "OM2000-BTS",
.states = om2k_bts_states,
.num_states = ARRAY_SIZE(om2k_bts_states),
.log_subsys = DNM,
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
.allstate_event_mask = S(OM2K_BTS_EVT_RESET),
.allstate_action = om2k_bts_allstate,
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
.event_names = om2k_bts_events,
.timer_cb = om2k_bts_timer_cb,
};
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
static struct osmo_fsm_inst *
om2k_bts_fsm_alloc(struct gsm_bts *bts)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
struct osmo_fsm_inst *fi;
struct om2k_bts_fsm_priv *obfp;
char idbuf[16];
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
OSMO_ASSERT(!bts->rbs2000.bts_fi);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
snprintf(idbuf, sizeof(idbuf), "%u", bts->nr);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
fi = osmo_fsm_inst_alloc(&om2k_bts_fsm, bts, NULL, LOGL_DEBUG, idbuf);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
if (!fi)
return NULL;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
fi->priv = obfp = talloc_zero(fi, struct om2k_bts_fsm_priv);
obfp->bts = bts;
return fi;
}
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
void om2k_bts_fsm_start(struct gsm_bts *bts)
{
OSMO_ASSERT(bts->rbs2000.bts_fi);
osmo_fsm_inst_dispatch(bts->rbs2000.bts_fi, OM2K_BTS_EVT_START, NULL);
}
void om2k_bts_fsm_reset(struct gsm_bts *bts)
{
OSMO_ASSERT(bts->rbs2000.bts_fi);
osmo_fsm_inst_dispatch(bts->rbs2000.bts_fi, OM2K_BTS_EVT_RESET, NULL);
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/***********************************************************************
* OM2000 Negotiation
***********************************************************************/
OM2000: Add VTY commands for connect/disconnect and op_info
2011-02-12 13:57:17 +00:00
static int abis_om2k_tx_negot_req_ack(struct gsm_bts *bts, const struct abis_om2k_mo *mo,
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
uint8_t *data, unsigned int len)
{
struct msgb *msg = om2k_msgb_alloc();
struct abis_om2k_hdr *o2k;
o2k = (struct abis_om2k_hdr *) msgb_put(msg, sizeof(*o2k));
OM2000: Add channel combination specific DIEs for TS config request ... and in order to do this reasonably well, we have to get rid of the pre-computed message length field in the OML header. Instead, we now simply compute it at abis_om2k_sendmsg() time based on the msgb_l2len().
2011-03-05 18:21:26 +00:00
fill_om2k_hdr(o2k, mo, OM2K_MSGT_NEGOT_REQ_ACK);
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
msgb_tlv_put(msg, OM2K_DEI_NEGOT_REC2, len, data);
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
DEBUGP(DNM, "Tx MO=%s %s\n", abis_om2k_mo_name(mo),
OM2000: Negotiation Resposne, send START on OML-up event
2011-02-12 12:44:14 +00:00
get_value_string(om2k_msgcode_vals, OM2K_MSGT_NEGOT_REQ_ACK));
return abis_om2k_sendmsg(bts, msg);
}
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
struct iwd_version {
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
char gen_char[3+1];
char rev_char[3+1];
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
};
struct iwd_type {
uint8_t num_vers;
struct iwd_version v[8];
};
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
static int om2k_rx_negot_req(struct msgb *msg)
{
src: use new msg->dst pointer instead of deprecated msg->trx This patch modifies openBSC code to use msg->dst which stores the pointer to the signalling link structure instead of the pointer to the transceiver structure. This patch prepares the introduction of libosmo-abis.
2011-08-17 20:43:54 +00:00
struct e1inp_sign_link *sign_link = (struct e1inp_sign_link *)msg->dst;
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
struct gsm_bts *bts = sign_link->trx->bts;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
struct abis_om2k_hdr *o2h = msgb_l2(msg);
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
struct iwd_type iwd_types[16];
uint8_t num_iwd_types = o2h->data[2];
uint8_t *cur = o2h->data+3;
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
unsigned int i;
int v;
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
uint8_t out_buf[1024];
uint8_t *out_cur = out_buf+1;
uint8_t out_num_types = 0;
memset(iwd_types, 0, sizeof(iwd_types));
/* Parse the RBS-supported IWD versions into iwd_types array */
for (i = 0; i < num_iwd_types; i++) {
uint8_t num_versions = *cur++;
uint8_t iwd_type = *cur++;
iwd_types[iwd_type].num_vers = num_versions;
for (v = 0; v < num_versions; v++) {
struct iwd_version *iwd_v = &iwd_types[iwd_type].v[v];
memcpy(iwd_v->gen_char, cur, 3);
OM2000: Fix Negotiation parsing
2011-02-12 17:13:37 +00:00
cur += 3;
memcpy(iwd_v->rev_char, cur, 3);
cur += 3;
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
DEBUGP(DNM, "\tIWD Type %u Gen %s Rev %s\n", iwd_type,
iwd_v->gen_char, iwd_v->rev_char);
}
}
/* Select the last version for each IWD type */
for (i = 0; i < ARRAY_SIZE(iwd_types); i++) {
struct iwd_type *type = &iwd_types[i];
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
struct iwd_version *sel_v = NULL, *alt_v = NULL;
uint16_t sel_ver, alt_ver = 0;
int gen, rev;
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
if (type->num_vers == 0)
continue;
out_num_types++;
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
for (v = type->num_vers-1; v >= 0; v--) {
if ((sscanf(type->v[v].gen_char, "G%2d", &gen) != 1) ||
(sscanf(type->v[v].rev_char, "R%2d", &rev) != 1))
continue;
sel_ver = (gen << 8) | rev;
if (!alt_v) {
alt_ver = sel_ver;
alt_v = &type->v[v];
}
if ((bts->rbs2000.om2k_version[i].limit != 0) &&
(bts->rbs2000.om2k_version[i].limit < sel_ver))
continue;
sel_v = &type->v[v];
break;
}
if (!sel_v) {
if (!alt_v) {
LOGP(DNM, LOGL_ERROR, "Couldn't find valid version for IWD Type %u."
"Skipping IWD ... this will most likely fail\n", i);
continue;
} else {
sel_v = alt_v;
sel_ver = alt_ver;
LOGP(DNM, LOGL_ERROR, "Couldn't find suitable version for IWD Type %u."
"Fallback to Gen %s Rev %s\n", i,
sel_v->gen_char, sel_v->rev_char);
}
}
bts->rbs2000.om2k_version[i].active = sel_ver;
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
*out_cur++ = i;
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
memcpy(out_cur, sel_v->gen_char, 3);
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
out_cur += 3;
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
memcpy(out_cur, sel_v->rev_char, 3);
OM2000: Implement proper parsing and response for Negotiation
2011-02-12 17:11:16 +00:00
out_cur += 3;
}
out_buf[0] = out_num_types;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
om2k: Add option to limit OML version during negotiation Starting from G12R13 the MCTR swiches to BSC controlled mode. And although we think we know how to configure it (via MCTR Conf Req), something doesn't work right and the timeslot configuration is not accepted. (TS Conf Result shows "Data not according to request"). So as a workaround for now, we use this version of the protocol where we don't configure the MCTR (it's in "BTS controlled mode") and with this protocol, the BTS accepts our timeslot config and we can bring the system up. This commit add a generic option to limit either OML or RSL IWD version to any value. It also keeps track of the actual negotation version so we can react to it in other places of the code. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8f0b0ba72056ea4250fe490e7a38630c77c04f65 better version limit Change-Id: Ia789f8ede3eab7eeca6c759da0109e0b53398f60
2020-05-05 19:07:58 +00:00
return abis_om2k_tx_negot_req_ack(bts, &o2h->mo, out_buf, out_cur - out_buf);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
}
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/***********************************************************************
* OM2000 Receive Message Handler
***********************************************************************/
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
static int om2k_rx_nack(struct msgb *msg)
{
struct abis_om2k_hdr *o2h = msgb_l2(msg);
uint16_t msg_type = ntohs(o2h->msg_type);
struct tlv_parsed tp;
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
LOGP(DNM, LOGL_ERROR, "Rx MO=%s %s", abis_om2k_mo_name(&o2h->mo),
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
get_value_string(om2k_msgcode_vals, msg_type));
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
abis_om2k_msg_tlv_parse(&tp, o2h);
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
if (TLVP_PRESENT(&tp, OM2K_DEI_REASON_CODE))
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGPC(DNM, LOGL_ERROR, ", Reason 0x%02x", *TLVP_VAL(&tp, OM2K_DEI_REASON_CODE));
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
if (TLVP_PRESENT(&tp, OM2K_DEI_RESULT_CODE))
LOGPC(DNM, LOGL_ERROR, ", Result %s",
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
get_value_string(om2k_result_strings, *TLVP_VAL(&tp, OM2K_DEI_RESULT_CODE)));
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
LOGPC(DNM, LOGL_ERROR, "\n");
return 0;
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
static int process_mo_state(struct gsm_bts *bts, struct om2k_decoded_msg *odm)
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
{
uint8_t mo_state;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
if (!TLVP_PRESENT(&odm->tp, OM2K_DEI_MO_STATE))
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
return -EIO;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
mo_state = *TLVP_VAL(&odm->tp, OM2K_DEI_MO_STATE);
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_DEBUG, "Rx MO=%s %s, MO State: %s\n", abis_om2k_mo_name(&odm->o2h.mo),
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
get_value_string(om2k_msgcode_vals, odm->msg_type),
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
get_value_string(om2k_mostate_vals, mo_state));
OM2000: Throw error when MO can not be enabled Throw warning message in case the MO state does not change to enabled after sendeing an Enable-Request message. Change-Id: Idfde8d6f71526e8acfea51835732515a4bee858e
2016-10-20 11:55:21 +00:00
/* Throw error message in case we see an enable rsponse that does
* not yield an enabled mo-state */
if (odm->msg_type == OM2K_MSGT_ENABLE_RES
&& mo_state != OM2K_MO_S_ENABLED) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "Rx MO=%s %s Failed to enable MO State!\n",
abis_om2k_mo_name(&odm->o2h.mo), get_value_string(om2k_msgcode_vals, odm->msg_type));
OM2000: Throw error when MO can not be enabled Throw warning message in case the MO state does not change to enabled after sendeing an Enable-Request message. Change-Id: Idfde8d6f71526e8acfea51835732515a4bee858e
2016-10-20 11:55:21 +00:00
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
update_mo_state(bts, &odm->o2h.mo, mo_state);
OM2000: Track the Operational Info and MO state
2011-03-06 20:20:38 +00:00
OM2000: Process MO STATE and ACCORDANCE INFO
2011-03-06 18:26:11 +00:00
return 0;
}
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
/* Display fault report bits (helper function of display_fault_maps()) */
OM2000: Fixup based on Coverity Scan suggestion This commit fixes Coverity Scan defect: CID 151901: Insecure data handling (TAINTED_SCALAR) Passing tainted variable "tag_len" to a tainted sink. Change-Id: Ic71ed6a3bbb228bc03e95bfc4a6f5fe09cf5a021
2016-11-15 18:27:20 +00:00
static bool display_fault_bits(const uint8_t *vect, uint16_t len,
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
uint8_t dei, const struct abis_om2k_mo *mo)
{
OM2000: Fixup based on Coverity Scan suggestion This commit fixes Coverity Scan defect: CID 151901: Insecure data handling (TAINTED_SCALAR) Passing tainted variable "tag_len" to a tainted sink. Change-Id: Ic71ed6a3bbb228bc03e95bfc4a6f5fe09cf5a021
2016-11-15 18:27:20 +00:00
uint16_t i;
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
int k;
bool faults_present = false;
int first = 1;
char string[255];
/* Check if errors are present at all */
for (i = 0; i < len; i++)
if (vect[i])
faults_present = true;
if (!faults_present)
return false;
sprintf(string, "Fault Report: %s (",
get_value_string(om2k_attr_vals, dei));
for (i = 0; i < len; i++) {
for (k = 0; k < 8; k++) {
if ((vect[i] >> k) & 1) {
if (!first)
sprintf(string + strlen(string), ",");
sprintf(string + strlen(string), "%d", k + i*8);
first = 0;
}
}
}
sprintf(string + strlen(string), ")\n");
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
DEBUGP(DNM, "Rx MO=%s %s", abis_om2k_mo_name(mo), string);
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
return true;
}
/* Display fault report maps */
static void display_fault_maps(const uint8_t *src, unsigned int src_len,
const struct abis_om2k_mo *mo)
{
uint8_t tag;
uint16_t tag_len;
const uint8_t *val;
int src_pos = 0;
int rc;
int tlv_count = 0;
uint16_t msg_code;
bool faults_present = false;
/* Chop off header */
src+=4;
src_len-=4;
/* Check message type */
msg_code = (*src & 0xff) << 8;
src++;
src_len--;
msg_code |= (*src & 0xff);
src++;
src_len--;
if (msg_code != OM2K_MSGT_FAULT_REP) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "Rx MO=%s Fault report: invalid message code!\n", abis_om2k_mo_name(mo));
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
return;
}
/* Chop off mo-interface */
src += 4;
src_len -= 4;
/* Iterate over each TLV element */
while (1) {
/* Bail if an the maximum number of TLV fields
* have been parsed */
Improve parsing of om2000 fault reports Fault Reports are commonly oberved with a TLV id 0xd2 as are reports with up to 20 TLVs. Let's not have these cause logging at level ERROR. Closes: OS#5593 Change-Id: Ibe0b38835362c59d1576a206b2f64cea4427295f
2022-06-28 02:22:06 +00:00
if (tlv_count >= 20) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "Rx MO=%s Fault Report: too many tlv elements!\n",
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
abis_om2k_mo_name(mo));
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
return;
}
/* Parse TLV field */
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
rc = tlv_parse_one(&tag, &tag_len, &val, &om2k_att_tlvdef, src + src_pos, src_len - src_pos);
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
if (rc > 0)
src_pos += rc;
else {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "Rx MO=%s Fault Report: invalid tlv element!\n",
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
abis_om2k_mo_name(mo));
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
return;
}
switch (tag) {
case OM2K_DEI_INT_FAULT_MAP_1A:
case OM2K_DEI_INT_FAULT_MAP_1B:
case OM2K_DEI_INT_FAULT_MAP_2A:
case OM2K_DEI_EXT_COND_MAP_1:
case OM2K_DEI_EXT_COND_MAP_2:
case OM2K_DEI_REPL_UNIT_MAP:
case OM2K_DEI_INT_FAULT_MAP_2A_EXT:
case OM2K_DEI_EXT_COND_MAP_2_EXT:
case OM2K_DEI_REPL_UNIT_MAP_EXT:
faults_present |= display_fault_bits(val, tag_len,
tag, mo);
break;
}
/* Stop when no further TLV elements can be expected */
if (src_len - src_pos < 2)
break;
tlv_count++;
}
if (!faults_present) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
DEBUGP(DNM, "Rx MO=%s Fault Report: All faults ceased!\n", abis_om2k_mo_name(mo));
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
}
}
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
int abis_om2k_rcvmsg(struct msgb *msg)
{
src: use new msg->dst pointer instead of deprecated msg->trx This patch modifies openBSC code to use msg->dst which stores the pointer to the signalling link structure instead of the pointer to the transceiver structure. This patch prepares the introduction of libosmo-abis.
2011-08-17 20:43:54 +00:00
struct e1inp_sign_link *sign_link = (struct e1inp_sign_link *)msg->dst;
struct gsm_bts *bts = sign_link->trx->bts;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
struct abis_om2k_hdr *o2h = msgb_l2(msg);
struct abis_om_hdr *oh = &o2h->om;
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
uint16_t msg_type = ntohs(o2h->msg_type);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
struct om2k_decoded_msg odm;
struct om2k_mo *mo;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
int rc = 0;
/* Various consistency checks */
if (oh->placement != ABIS_OM_PLACEMENT_ONLY) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "ABIS OML placement 0x%x not supported\n", oh->placement);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
if (oh->placement != ABIS_OM_PLACEMENT_FIRST)
return -EINVAL;
}
if (oh->sequence != 0) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "ABIS OML sequence 0x%x != 0x00\n", oh->sequence);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
return -EINVAL;
}
msg->l3h = (unsigned char *)o2h + sizeof(*o2h);
if (oh->mdisc != ABIS_OM_MDISC_FOM) {
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
LOGP(DNM, LOGL_ERROR, "unknown ABIS OM2000 message discriminator 0x%x\n", oh->mdisc);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
return -EINVAL;
}
abis_om2000: make om2k_mo_name() an exported function Change-Id: Idb05bcad8059ab2b2be6c7057495d0279a4b62c7
2020-12-21 09:29:08 +00:00
DEBUGP(DNM, "Rx MO=%s %s (%s)\n", abis_om2k_mo_name(&o2h->mo),
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
get_value_string(om2k_msgcode_vals, msg_type), osmo_hexdump(msg->l2h, msgb_l2len(msg)));
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
om2k_decode_msg(&odm, msg);
process_mo_state(bts, &odm);
OM2000: Fix computation of length field in header, direct transmit of msgs
2011-02-12 12:09:38 +00:00
switch (msg_type) {
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
case OM2K_MSGT_CAL_TIME_REQ:
rc = abis_om2k_cal_time_resp(bts);
om2000: Don't print "should not generate any message" on CAL_TIME_REQ Change-Id: Ic6f576250d56707cf325886d70209a3744163d79 Closes: OS#4670
2022-01-02 14:12:12 +00:00
/* we receive this from MOs without FSM (https://osmocom.org/issues/4670) */
goto no_mo;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
case OM2K_MSGT_FAULT_REP:
OM2000: Add fault report parsing This patch adds parsing for OM2000 MO fault report map parsing, the bits in the fault maps are counted out and displayed. Change-Id: I6e2928f39b09bc08e9ab78bc10bc81e07f7eb55d
2016-10-19 08:14:35 +00:00
display_fault_maps(msg->l2h, msgb_l2len(msg), &o2h->mo);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_FAULT_REP_ACK);
om2000: Don't print "should not generate any message" on FAULT_REP Change-Id: I65184ef5aa90a993491f9393614bb499c6998169 Closes: OS#4643
2022-01-02 14:14:35 +00:00
/* we receive this from MOs without FSM (https://osmocom.org/issues/4643) */
goto no_mo;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
case OM2K_MSGT_NEGOT_REQ:
rc = om2k_rx_negot_req(msg);
break;
case OM2K_MSGT_START_RES:
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* common processing here */
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_START_RES_ACK);
/* below we dispatch into MO */
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
break;
case OM2K_MSGT_IS_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_IS_CONF_RES_ACK);
break;
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
case OM2K_MSGT_CON_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_CON_CONF_RES_ACK);
break;
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MSGT_MCTR_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_MCTR_CONF_RES_ACK);
break;
OM2000: Add support for sending TX, RX and TS configuration requests They can be triggered from the VTY
2011-03-05 13:13:14 +00:00
case OM2K_MSGT_TX_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_TX_CONF_RES_ACK);
break;
case OM2K_MSGT_RX_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_RX_CONF_RES_ACK);
break;
case OM2K_MSGT_TS_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_TS_CONF_RES_ACK);
break;
OM2000: Add support for configuring the TF (Timing Function)
2011-03-05 13:36:47 +00:00
case OM2K_MSGT_TF_CONF_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_TF_CONF_RES_ACK);
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
break;
OM2000: Make sure we ACK an ENABLE RESULT
2011-02-13 11:43:44 +00:00
case OM2K_MSGT_ENABLE_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_ENABLE_RES_ACK);
break;
OM2000: Handle various ACKs better, implement CON CONF REQ
2011-02-14 14:31:43 +00:00
case OM2K_MSGT_DISABLE_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_DISABLE_RES_ACK);
break;
case OM2K_MSGT_TEST_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_TEST_RES_ACK);
break;
om2k: Acknowledge the HW Infos Reports Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Ida92c9ca59a608854737ba9e330a5394cceff0fe
2020-05-05 18:16:16 +00:00
case OM2K_MSGT_CAPA_HW_INFOS_REP:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_CAPA_HW_INFOS_REP_ACK);
break;
om2000: Add support for querying RBS capabilities Change-Id: Id1fbaf41286f3651ce8f210eb8da05fb51179c96
2016-10-15 19:24:57 +00:00
case OM2K_MSGT_CAPA_RES:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_CAPA_RES_ACK);
break;
om2k: Properly name message 0x0136, found to be MCTR Statistics Report Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I4b4389fd2e7bf0b3078b8ff60f6ea109114a4475
2020-05-06 05:25:25 +00:00
case OM2K_MSGT_MCTR_STATS_REP:
rc = abis_om2k_tx_simple(bts, &o2h->mo, OM2K_MSGT_MCTR_STATS_REP_ACK);
om2k: Acknowledge the unknown MCTR messages we get from time to time This is probably a fault report of some kind, but didn't get any confirmation or naming from anywhere for it yet. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8333093d09f27f61094c7f5854573aa16e4bf28c
2020-05-05 18:45:03 +00:00
break;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* ERrors */
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
case OM2K_MSGT_START_REQ_REJ:
OM2000: Handle (Dis)Connect Reject in NACK handler
2011-03-05 16:58:13 +00:00
case OM2K_MSGT_CONNECT_REJ:
OM2000: Consider OP_INFO_REJ and TEST_REQ_REJ as NACK
2011-03-05 18:42:16 +00:00
case OM2K_MSGT_OP_INFO_REJ:
OM2000: Handle (Dis)Connect Reject in NACK handler
2011-03-05 16:58:13 +00:00
case OM2K_MSGT_DISCONNECT_REJ:
OM2000: Consider OP_INFO_REJ and TEST_REQ_REJ as NACK
2011-03-05 18:42:16 +00:00
case OM2K_MSGT_TEST_REQ_REJ:
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
case OM2K_MSGT_CON_CONF_REQ_REJ:
case OM2K_MSGT_IS_CONF_REQ_REJ:
om2k: Add support for MCTR configuration Currently only supports a single MCTR with fixed configuration Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I96b8bb2c01c05bf153fc924f62bd6aafa96725ee
2020-05-05 18:14:44 +00:00
case OM2K_MSGT_MCTR_CONF_REQ_REJ:
OM2000: Print result/reason code when receiving a reject
2011-03-05 16:52:09 +00:00
case OM2K_MSGT_TX_CONF_REQ_REJ:
case OM2K_MSGT_RX_CONF_REQ_REJ:
case OM2K_MSGT_TS_CONF_REQ_REJ:
case OM2K_MSGT_TF_CONF_REQ_REJ:
case OM2K_MSGT_ENABLE_REQ_REJ:
case OM2K_MSGT_ALARM_STATUS_REQ_REJ:
case OM2K_MSGT_DISABLE_REQ_REJ:
rc = om2k_rx_nack(msg);
break;
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* Resolve the MO for this message */
mo = get_om2k_mo(bts, &o2h->mo);
if (!mo) {
LOGP(DNM, LOGL_ERROR, "Couldn't resolve MO for OM2K msg "
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
"%s: %s\n", get_value_string(om2k_msgcode_vals, msg_type), msgb_hexdump(msg));
om2000: Fix memory leak in OM2000 message handling Change-Id: I7a94320f3b3af65003df67c11fe7221dfc7d7d62
2022-01-02 14:10:56 +00:00
goto no_mo;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
abis_om2k: protect MO FSMs by NULL check Also set MO FSMs to NULL after freeing them. Change-Id: I30df0b9ab8bc47ba9756c8388e977deed0e40200
2016-11-02 01:41:41 +00:00
if (!mo->fsm) {
LOGP(DNM, LOGL_ERROR, "MO object should not generate any message. fsm == NULL "
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
"%s: %s\n", get_value_string(om2k_msgcode_vals, msg_type), msgb_hexdump(msg));
om2000: Fix memory leak in OM2000 message handling Change-Id: I7a94320f3b3af65003df67c11fe7221dfc7d7d62
2022-01-02 14:10:56 +00:00
goto no_mo;
abis_om2k: protect MO FSMs by NULL check Also set MO FSMs to NULL after freeing them. Change-Id: I30df0b9ab8bc47ba9756c8388e977deed0e40200
2016-11-02 01:41:41 +00:00
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
/* Dispatch message to that MO */
om2k_mo_fsm_recvmsg(bts, mo, &odm);
om2000: Fix memory leak in OM2000 message handling Change-Id: I7a94320f3b3af65003df67c11fe7221dfc7d7d62
2022-01-02 14:10:56 +00:00
no_mo:
[RBS2000] Add initial code for Ericsson OM2000 O&M protocol ... and link it from the RBS2000 BTS support
2011-02-12 11:33:06 +00:00
msgb_free(msg);
return rc;
}
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
static void om2k_mo_init(struct om2k_mo *mo, uint8_t class, uint8_t bts_nr, uint8_t assoc_so, uint8_t inst)
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
{
mo->addr.class = class;
mo->addr.bts = bts_nr;
mo->addr.assoc_so = assoc_so;
mo->addr.inst = inst;
}
/* initialize the OM2K_MO members of gsm_bts_trx and its timeslots */
void abis_om2k_trx_init(struct gsm_bts_trx *trx)
{
struct gsm_bts *bts = trx->bts;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
struct osmo_fsm_inst *trx_fi;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
unsigned int i;
OSMO_ASSERT(bts->type == GSM_BTS_TYPE_RBS2000);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
trx_fi = om2k_trx_fsm_alloc(trx->bts->rbs2000.bts_fi, trx, OM2K_BTS_EVT_TRX_DONE);
trx->rbs2000.trx_fi = trx_fi;
Add command to enable RX diversity to RBS2000 Allow selection of RX diversity from VTY Options are a,ab,b Default is 'a' so there is no change from previous behavior Change-Id: I430762b8cfa51060841d90ba4446de73bd557c6c
2021-03-11 00:00:25 +00:00
trx->rbs2000.rx_diversity = OM2K_RX_DIVERSITY_A;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&trx->rbs2000.trxc.om2k_mo, OM2K_MO_CLS_TRXC, bts->nr, 255, trx->nr);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(trx_fi, OM2K_TRX_EVT_TRXC_DONE, trx, &trx->rbs2000.trxc.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&trx->rbs2000.tx.om2k_mo, OM2K_MO_CLS_TX, bts->nr, 255, trx->nr);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(trx_fi, OM2K_TRX_EVT_TX_DONE, trx, &trx->rbs2000.tx.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&trx->rbs2000.rx.om2k_mo, OM2K_MO_CLS_RX, bts->nr, 255, trx->nr);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(trx_fi, OM2K_TRX_EVT_RX_DONE, trx, &trx->rbs2000.rx.om2k_mo);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
for (i = 0; i < ARRAY_SIZE(trx->ts); i++) {
dyn TS: init only when both RSL and the Channel OM are established Recent Icf6e25ff068e8a2600562d52726ead65e864ec02 changed the dyn_ts_init() hook from bootstrap_rsl() to the Channel OPSTART ACK, but this is not sufficient. Now RBS2k never calls dyn_ts_init(), and we may need to wait for RSL: Dyn TS should actually be initialized only when *both* OML opstart and RSL link are established. To that end, introduce a generalized API to query OML and RSL status and to trigger a timeslot init at the appropriate time. Add gsm_ts_check_init() to be called both when RSL and OML opstart are established: trigger gsm_ts_init() only when both are given. Add gsm_bts_trx_ts->initialized flag to mark whether initialization has already taken place. Add gsm_bts_mark_all_ts_uninitialized() to conveniently clear this flag for all TS in a BTS. Add gsm_bts_model.oml_is_ts_ready() callback so that each BTS implementation can return the OML status of a timeslot in its own OML implementation. Actually, currently all BTS models that need this init mechanism store the TS' OML status in ts->mo.nm_state. While we would in practice correctly init dyn TS by just looking at ts->mo.nm_state, semantically, the decision whether the TS is ready is up to the BTS models' specific OML implementations. From bootstrap_rsl(), call gsm_ts_check_init(), in case the TS OML Opstart has happened before RSL is established -- applies to all BTS models. For all BTS models: - call gsm_{bts,trx}_mark_all_ts_uninitialized() when OM is torn down, to make sure the TS init mechanism will work a second time. For all BTS models supporting dyn TS, i.e. osmo-bts, nanobts and RBS2k: - implement oml_is_ts_ready(). - call gsm_ts_check_init() when a Channel OM is taken into operation. Any BTS models that don't set oml_is_ts_ready() will see a ts init as soon as RSL is bootstrapped (incidentally, the old dyn TS behavior before recent Icf6e25ff068e8a2600562d52726ead65e864ec02). This firstly fixes dyn TS for RBS2k by re-adding the initial switch to PDCH, and furthermore does so only after both OML TS opstart and RSL are through. This fixes the ttcn3-bsc-tests around dyn TS, since for the osmo-bts-virtual, the RSL is established only after OML opstart on the TS, which was broken by Icf6e25ff068e8a2600562d52726ead65e864ec02. Nokia Site and Siemens BS11 practically do not require this init mechanism, since all that happens there so far is dyn TS init, and these BTS models do not support dyn TS of any kind. A future patch may add oml_is_ts_ready(). Related: OS#3205 Change-Id: I99f29d2ba079f6f4b77f0af12d9784588d2f56b3
2018-05-09 15:46:22 +00:00
struct gsm_bts_trx_ts *ts = &trx->ts[i];
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&ts->rbs2000.om2k_mo, OM2K_MO_CLS_TS, bts->nr, trx->nr, i);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(trx_fi, OM2K_TRX_EVT_TS_DONE, trx, &ts->rbs2000.om2k_mo);
large refactoring: use FSMs for lchans; add inter-BSC HO Add FSMs: - timeslot_fsm: handle dynamic timeslots and OML+RSL availability. - lchan_fsm: handle an individual lchan activation, RTP stream and release, signal the appropriate calling FSMs on success, failure, release. - mgw_endpoint_fsm: handle one entire endpoint with several CI. - assignment_fsm: BSSMAP Assignment Request. - handover_fsm: all of intra, inter-MO and inter-MT handover. Above FSMs absorb large parts of the gscon FSM. The gscon FSM was surpassing the maximum amount events (32), and it is more logical to treat assignment, handover and MGW procedures in separate FSMs. - Add logging macros for each FSM type: - LOG_TS() - LOG_LCHAN() - LOG_MGWEP(), LOG_CI() - LOG_ASSIGNMENT() - LOG_HO() These log with the osmo_fsm_inst where present. New style decision: logging without a final newline char is awkward, especially for gsmtap logging and when other logs interleave LOGPC() calls; we have various cases where the final \n goes missing, and also this invokes the log category checking N times instead of once. So I decided to make these macros *always* append a newline, but only if there is no final newline yet. I hope that the compiler optimizes the strlen() of the constant format strings away. Thus I can log with or without typing "\n" and always get an \n termination anyway. General: - replace osmo_timers, state enums and program-wide osmo_signal_dispatch() with dedicated FSM timeouts, states and events. - introduce a common way to handle Tnnn timers: gsm_timers.h/.c: struct T_def. These can be used (with some macro magic) to define a state's timeout once, and not make mistakes for each osmo_fsm_inst_state_chg(). Details: bsc_subscr_conn_fsm.c: - move most states of this FSM to lchan_fsm, assignment_fsm, handover_fsm and mgw_endpoint_fsm. - There is exactly one state for an ongoing Assignment, with all details handled in conn->assignment.fi. The state relies on the assignment_fsm's timeout. - There is one state for an ongoing Handover; except for an incoming Handover from a remote BSS, the gscon remains in ST_INIT until the new lchan and conn are both established. - move bssmap_add_lcls_status() to osmo_bsc_lcls.c abis_rsl.c: - move all dynamic timeslot logic away into timeslot_fsm. Only keep plain send/receive functions in abis_rsl.c - reduce some rsl functions to merely send a message, rename to "_tx_". - rsl_ipacc_mdcx(): add '_tx_' in the name; move parts that change the lchan state out into the lchan_fsm, the lchan->abis_ip.* are now set there prior to invoking this function. - move all timers and error/release handling away into various FSMs. - tweak ipa_smod_s_for_lchan() and ipa_rtp_pt_for_lchan() to not require an lchan passed, but just mode,type that they require. Rename to ipacc_speech_mode*() and ipacc_payload_type(). - add rsl_forward_layer3_info, used for inter-BSC HO MO, to just send the RR message received during BSSMAP Handover Command. - move various logging to LOG_LCHAN() in order to log with the lchan FSM instance. One drawback is that the lchan FSM is limited to one logging category, i.e. this moves some logging from DRR to DRSL. It might actually make sense to combine those categories. - lose LOGP...LOGPC logging cascades: they are bad for gsmtap logging and for performance. - handle_classmark_chg(): change logging, move cm2 len check out of the cm3 condition (I hope that's correct). - gsm48_send_ho_cmd(): split off gsm48_make_ho_cmd() which doesn't send right away, so that during inter-bsc HO we can make an RR Handover Command to send via the MSC to the remote BSS. assignment_fsm.c: - the Chan Mode Modify in case of re-using the same lchan is not implemented yet, because this was also missing in the previous implementation (OS#3357). osmo_bsc_api.c: - simplify bsc_mr_config() and move to lchan_fsm.c, the only caller; rename to lchan_mr_config(). (bsc_mr_config() used to copy the values to mr_bts_lv twice, once by member assignment and then again with a memcpy.) - During handover, we used to copy the MR config from the old lchan. Since we may handover between FR and HR, rather set the MR Config anew every time, so that FR rates are always available on FR lchans, and never on HR lchans. Depends: I03ee7ce840ecfa0b6a33358e7385528aabd4873f (libosmocore), I1f2918418c38918c5ac70acaa51a47adfca12b5e (libosmocore) Change-Id: I82e3f918295daa83274a4cf803f046979f284366
2018-05-14 16:14:15 +00:00
OSMO_ASSERT(ts->fi);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
}
/* initialize the OM2K_MO members of gsm_bts */
void abis_om2k_bts_init(struct gsm_bts *bts)
{
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
struct osmo_fsm_inst *bts_fi;
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
OSMO_ASSERT(bts->type == GSM_BTS_TYPE_RBS2000);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
bts_fi = om2k_bts_fsm_alloc(bts);
bts->rbs2000.bts_fi = bts_fi;
Add vty command for Ericsson RBS2000 sync This commit adds support for Selection of syncronization source. Options are internal for E1 and external for GPS Change-Id: Ia3d1acd6b3442238b35fc911092e12a6ac989adb
2021-03-10 23:33:46 +00:00
bts->rbs2000.sync_src = OM2K_SYNC_SRC_INTERNAL;
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.cf.om2k_mo, OM2K_MO_CLS_CF, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(bts_fi, OM2K_BTS_EVT_CF_DONE, bts->c0, &bts->rbs2000.cf.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.is.om2k_mo, OM2K_MO_CLS_IS, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(bts_fi, OM2K_BTS_EVT_IS_DONE, bts->c0, &bts->rbs2000.is.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.con.om2k_mo, OM2K_MO_CLS_CON, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(bts_fi, OM2K_BTS_EVT_CON_DONE, bts->c0, &bts->rbs2000.con.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.dp.om2k_mo, OM2K_MO_CLS_DP, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.tf.om2k_mo, OM2K_MO_CLS_TF, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(bts_fi, OM2K_BTS_EVT_TF_DONE, bts->c0, &bts->rbs2000.tf.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
om2k_mo_init(&bts->rbs2000.mctr.om2k_mo, OM2K_MO_CLS_MCTR, bts->nr, 0xFF, 0);
abis_om2000: keep OM2K FSMs around, don't terminate The existing code uses short-lived FSMs which are allocated straight before START, and which are free'd after DONE state is reached. While that works, it makes state introspection a bit hard, as one cannot show the FSM states, etc. Let's change to a different model where the per-OM2k-MO FSMs are always around (in state INIT after object creation). While at it, also introduce a RESET event that can reset each FSM instance back to INIT state, i.e. in case of OML link failure. Change-Id: Ia37cffff5c451e1d79a52ccae41ab5718b4661d4
2020-12-20 22:53:24 +00:00
om2k_mo_fsm_alloc(bts_fi, OM2K_BTS_EVT_MCTR_DONE, bts->c0, &bts->rbs2000.mctr.om2k_mo);
[cosmetic] abis_om2000: Re-format to use longer lines For several years now we are using line lengths longer than 80chars, let's re-format this code more in-line with our new style. Change-Id: I8143280358c0e23ca7e32b882e952eac991e105b
2020-12-21 10:11:41 +00:00
// FIXME: There can be multiple MCTRs ...
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}
static __attribute__((constructor)) void abis_om2k_init(void)
{
check for osmo_fsm_register() error return value Change-Id: I77e8ace007a3d6b9c40d3e158d1cdb7576aab77b
2019-12-01 12:18:20 +00:00
OSMO_ASSERT(osmo_fsm_register(&om2k_mo_fsm) == 0);
OSMO_ASSERT(osmo_fsm_register(&om2k_bts_fsm) == 0);
OSMO_ASSERT(osmo_fsm_register(&om2k_trx_fsm) == 0);
WIP: OM2000: Full state machine implementation using osmo_fsm Our existing OM2000 code for initializing all Managed Objects of a BTS at startup was never complete. Rather than trying to fix the old-style code, introudce a hierarchy of osmo_fsm's reflecting the full protocol hand-shake and sequence of bringing up the individual MO's. If this works out well, it mihgt make sense to convert the TS 12.21 OML code for other BTS models, too. Change-Id: I3e11b28ba22b8c227e0401e6207fdda5381dda8c
2016-07-09 20:20:57 +00:00
}