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

544 lines
14 KiB
C
Raw Normal View History

no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
/* GSM Channel allocation routines
*
* (C) 2008 by Harald Welte <laforge@gnumonks.org>
[rr] Send a Channel Release before deactivating the channel After auto releasing a channel the next paging request will not be immediately answered. The hypothesis was that we do not release the channel properly. Implementing Channel Release of GSM 04.08 should have fixed it, but it didn't. According to the wireshark dissectors the message is correct though. - Add the RR cause values to gsm_04_08. - Implement the Channel Release message - Invoke the release channel function before deallocating the lchan.
2009-02-06 21:55:37 +00:00
* (C) 2008, 2009 by Holger Hans Peter Freyther <zecke@selfish.org>
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
*
* All Rights Reserved
*
* This program is free software; you can redistribute it and/or modify
License change: We are now AGPLv3+ instead of GPLv2+ The reason for this is quite simple: We want to make sure anyone running a customized version of OpenBSC to operate a network will have to release all custom modifiations to the source code.
2011-01-01 14:25:50 +00:00
* 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
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
* (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
License change: We are now AGPLv3+ instead of GPLv2+ The reason for this is quite simple: We want to make sure anyone running a customized version of OpenBSC to operate a network will have to release all custom modifiations to the source code.
2011-01-01 14:25:50 +00:00
* GNU Affero General Public License for more details.
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
*
License change: We are now AGPLv3+ instead of GPLv2+ The reason for this is quite simple: We want to make sure anyone running a customized version of OpenBSC to operate a network will have to release all custom modifiations to the source code.
2011-01-01 14:25:50 +00:00
* 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/>.
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
*
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
misc: Remove gsm_subscriber.h include to reduce dependencies.
2010-12-22 17:13:00 +00:00
#include <openbsc/gsm_subscriber.h>
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
#include <openbsc/chan_alloc.h>
now we get up to the SETUP of MO calls
2008-12-27 01:55:51 +00:00
#include <openbsc/abis_nm.h>
Do not call rsl_chan_release directly but use the use_count of the lchan Call use_lchan early in allocate_loc_updating_req, do not directly call rsl_chan_release but go through channel alloc to take the use_count into account.
2009-01-01 03:46:11 +00:00
#include <openbsc/abis_rsl.h>
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
#include <openbsc/debug.h>
lchan: Close a possible still open RTP Socket In case of a reset (loss of the BTS) close down all remaining RTP Proxy sockets. In case of a lchan_free shout if the rtp proxy is still open. I would prefer if the proxy code sits inside the gsm subscriber connection.
2012-01-17 14:27:33 +00:00
#include <openbsc/rtp_proxy.h>
lchan: Handle the abnormal case of channel getting closed The abnormal case is that lchan_free ist getting called due a RSL_MT_CHAN_REL_ACK in the RSL but the refcount of this channel is not zero. This means that some "logical operation" is still going on that needs to be cancelled. Instead of always queuing up all operations in the struct gsm_lchan use the signal framework to inform higher layers about this abnormal case. In gsm_04_08.c a signal handler is installed and in the abnormal case the location updating request operation is freed.
2009-06-06 13:54:35 +00:00
#include <openbsc/signal.h>
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
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/talloc.h>
abis_rsl: Send the IMMEDIATE Assignment after the Channel Ack. The Channel Activate might be sent to a different TRX than the Immediate Assignment. So we need to make sure that the channel is activated before we send the immediate assignment for the RACH. Another reason for that is according to GSM 08.58 we should take the frame number from the activate and use it for the starting time inside the immediate assignment message. We obviously do not do this yet. The code assumes that the BTS will either respond with a CHAN ACK or a CHAN NACK if not the lchan will remain in the request state.
2010-06-22 04:11:59 +00:00
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
static int ts_is_usable(struct gsm_bts_trx_ts *ts)
{
/* FIXME: How does this behave for BS-11 ? */
if (is_ipaccess_bts(ts->trx->bts)) {
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
if (!nm_is_running(&ts->mo.nm_state))
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
return 0;
}
comments: clarify some dynamic TS comments A new type of dynamic channel will be introduced soon, so prepare some comments to name the dynamic TS kind more specifically. Change-Id: I51fa8c2ebba507299e55a5cb7e67e48a6c8471f7
2016-07-14 00:51:13 +00:00
/* If a TCH/F_PDCH TS is busy changing, it is already taken or not
dyn PDCH: set lchan->state after PDCH DEACT / before PDCH ACT Do the PDCH DE/ACT before we set the lchan->state to De-/Activation Requested. It makes more sense semantically to change PDCH mode while the lchan is still in NONE status. Thus slightly move some invocations: PDCH ACT: Free the lchan before PDCH activation. Hence remove the lchan_free() call from the rsl_rx_pdch_act_ack() code path; it used to do the PDCH activation first and call lchan_free() in the callback. PDCH DEACT: Set the (TCH) Activation Requested state only within rsl_chan_activate_lchan(), after the PDCH deact is complete. Channel allocator: don't pick channels that have a PDCH PENDING flag set, to avoid using channels that are still in PDCH switchover (despite their state being NONE). The lchan_may_change_pdch() sanity checks are becoming a lot simpler. Change-Id: I4206dd4808e21c3e59393ea7f5ab4f438afff066
2016-06-14 11:12:00 +00:00
* yet available. */
if (ts->pchan == GSM_PCHAN_TCH_F_PDCH) {
if (ts->flags & TS_F_PDCH_PENDING_MASK)
return 0;
}
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/* If a dynamic channel is busy changing, it is already taken or not
* yet available. */
if (ts->pchan == GSM_PCHAN_TCH_F_TCH_H_PDCH) {
if (ts->dyn.pchan_is != ts->dyn.pchan_want)
return 0;
}
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
return 1;
}
export the 'trx_is_usable() function and use it from paging this ensures we don't send paging requests to currently inactive transceivers.
2009-12-28 15:36:28 +00:00
int trx_is_usable(struct gsm_bts_trx *trx)
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
{
/* FIXME: How does this behave for BS-11 ? */
if (is_ipaccess_bts(trx->bts)) {
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
if (!nm_is_running(&trx->mo.nm_state) ||
!nm_is_running(&trx->bb_transc.mo.nm_state))
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
return 0;
}
return 1;
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
static struct gsm_lchan *
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
_lc_find_trx(struct gsm_bts_trx *trx, enum gsm_phys_chan_config pchan,
enum gsm_phys_chan_config dyn_as_pchan)
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
{
struct gsm_bts_trx_ts *ts;
dyn PDCH: Fix free slot search for chan_alloc_reverse == true For chan_alloc_reverse, _lc_find_trx() should return the last free slot instead of the first. Original patch by jolly, but split in two by nhofmeyr. Change-Id: Iff980242b9b5cb39345aaad0350ee368537677cd
2013-10-11 11:05:16 +00:00
int j, start, stop, dir, ss;
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
int check_subslots;
use list iteration macros to iterate over list of TRX
2009-08-10 11:25:55 +00:00
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
if (!trx_is_usable(trx))
return NULL;
dyn PDCH: Fix free slot search for chan_alloc_reverse == true For chan_alloc_reverse, _lc_find_trx() should return the last free slot instead of the first. Original patch by jolly, but split in two by nhofmeyr. Change-Id: Iff980242b9b5cb39345aaad0350ee368537677cd
2013-10-11 11:05:16 +00:00
if (trx->bts->chan_alloc_reverse) {
/* check TS 7..0 */
start = 7;
stop = -1;
dir = -1;
} else {
/* check TS 0..7 */
start = 0;
stop = 8;
dir = 1;
}
for (j = start; j != stop; j += dir) {
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
ts = &trx->ts[j];
channel allocator: don't allocate channels of unavalable TRXs In case we have multiple TRX configured, but not all of them are actually active/operational, we should not try to allocate channels from such transceivers.
2009-12-24 09:10:16 +00:00
if (!ts_is_usable(ts))
continue;
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
if (ts->pchan != pchan)
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
continue;
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/*
* Allocation for fully dynamic timeslots
* (does not apply for ip.access style GSM_PCHAN_TCH_F_PDCH)
*
* Note the special nature of a dynamic timeslot in PDCH mode:
* in PDCH mode, typically, lchan->type is GSM_LCHAN_NONE and
* lchan->state is LCHAN_S_NONE -- an otherwise unused slot
* becomes PDCH implicitly. In the same sense, this channel
* allocator will never be asked to find an available PDCH
* slot; only TCH/F or TCH/H will be requested, and PDCH mode
* means that it is available for switchover.
*
* A dynamic timeslot in PDCH mode may be switched to TCH/F or
* TCH/H. If a dyn TS is already in TCH/F or TCH/H mode, it
* means that it is in use and its mode can't be switched.
*
* The logic concerning channels for TCH/F is trivial: there is
* only one channel, so a dynamic TS in TCH/F mode is already
* taken and not available for allocation. For TCH/H, we need
* to check whether a dynamic timeslot is already in TCH/H mode
* and whether one of the two channels is still available.
*/
cosmetic: chan_alloc: use switch instead of if-cascade Preparing cosmetically for a subsequent commit which will add another pchan kind to be checked, rather use a "switch (pchan) {}". Also reverse one if() branch to "early-exit" style. Change-Id: Ie5eb0fa859c4f225616095dc56d52ce0f2dc8bdc
2016-12-05 15:48:36 +00:00
switch (pchan) {
case GSM_PCHAN_TCH_F_TCH_H_PDCH:
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
if (ts->dyn.pchan_is != ts->dyn.pchan_want) {
/* The TS's mode is being switched. Not
* available anymore/yet. */
DEBUGP(DRLL, "%s already in switchover\n",
gsm_ts_and_pchan_name(ts));
continue;
}
if (ts->dyn.pchan_is == GSM_PCHAN_PDCH) {
/* This slot is available. Still check for
* error states to be sure; in all cases the
* first lchan will be used. */
if (ts->lchan->state != LCHAN_S_NONE
&& ts->lchan->state != LCHAN_S_ACTIVE)
continue;
return ts->lchan;
}
cosmetic: chan_alloc: use switch instead of if-cascade Preparing cosmetically for a subsequent commit which will add another pchan kind to be checked, rather use a "switch (pchan) {}". Also reverse one if() branch to "early-exit" style. Change-Id: Ie5eb0fa859c4f225616095dc56d52ce0f2dc8bdc
2016-12-05 15:48:36 +00:00
if (ts->dyn.pchan_is != dyn_as_pchan)
/* not applicable. */
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
continue;
cosmetic: chan_alloc: use switch instead of if-cascade Preparing cosmetically for a subsequent commit which will add another pchan kind to be checked, rather use a "switch (pchan) {}". Also reverse one if() branch to "early-exit" style. Change-Id: Ie5eb0fa859c4f225616095dc56d52ce0f2dc8bdc
2016-12-05 15:48:36 +00:00
/* The requested type matches the dynamic timeslot's
* current mode. A channel may still be available
* (think TCH/H). */
check_subslots = ts_subslots(ts);
break;
Fix TCH/F_PDCH: no need to check ts subslots for PDCH For TCH/F_PDCH in PDCH mode, directly return the lchan to use, in order to switch it to TCH/F. To check the pchan type in chan_alloc.c, make ts_pchan() public in gsm_data_shared.h. Commit c3f72f63afde926dfc46827d6880055597515fb6 broke TCH/F_PDCH, as a fallout of setting the GSM_PCHAN_PDCH subslots number to 0. This is sane and correct, but the chan_alloc code failed to see a ts as available if it has no subslots. Explanation: _lc_find_trx() checks each timeslot. For normal, static TCH timeslots we determine the number of logical subslots contained and check whether one of them is free. For dynamic TS, we can do the same when in TCH mode, but when in PDCH mode, we already know that it is available for immediate switchover for voice and hence can return it right away. TCH/F_TCH/H_PDCH already has a special check for that. TCH/F_PDCH doesn't, but this worked for TCH/F_PDCH as long as ts_subslots() returned 1 for PDCH: the for-loop at the bottom of _lc_find_trx() checked one subslot, which succeeded on an lchan in PDCH mode, since PDCH lchans are always marked type == NONE and state == NONE. Now we more accurately acknowledge that a PDCH timeslot has zero subslots and that a dynamic timeslot in PDCH mode can always be switched to voice immediately, without checking lchan type or state. So, above mentioned commit set PDCH to zero subslots, and the for-loop to check the (zero) subslots never ran and hence never returned the lchan. This fix adds a special condition for TCH/F_PDCH in PDCH mode, same as TCH/F_TCH/H_PDCH. (Todo: ts_pchan() can probably be used in other places as well to remove some code dup. Leaving that for another patch.) Fixes: OS#1868 Change-Id: I5d555d018a5bcb8d948e54059d32ec4c9b3070d0
2016-12-05 15:50:47 +00:00
case GSM_PCHAN_TCH_F_PDCH:
/* Available for voice when in PDCH mode */
if (ts_pchan(ts) != GSM_PCHAN_PDCH)
continue;
/* Subslots of a PDCH ts don't need to be checked. */
return ts->lchan;
cosmetic: chan_alloc: use switch instead of if-cascade Preparing cosmetically for a subsequent commit which will add another pchan kind to be checked, rather use a "switch (pchan) {}". Also reverse one if() branch to "early-exit" style. Change-Id: Ie5eb0fa859c4f225616095dc56d52ce0f2dc8bdc
2016-12-05 15:48:36 +00:00
default:
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/* Not a dynamic channel, there is only one pchan kind: */
chan_alloc.c: use ts_subslots() instead of subslots_per_pchan[] The array will move to gsm_data_shared.c; to prepare, use the function instead. Change-Id: Icbea7dbd78abf6144e5291f531a97f96507d8cbf
2016-07-29 16:13:38 +00:00
check_subslots = ts_subslots(ts);
cosmetic: chan_alloc: use switch instead of if-cascade Preparing cosmetically for a subsequent commit which will add another pchan kind to be checked, rather use a "switch (pchan) {}". Also reverse one if() branch to "early-exit" style. Change-Id: Ie5eb0fa859c4f225616095dc56d52ce0f2dc8bdc
2016-12-05 15:48:36 +00:00
break;
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
}
/* Is a sub-slot still available? */
for (ss = 0; ss < check_subslots; ss++) {
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
struct gsm_lchan *lc = &ts->lchan[ss];
Introduce new ACT_REQ state to prevent race condition during channel allocation When we allocate a channel, we send the RSL CHAN ACT REQ and wait until we get a CHAN ACT ACK. Only the ACK will change the state, so there is a race where we allocate that same channel to a different channel request before we get the ACT ACK. Introducing a new ACT_REQ state resolves this issue.
2009-12-27 17:12:29 +00:00
if (lc->type == GSM_LCHAN_NONE &&
lc->state == LCHAN_S_NONE)
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
return lc;
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
}
}
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
return NULL;
}
static struct gsm_lchan *
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
_lc_dyn_find_bts(struct gsm_bts *bts, enum gsm_phys_chan_config pchan,
enum gsm_phys_chan_config dyn_as_pchan)
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
{
struct gsm_bts_trx *trx;
struct gsm_lchan *lc;
if (bts->chan_alloc_reverse) {
llist_for_each_entry_reverse(trx, &bts->trx_list, list) {
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
lc = _lc_find_trx(trx, pchan, dyn_as_pchan);
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
if (lc)
return lc;
}
} else {
llist_for_each_entry(trx, &bts->trx_list, list) {
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
lc = _lc_find_trx(trx, pchan, dyn_as_pchan);
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
if (lc)
return lc;
}
}
chan_alloc: remove ts_alloc() and ts_free() The idea of ts_alloc()/ts_free() dates back to the very early days of OpenBSC, where we didn't yet have a fixed PCHAN type assigned for every lchan in a BTS. However, ever since, PCHAN types (channel combinations) are configured by OML in a certain way, and we only allocate LCHANs inside PCHANs of a matching type. There should be no PCHANs with type GSM_PCHAN_NONE, unless those that you don't want to use for administraive reasons or the like.
2015-01-01 11:46:26 +00:00
return NULL;
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
}
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
static struct gsm_lchan *
_lc_find_bts(struct gsm_bts *bts, enum gsm_phys_chan_config pchan)
{
return _lc_dyn_find_bts(bts, pchan, GSM_PCHAN_NONE);
}
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
/* Allocate a logical channel.
*
comments: clarify some dynamic TS comments A new type of dynamic channel will be introduced soon, so prepare some comments to name the dynamic TS kind more specifically. Change-Id: I51fa8c2ebba507299e55a5cb7e67e48a6c8471f7
2016-07-14 00:51:13 +00:00
* Dynamic channel types: we always prefer a dedicated TS, and only pick +
* switch a dynamic TS if no pure TS of the requested PCHAN is available.
*
* TCH_F/PDCH: if we pick a PDCH ACT style dynamic TS as TCH/F channel, PDCH
* will be disabled in rsl_chan_activate_lchan(); there is no need to check
* whether PDCH mode is currently active, here.
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
*/
[alloc] Assign a TCH for LU when all SDCCHs are occupied. When the cell becomes visible we will be bombed with location updating requests and to reduce the load on the network we should assign as many channels for it as possible. During load peek it is even more important than to have a spare voice channel and in general the LU procedure is pretty fast.
2010-09-06 00:58:42 +00:00
struct gsm_lchan *lchan_alloc(struct gsm_bts *bts, enum gsm_chan_t type,
int allow_bigger)
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
{
struct gsm_lchan *lchan = NULL;
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
enum gsm_phys_chan_config first, first_cbch, second, second_cbch;
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
switch (type) {
case GSM_LCHAN_SDCCH:
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
if (bts->chan_alloc_reverse) {
first = GSM_PCHAN_SDCCH8_SACCH8C;
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
first_cbch = GSM_PCHAN_SDCCH8_SACCH8C_CBCH;
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
second = GSM_PCHAN_CCCH_SDCCH4;
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
second_cbch = GSM_PCHAN_CCCH_SDCCH4_CBCH;
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
} else {
first = GSM_PCHAN_CCCH_SDCCH4;
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
first_cbch = GSM_PCHAN_CCCH_SDCCH4_CBCH;
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
second = GSM_PCHAN_SDCCH8_SACCH8C;
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
second_cbch = GSM_PCHAN_SDCCH8_SACCH8C_CBCH;
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
}
lchan = _lc_find_bts(bts, first);
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
if (lchan == NULL)
lchan = _lc_find_bts(bts, first_cbch);
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
if (lchan == NULL)
when channel allocator operates in reverse order, prefer SDCCH/8 over SDCCH/4
2009-08-10 12:44:24 +00:00
lchan = _lc_find_bts(bts, second);
Add basic support for CBCH / SMS-CB (Cell Brroadcast) We can now configure the pyisical channel types for CBCH either in the CCCH+SDCCH4 or in the SDCCH8 chanel combination. Depending on whether a CBCH exists on the BTS, we also generate the SI4 with matching CBCH channel description to notify the phones of the existance of the CBCH. There is now a VTY command how a SMS-CB message can be sent to a given BTS. We do not yet have any logic at all for actual scheduling of multiple CBCH RSL messages towards one or multiple BTSs yet, though.
2014-12-28 14:00:45 +00:00
if (lchan == NULL)
lchan = _lc_find_bts(bts, second_cbch);
[alloc] Assign a TCH for LU when all SDCCHs are occupied. When the cell becomes visible we will be bombed with location updating requests and to reduce the load on the network we should assign as many channels for it as possible. During load peek it is even more important than to have a spare voice channel and in general the LU procedure is pretty fast.
2010-09-06 00:58:42 +00:00
/* allow to assign bigger channels */
if (allow_bigger) {
if (lchan == NULL) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_H);
lchan_alloc(): on alloc failure, report original type In lchan_alloc(), there are several decisions to fall back to another type of channel, followed by setting the channel type to the fall back type. So far, this was set regardless of allocation success or failure. If such fall back type is not available, do not modify the local type variable and thus report an S_CHALLOC_ALLOC_FAIL on the type originally requested (report is at the end of lchan_alloc()). Change-Id: Ie3d4cb74f91db0b8c4f5e595a963099de339ad1a
2016-06-06 10:57:22 +00:00
if (lchan)
type = GSM_LCHAN_TCH_H;
[alloc] Assign a TCH for LU when all SDCCHs are occupied. When the cell becomes visible we will be bombed with location updating requests and to reduce the load on the network we should assign as many channels for it as possible. During load peek it is even more important than to have a spare voice channel and in general the LU procedure is pretty fast.
2010-09-06 00:58:42 +00:00
}
if (lchan == NULL) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F);
lchan_alloc(): on alloc failure, report original type In lchan_alloc(), there are several decisions to fall back to another type of channel, followed by setting the channel type to the fall back type. So far, this was set regardless of allocation success or failure. If such fall back type is not available, do not modify the local type variable and thus report an S_CHALLOC_ALLOC_FAIL on the type originally requested (report is at the end of lchan_alloc()). Change-Id: Ie3d4cb74f91db0b8c4f5e595a963099de339ad1a
2016-06-06 10:57:22 +00:00
if (lchan)
type = GSM_LCHAN_TCH_F;
[alloc] Assign a TCH for LU when all SDCCHs are occupied. When the cell becomes visible we will be bombed with location updating requests and to reduce the load on the network we should assign as many channels for it as possible. During load peek it is even more important than to have a spare voice channel and in general the LU procedure is pretty fast.
2010-09-06 00:58:42 +00:00
}
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
/* try dynamic TCH/F_PDCH */
if (lchan == NULL) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F_PDCH);
/* TCH/F_PDCH will be used as TCH/F */
if (lchan)
type = GSM_LCHAN_TCH_F;
}
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/* try fully dynamic TCH/F_TCH/H_PDCH */
if (lchan == NULL) {
lchan = _lc_dyn_find_bts(bts, GSM_PCHAN_TCH_F_TCH_H_PDCH,
GSM_PCHAN_TCH_H);
if (lchan)
type = GSM_LCHAN_TCH_H;
}
/*
* No need to check fully dynamic channels for TCH/F:
* if no TCH/H was available, neither will be TCH/F.
*/
[alloc] Assign a TCH for LU when all SDCCHs are occupied. When the cell becomes visible we will be bombed with location updating requests and to reduce the load on the network we should assign as many channels for it as possible. During load peek it is even more important than to have a spare voice channel and in general the LU procedure is pretty fast.
2010-09-06 00:58:42 +00:00
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
break;
case GSM_LCHAN_TCH_F:
introduce bts->chan_alloc_reverse flag The channel allocator can be set in ascending or descending order. Ascnending means we first try to allocate channels on TRX0, then TRX1, etc. Descending means we first try to allocate cahnnels on TRXn, then n-1 down to 0.
2009-08-10 11:46:55 +00:00
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F);
chan_alloc: Fall-back to TCH/H, if we cannot find a TCH/F I'm not entirely sure if this is the best approach. However, there are phones that send a RACH request for TCH/F on MO calls, even though they actually do support TCH/H channels.
2014-05-18 20:23:26 +00:00
/* If we don't have TCH/F available, fall-back to TCH/H */
if (!lchan) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_H);
lchan_alloc(): on alloc failure, report original type In lchan_alloc(), there are several decisions to fall back to another type of channel, followed by setting the channel type to the fall back type. So far, this was set regardless of allocation success or failure. If such fall back type is not available, do not modify the local type variable and thus report an S_CHALLOC_ALLOC_FAIL on the type originally requested (report is at the end of lchan_alloc()). Change-Id: Ie3d4cb74f91db0b8c4f5e595a963099de339ad1a
2016-06-06 10:57:22 +00:00
if (lchan)
type = GSM_LCHAN_TCH_H;
chan_alloc: Fall-back to TCH/H, if we cannot find a TCH/F I'm not entirely sure if this is the best approach. However, there are phones that send a RACH request for TCH/F on MO calls, even though they actually do support TCH/H channels.
2014-05-18 20:23:26 +00:00
}
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
/* If we don't have TCH/H either, try dynamic TCH/F_PDCH */
if (!lchan) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F_PDCH);
/* TCH/F_PDCH used as TCH/F -- here, type is already
* set to GSM_LCHAN_TCH_F, but for clarity's sake... */
if (lchan)
type = GSM_LCHAN_TCH_F;
}
dyn TS: OS#1778 workaround: disable TCH/F on dyn TS for nitb To avoid two phones picking mismatching TCH pchans, never pick TCH/F on dynamic TS in osmo-nitb. Add gsm_network flag dyn_ts_allow_tch_f, set to true by default in gsm_network_init(). Set this flag to false in osmo-nitb's main(). See http://osmocom.org/issues/1778 Reasoning about ways to solve this: * a compile time switch doesn't work because libbsc is first compiled and then linked to both osmo-nitb and osmo-bsc. * we could test net->bsc_api == msc_bsc_api(), but I have the so-called MSC split waiting on branch sysmocom/cscn, which will result in msc_bsc_api() not being linked in the osmo-bsc binary. * have a function am_i_nitb() with different implementations in osmo-nitb and osmo-bsc, but then we'd need to add implementations to all tests and other binaries linking lchan_alloc(). * have a flag in struct bsc_api, but so far there are only function pointers there. Having a "global" flag in gsm_network allows to add a VTY command in case we decide to keep this feature (#1781), has no linking implications and is nicely explicit. Tested that osmo-bsc still picks TCH/F on dyn TS indirectly, since I have no standalone MSC available: when compiling osmo-nitb with the line that sets dyn_ts_allow_tch_f = false commented out, TCH/F is picked as described in OS#1778; and by printf-verifying that dyn_ts_allow_tch_f == true in osmo-bsc main(), only osmo-nitb should have TCH/F disabled. Related: OS#1778, OS#1781 Change-Id: If7e4797a72815fc6e2bbef27756ea5df69f4bde7
2016-07-23 18:15:28 +00:00
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/* Try fully dynamic TCH/F_TCH/H_PDCH as TCH/F... */
dyn TS: OS#1778 workaround: disable TCH/F on dyn TS for nitb To avoid two phones picking mismatching TCH pchans, never pick TCH/F on dynamic TS in osmo-nitb. Add gsm_network flag dyn_ts_allow_tch_f, set to true by default in gsm_network_init(). Set this flag to false in osmo-nitb's main(). See http://osmocom.org/issues/1778 Reasoning about ways to solve this: * a compile time switch doesn't work because libbsc is first compiled and then linked to both osmo-nitb and osmo-bsc. * we could test net->bsc_api == msc_bsc_api(), but I have the so-called MSC split waiting on branch sysmocom/cscn, which will result in msc_bsc_api() not being linked in the osmo-bsc binary. * have a function am_i_nitb() with different implementations in osmo-nitb and osmo-bsc, but then we'd need to add implementations to all tests and other binaries linking lchan_alloc(). * have a flag in struct bsc_api, but so far there are only function pointers there. Having a "global" flag in gsm_network allows to add a VTY command in case we decide to keep this feature (#1781), has no linking implications and is nicely explicit. Tested that osmo-bsc still picks TCH/F on dyn TS indirectly, since I have no standalone MSC available: when compiling osmo-nitb with the line that sets dyn_ts_allow_tch_f = false commented out, TCH/F is picked as described in OS#1778; and by printf-verifying that dyn_ts_allow_tch_f == true in osmo-bsc main(), only osmo-nitb should have TCH/F disabled. Related: OS#1778, OS#1781 Change-Id: If7e4797a72815fc6e2bbef27756ea5df69f4bde7
2016-07-23 18:15:28 +00:00
if (!lchan && bts->network->dyn_ts_allow_tch_f) {
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
lchan = _lc_dyn_find_bts(bts,
GSM_PCHAN_TCH_F_TCH_H_PDCH,
GSM_PCHAN_TCH_F);
if (lchan)
type = GSM_LCHAN_TCH_F;
}
/* ...and as TCH/H. */
if (!lchan) {
lchan = _lc_dyn_find_bts(bts,
GSM_PCHAN_TCH_F_TCH_H_PDCH,
GSM_PCHAN_TCH_H);
if (lchan)
type = GSM_LCHAN_TCH_H;
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
break;
case GSM_LCHAN_TCH_H:
fix mistypes, spaces and tabs Change-Id: I651473f638365bbd333c4c80d223cdd023dd2c51
2016-08-29 16:40:02 +00:00
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_H);
If we're trying to allocate TCH/H but don't have it, fallback to TCH/F The rationale is as following: If we have NECI=1, then the phone will request a channel with CHREQ "0100xxxx Originating speech call from dual-rate mobile station when TCH/H is sufficient and supported by the MS for speech calls", then we will try to allocate a TCH/H [as it is sufficient]. However, if there are no free TCH/H slots on the BTS, we abandon and can't handle the MO call at all :(
2009-12-12 19:58:20 +00:00
/* If we don't have TCH/H available, fall-back to TCH/F */
make speech calls with NECI=1 work Fix minor bug with speech calls in case of NECI=1
2009-12-12 20:16:38 +00:00
if (!lchan) {
If we're trying to allocate TCH/H but don't have it, fallback to TCH/F The rationale is as following: If we have NECI=1, then the phone will request a channel with CHREQ "0100xxxx Originating speech call from dual-rate mobile station when TCH/H is sufficient and supported by the MS for speech calls", then we will try to allocate a TCH/H [as it is sufficient]. However, if there are no free TCH/H slots on the BTS, we abandon and can't handle the MO call at all :(
2009-12-12 19:58:20 +00:00
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F);
lchan_alloc(): on alloc failure, report original type In lchan_alloc(), there are several decisions to fall back to another type of channel, followed by setting the channel type to the fall back type. So far, this was set regardless of allocation success or failure. If such fall back type is not available, do not modify the local type variable and thus report an S_CHALLOC_ALLOC_FAIL on the type originally requested (report is at the end of lchan_alloc()). Change-Id: Ie3d4cb74f91db0b8c4f5e595a963099de339ad1a
2016-06-06 10:57:22 +00:00
if (lchan)
type = GSM_LCHAN_TCH_F;
make speech calls with NECI=1 work Fix minor bug with speech calls in case of NECI=1
2009-12-12 20:16:38 +00:00
}
dyn TS: enhance channel allocator for dynamic TS Change _lc_find_bts() to _lc_dyn_find_bts() with added dyn_as_pchan arg to pass exactly as which pchan we'd like to allocate on a dynamic TS. Add _lc_find_bts() as wrapper so non-dynamic-TS callers remain unchanged. Also add dyn_as_pchan arg to _lc_find_trx() (not renaming to dyn and wrapping because there is only one caller). Implement dynamic allocator logic in _lc_find_trx() and lchan_alloc(). A returned dynamic channel still needs to be switched to the proper mode, which will follow in another commit. Replace a fixme comment with a normal comment in subslots_per_pchan[], because handling of dynamic TS is now defined. Change-Id: I18da7679300c43220d9baa6a304e8df74d366249
2016-07-14 01:01:24 +00:00
/* No dedicated TCH/x available -- try fully dynamic
* TCH/F_TCH/H_PDCH */
if (!lchan) {
lchan = _lc_dyn_find_bts(bts,
GSM_PCHAN_TCH_F_TCH_H_PDCH,
GSM_PCHAN_TCH_H);
if (lchan)
type = GSM_LCHAN_TCH_H;
}
/*
* No need to check TCH/F_TCH/H_PDCH channels for TCH/F:
* if no TCH/H was available, neither will be TCH/F.
*/
dyn PDCH: allow allocating TCH/F on TCH/F_PDCH slots Remove check for dyn PDCH in _lc_find_trx(), instead call _lc_find_trx() via _lc_find_bts() several times, so that pure TCH/F is preferred to TCH/F_PDCH. Add this logic next to the other channel match decisions in chan_alloc(). BTW, the removed check in _lc_find_trx() whether PDCH is active is not necessary, as described in the added comment for lchan_alloc(). Original patch idea by jolly, but split in two and implemented differently by nhofmeyr. Change-Id: I0c728b922656be03588b775638b610a93f8187d5
2016-06-06 10:53:25 +00:00
/* If we don't have TCH/F either, try dynamic TCH/F_PDCH */
if (!lchan) {
lchan = _lc_find_bts(bts, GSM_PCHAN_TCH_F_PDCH);
if (lchan)
type = GSM_LCHAN_TCH_F;
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
break;
default:
logging: introduce log levels at caller site This introduces a new LOGP() macro together with LOGL_* definition to support multiple log levels (severities) throughout the codebase. Please note that the actual logging system does not use them yet, in this patch we simply introduce the new macros at the caller site.
2009-12-17 22:10:46 +00:00
LOGP(DRLL, LOGL_ERROR, "Unknown gsm_chan_t %u\n", type);
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
}
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
if (lchan) {
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
lchan->type = type;
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
log lchan_alloc() result It is particularly interesting to see whether a given lchan type is allocated on a dynamic timeslot. Change-Id: I8a0bca6d9cd583a0988e5ee8f4e6f74f218f4185
2016-07-25 10:33:02 +00:00
LOGP(DRLL, LOGL_INFO, "%s Allocating lchan=%u as %s\n",
gsm_ts_and_pchan_name(lchan->ts),
lchan->nr, gsm_lchant_name(lchan->type));
[lchan] Keep track which SAPIs has been established with the BTS Keep track of which SAPIs have been established either by the BTS (from the MS) or by us. This can be used by the on-waves BSC code to figure out if a new request should be made.
2009-10-28 13:23:39 +00:00
/* clear sapis */
[lchan] Fix the SAPI reset in the chan allocation... Reset the whole array instead of just the first element.
2009-11-18 21:59:51 +00:00
memset(lchan->sapis, 0, ARRAY_SIZE(lchan->sapis));
[lchan] Keep track which SAPIs has been established with the BTS Keep track of which SAPIs have been established either by the BTS (from the MS) or by us. This can be used by the on-waves BSC code to figure out if a new request should be made.
2009-10-28 13:23:39 +00:00
[lchan] RSL and RR need the multirate config, place it in the lchan Both GSM 04.08 RR and GSM 08.58 RSL need the multirate config in the channel modify. Place the config in the lchan, change the gsm48 methods to not take the argument, change the RSL implementation to make use of it with the right IE. The other code should use the t(l)v_put routines as well but were left untouched for now.
2009-11-18 21:57:02 +00:00
/* clear multi rate config */
Add full AMR multirate IE support with VTY config for MS and BTS side
2014-01-19 10:47:44 +00:00
memset(&lchan->mr_ms_lv, 0, sizeof(lchan->mr_ms_lv));
memset(&lchan->mr_bts_lv, 0, sizeof(lchan->mr_bts_lv));
lchan: Remember why a channel is broken using static strings Remember why a channel is being marked as broken. So we can maybe understand what happend.
2014-12-28 11:08:28 +00:00
lchan->broken_reason = "";
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
} else {
struct challoc_signal_data sig;
log lchan_alloc() result It is particularly interesting to see whether a given lchan type is allocated on a dynamic timeslot. Change-Id: I8a0bca6d9cd583a0988e5ee8f4e6f74f218f4185
2016-07-25 10:33:02 +00:00
LOGP(DRLL, LOGL_ERROR, "Failed to allocate %s channel\n",
gsm_lchant_name(type));
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
sig.bts = bts;
sig.type = type;
src: use namespace prefix osmo_signal* Summary of changes: s/signal_cbfn/osmo_signal_cbfn/g s/register_signal_handler/osmo_signal_register_handler/g s/unregister_signal_handler/osmo_signal_unregister_handler/g s/dispatch_signal/osmo_signal_dispatch/g
2011-05-06 10:12:31 +00:00
osmo_signal_dispatch(SS_CHALLOC, S_CHALLOC_ALLOC_FAIL, &sig);
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
return lchan;
}
/* Free a logical channel */
void lchan_free(struct gsm_lchan *lchan)
{
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
struct challoc_signal_data sig;
introduce cache of 6 last recently received measurement reports for each lchan
2009-12-15 20:36:05 +00:00
int i;
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
sig.type = lchan->type;
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
lchan->type = GSM_LCHAN_NONE;
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
bsc_api: Allocate the subscriber_connection dynamically This is a big change to the way we use the subscriber connection. From now on it is is dynamically allocated and we will slowly move from a 1:1 lchan to conn to having more than one lchan per connection. This is the first commit, the subscr_con* methods will move to gsm_data once the use_count is removed from the connection, the freeing of the connection will also change.
2010-06-28 07:47:12 +00:00
if (lchan->conn) {
lchan: Every SS_LCHAN signal now sends a struct lchan_sig_data The SS_LCHAN signals now always include the lchan_sig_data. For the measurement report it will optionally include the measurement report as well. Attempt to update all handlers of this signal as well
2010-12-27 12:28:20 +00:00
struct lchan_signal_data sig;
bsc_api: Allocate the subscriber_connection dynamically This is a big change to the way we use the subscriber connection. From now on it is is dynamically allocated and we will slowly move from a 1:1 lchan to conn to having more than one lchan per connection. This is the first commit, the subscr_con* methods will move to gsm_data once the use_count is removed from the connection, the freeing of the connection will also change.
2010-06-28 07:47:12 +00:00
/* We might kill an active channel... */
lchan: Every SS_LCHAN signal now sends a struct lchan_sig_data The SS_LCHAN signals now always include the lchan_sig_data. For the measurement report it will optionally include the measurement report as well. Attempt to update all handlers of this signal as well
2010-12-27 12:28:20 +00:00
sig.lchan = lchan;
sig.mr = NULL;
src: use namespace prefix osmo_signal* Summary of changes: s/signal_cbfn/osmo_signal_cbfn/g s/register_signal_handler/osmo_signal_register_handler/g s/unregister_signal_handler/osmo_signal_unregister_handler/g s/dispatch_signal/osmo_signal_dispatch/g
2011-05-06 10:12:31 +00:00
osmo_signal_dispatch(SS_LCHAN, S_LCHAN_UNEXPECTED_RELEASE, &sig);
Reset the use_count to zero in chan_free It is possible that the BTS is closing the channel even when our upper layers are doing work. Reset the use_count add a fixme to call cancellations for pending operations. Cancellation of the call state (state machines in general) and such come into mind...
2009-01-09 21:39:17 +00:00
}
lchan: Close a possible still open RTP Socket In case of a reset (loss of the BTS) close down all remaining RTP Proxy sockets. In case of a lchan_free shout if the rtp proxy is still open. I would prefer if the proxy code sits inside the gsm subscriber connection.
2012-01-17 14:27:33 +00:00
if (lchan->abis_ip.rtp_socket) {
LOGP(DRLL, LOGL_ERROR, "%s RTP Proxy Socket remained open.\n",
gsm_lchan_name(lchan));
rtp_socket_free(lchan->abis_ip.rtp_socket);
lchan->abis_ip.rtp_socket = NULL;
}
bsc_api: Allocate the subscriber_connection dynamically This is a big change to the way we use the subscriber connection. From now on it is is dynamically allocated and we will slowly move from a 1:1 lchan to conn to having more than one lchan per connection. This is the first commit, the subscr_con* methods will move to gsm_data once the use_count is removed from the connection, the freeing of the connection will also change.
2010-06-28 07:47:12 +00:00
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
/* stop the timer */
src: use namespace prefix osmo_timer* for timer functions Summary of changes: s/struct timer_list/struct osmo_timer_list/g s/bsc_add_timer/osmo_timer_add/g s/bsc_schedule_timer/osmo_timer_schedule/g s/bsc_del_timer/osmo_timer_del/g s/bsc_timer_pending/osmo_timer_pending/g s/bsc_nearest_timer/osmo_timers_nearest/g s/bsc_prepare_timers/osmo_timers_prepare/g s/bsc_update_timers/osmo_timers_update/g s/bsc_timer_check/osmo_timers_check/g
2011-05-06 10:11:06 +00:00
osmo_timer_del(&lchan->T3101);
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
introduce cache of 6 last recently received measurement reports for each lchan
2009-12-15 20:36:05 +00:00
/* clear cached measuement reports */
lchan->meas_rep_idx = 0;
for (i = 0; i < ARRAY_SIZE(lchan->meas_rep); i++) {
lchan->meas_rep[i].flags = 0;
lchan->meas_rep[i].nr = 0;
}
[handover] Real handover algorithm This implements the handover algorithm (and associated parameters) as described in Chapter 8 of the book "Performance Enhancements in a Frequency |Hopping GSM Network" by Thomas Toftegard Nielsen and Jeroen Wigard. The parameters such as averaging windows are configured in struct gsm_network. We keep some state to trakc up to 10 neighbors as they are being reported from the MS. This has so far only been tested in a network with two BTS that have each other as neighbor. Networks with morge neighbors might encounter bugs.
2009-12-21 12:30:17 +00:00
for (i = 0; i < ARRAY_SIZE(lchan->neigh_meas); i++)
lchan->neigh_meas[i].arfcn = 0;
introduce cache of 6 last recently received measurement reports for each lchan
2009-12-15 20:36:05 +00:00
abis_rsl: Send the IMMEDIATE Assignment after the Channel Ack. The Channel Activate might be sent to a different TRX than the Immediate Assignment. So we need to make sure that the channel is activated before we send the immediate assignment for the RACH. Another reason for that is according to GSM 08.58 we should take the frame number from the activate and use it for the starting time inside the immediate assignment message. We obviously do not do this yet. The code assumes that the BTS will either respond with a CHAN ACK or a CHAN NACK if not the lchan will remain in the request state.
2010-06-22 04:11:59 +00:00
if (lchan->rqd_ref) {
talloc_free(lchan->rqd_ref);
lchan->rqd_ref = NULL;
lchan->rqd_ta = 0;
}
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
sig.lchan = lchan;
sig.bts = lchan->ts->trx->bts;
src: use namespace prefix osmo_signal* Summary of changes: s/signal_cbfn/osmo_signal_cbfn/g s/register_signal_handler/osmo_signal_register_handler/g s/unregister_signal_handler/osmo_signal_unregister_handler/g s/dispatch_signal/osmo_signal_dispatch/g
2011-05-06 10:12:31 +00:00
osmo_signal_dispatch(SS_CHALLOC, S_CHALLOC_FREED, &sig);
chan_alloc: Support allocating TCH/F of a dynamic TCH/F + PDCH This code simply enables the channel allocator to understand the dynamic TCH/F / PDCH channel type as used by ip.access nanoBTS. It does not actually try to switch the dynamic mode, but instead sends signals to a (not yet present) dynamic switching algorithm.
2010-03-28 07:58:03 +00:00
bsc_api: Allocate the subscriber_connection dynamically This is a big change to the way we use the subscriber connection. From now on it is is dynamically allocated and we will slowly move from a 1:1 lchan to conn to having more than one lchan per connection. This is the first commit, the subscr_con* methods will move to gsm_data once the use_count is removed from the connection, the freeing of the connection will also change.
2010-06-28 07:47:12 +00:00
if (lchan->conn) {
bsc_msc: Remove use_count from the subscriber connection A channel will be released in case of * Errors via the clear_request callback... * no more transactions and operations are going on. This means that if we do something without a transaction the channel might be closed down right away. The bug fix will be to create a transaction/operation.
2010-06-28 09:09:29 +00:00
LOGP(DRLL, LOGL_ERROR, "the subscriber connection should be gone.\n");
bsc_api: Allocate the subscriber_connection dynamically This is a big change to the way we use the subscriber connection. From now on it is is dynamically allocated and we will slowly move from a 1:1 lchan to conn to having more than one lchan per connection. This is the first commit, the subscr_con* methods will move to gsm_data once the use_count is removed from the connection, the freeing of the connection will also change.
2010-06-28 07:47:12 +00:00
lchan->conn = NULL;
}
no reasonable commit message possible. tons of changes of the last 2 days
2008-12-25 23:28:35 +00:00
/* FIXME: ts_free() the timeslot, if we're the last logical
* channel using it */
}
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
[bsc_init] When the RSL/OML connection drops, free all lchans Free all allocated channels on the TRX that failed, go through lchan_free to signal higher layers and then force a reset of the channel. Make the TRX and TS unusable by setting the operational set to 0 (not really defined) which should be reset once the RSL is coming up again.
2010-04-15 09:21:02 +00:00
/*
* There was an error with the TRX and we need to forget
* any state so that a lchan can be allocated again after
* the trx is fully usable.
lchan: Document why lchan_reset is there... at least try to.
2010-06-22 04:19:06 +00:00
*
* This should be called after lchan_free to force a channel
* be available for allocation again. This means that this
* method will stop the "delay after error"-timer and set the
* state to LCHAN_S_NONE.
[bsc_init] When the RSL/OML connection drops, free all lchans Free all allocated channels on the TRX that failed, go through lchan_free to signal higher layers and then force a reset of the channel. Make the TRX and TS unusable by setting the operational set to 0 (not really defined) which should be reset once the RSL is coming up again.
2010-04-15 09:21:02 +00:00
*/
void lchan_reset(struct gsm_lchan *lchan)
{
src: use namespace prefix osmo_timer* for timer functions Summary of changes: s/struct timer_list/struct osmo_timer_list/g s/bsc_add_timer/osmo_timer_add/g s/bsc_schedule_timer/osmo_timer_schedule/g s/bsc_del_timer/osmo_timer_del/g s/bsc_timer_pending/osmo_timer_pending/g s/bsc_nearest_timer/osmo_timers_nearest/g s/bsc_prepare_timers/osmo_timers_prepare/g s/bsc_update_timers/osmo_timers_update/g s/bsc_timer_check/osmo_timers_check/g
2011-05-06 10:11:06 +00:00
osmo_timer_del(&lchan->T3101);
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
osmo_timer_del(&lchan->T3109);
src: use namespace prefix osmo_timer* for timer functions Summary of changes: s/struct timer_list/struct osmo_timer_list/g s/bsc_add_timer/osmo_timer_add/g s/bsc_schedule_timer/osmo_timer_schedule/g s/bsc_del_timer/osmo_timer_del/g s/bsc_timer_pending/osmo_timer_pending/g s/bsc_nearest_timer/osmo_timers_nearest/g s/bsc_prepare_timers/osmo_timers_prepare/g s/bsc_update_timers/osmo_timers_update/g s/bsc_timer_check/osmo_timers_check/g
2011-05-06 10:11:06 +00:00
osmo_timer_del(&lchan->T3111);
osmo_timer_del(&lchan->error_timer);
[bsc_init] When the RSL/OML connection drops, free all lchans Free all allocated channels on the TRX that failed, go through lchan_free to signal higher layers and then force a reset of the channel. Make the TRX and TS unusable by setting the operational set to 0 (not really defined) which should be reset once the RSL is coming up again.
2010-04-15 09:21:02 +00:00
lchan->type = GSM_LCHAN_NONE;
lchan->state = LCHAN_S_NONE;
lchan: Close a possible still open RTP Socket In case of a reset (loss of the BTS) close down all remaining RTP Proxy sockets. In case of a lchan_free shout if the rtp proxy is still open. I would prefer if the proxy code sits inside the gsm subscriber connection.
2012-01-17 14:27:33 +00:00
if (lchan->abis_ip.rtp_socket) {
rtp_socket_free(lchan->abis_ip.rtp_socket);
lchan->abis_ip.rtp_socket = NULL;
}
[bsc_init] When the RSL/OML connection drops, free all lchans Free all allocated channels on the TRX that failed, go through lchan_free to signal higher layers and then force a reset of the channel. Make the TRX and TS unusable by setting the operational set to 0 (not really defined) which should be reset once the RSL is coming up again.
2010-04-15 09:21:02 +00:00
}
chan_alloc: Change Channel Release to release SAPIs, then the channel Currently every SAPI release indication will trigger the channel. It was possible that we had SAPI=3 and SAPI=0 allocated and we tried to release the channel by sending a RF Channel Release, the BTS answered with a RF Channel Release ACK but also sent the SAPI Release Indication which triggered a channel release here. So it was possible that we would have released a newly allocated channel because of the SAPI release of the old connection. This code now works by releasing all SAPIs from highest to lowest, then sending a SACH Deactivate and finally releasing the channel. This approach is in use on the on-waves/bsc-master.
2010-07-29 09:09:36 +00:00
/* Drive the release process of the lchan */
lchan: Release the lchan more quickly, align with GSM 04.08 * Release all channels with SAPI > 0 with the "local end release" (as of NOTE 1 of GSM 04.08). * No need to wait for all SAPIs to be torn down and the normal REL_IND/REL_CONF will call rsl_handle_release and the channel should be released. * Update the documentation
2011-12-27 21:24:17 +00:00
static void _lchan_handle_release(struct gsm_lchan *lchan,
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
int sacch_deact, int mode)
Recylce a gsm_lchan when the refcount drops to zero When a channel is allocated, start a timeout, when a lchan_use is used the timer will be restarted, when the timeout fires we will try to recycle or restart the timer.
2008-12-30 19:18:21 +00:00
{
lchan: Release the lchan more quickly, align with GSM 04.08 * Release all channels with SAPI > 0 with the "local end release" (as of NOTE 1 of GSM 04.08). * No need to wait for all SAPIs to be torn down and the normal REL_IND/REL_CONF will call rsl_handle_release and the channel should be released. * Update the documentation
2011-12-27 21:24:17 +00:00
/* Release all SAPIs on the local end and continue */
rsl_release_sapis_from(lchan, 1, RSL_REL_LOCAL_END);
/*
* Shall we send a RR Release, start T3109 and wait for the
* release indication from the BTS or just take it down (e.g.
* on assignment requests)
*/
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
if (sacch_deact) {
[rr] Send a Channel Release before deactivating the channel After auto releasing a channel the next paging request will not be immediately answered. The hypothesis was that we do not release the channel properly. Implementing Channel Release of GSM 04.08 should have fixed it, but it didn't. According to the wireshark dissectors the message is correct though. - Add the RR cause values to gsm_04_08. - Implement the Channel Release message - Invoke the release channel function before deallocating the lchan.
2009-02-06 21:55:37 +00:00
gsm48_send_rr_release(lchan);
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
/* Deactivate the SACCH on the BTS side */
rsl_deact_sacch(lchan);
rsl_start_t3109(lchan);
gsm: Do not attempt to release SAPI=0 if it was never allocated In case of handover (but probably on RACH) we would send a RLL for SAPI=0 even if this SAPI was never established. After we have released all SAPI>0 locally check that SAPI=0 is established and if not release the rf channel directly.
2012-12-25 22:45:14 +00:00
} else if (lchan->sapis[0] == LCHAN_SAPI_UNUSED) {
rsl_direct_rf_release(lchan);
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
} else {
lchan: Release the lchan more quickly, align with GSM 04.08 * Release all channels with SAPI > 0 with the "local end release" (as of NOTE 1 of GSM 04.08). * No need to wait for all SAPIs to be torn down and the normal REL_IND/REL_CONF will call rsl_handle_release and the channel should be released. * Update the documentation
2011-12-27 21:24:17 +00:00
rsl_release_request(lchan, 0, mode);
lchan: Introduce T3109 handling for the release procedure T3109 is started when the SACCH is deactivated. It is stopped when the phones sends the DISC/UA/UM on LAPDm for the main signalling link. In case of timeout the abnormal release procedure will be initiated. Make sure to not issue the SACCH Deactivate twice to avoid confusing the equipment. This is still not fully spec compliant. In case of a timeout the abnormal release handling will be started which involves starting T3111+2. The error handling should be split out of the rf channel release method, e.g. lchan_release should be called and check if the channel release was already initiated.
2011-12-28 15:21:05 +00:00
}
chan_alloc: Change Channel Release to release SAPIs, then the channel Currently every SAPI release indication will trigger the channel. It was possible that we had SAPI=3 and SAPI=0 allocated and we tried to release the channel by sending a RF Channel Release, the BTS answered with a RF Channel Release ACK but also sent the SAPI Release Indication which triggered a channel release here. So it was possible that we would have released a newly allocated channel because of the SAPI release of the old connection. This code now works by releasing all SAPIs from highest to lowest, then sending a SACH Deactivate and finally releasing the channel. This approach is in use on the on-waves/bsc-master.
2010-07-29 09:09:36 +00:00
}
/* Consider releasing the channel now */
rsl: Rename the reason to release_mode and use the enum value The API with 0 and 1 as legal values is hard to understand. Use the recently introduced enum values instead.
2012-12-06 11:01:38 +00:00
int lchan_release(struct gsm_lchan *lchan, int sacch_deact, enum rsl_rel_mode mode)
chan_alloc: Change Channel Release to release SAPIs, then the channel Currently every SAPI release indication will trigger the channel. It was possible that we had SAPI=3 and SAPI=0 allocated and we tried to release the channel by sending a RF Channel Release, the BTS answered with a RF Channel Release ACK but also sent the SAPI Release Indication which triggered a channel release here. So it was possible that we would have released a newly allocated channel because of the SAPI release of the old connection. This code now works by releasing all SAPIs from highest to lowest, then sending a SACH Deactivate and finally releasing the channel. This approach is in use on the on-waves/bsc-master.
2010-07-29 09:09:36 +00:00
{
DEBUGP(DRLL, "%s starting release sequence\n", gsm_lchan_name(lchan));
rsl_lchan_set_state(lchan, LCHAN_S_REL_REQ);
bsc_api: Clear lchan->conn on lchan_release Set conn to NULL before starting the release procedure, remove code to check for a lchan->conn as it can not happen. If there are any memleaks we will notice them. Detach the lchan->conn from the channel before detaching.
2010-11-14 15:04:46 +00:00
lchan->conn = NULL;
lchan: Release the lchan more quickly, align with GSM 04.08 * Release all channels with SAPI > 0 with the "local end release" (as of NOTE 1 of GSM 04.08). * No need to wait for all SAPIs to be torn down and the normal REL_IND/REL_CONF will call rsl_handle_release and the channel should be released. * Update the documentation
2011-12-27 21:24:17 +00:00
_lchan_handle_release(lchan, sacch_deact, mode);
Do not call rsl_chan_release directly but use the use_count of the lchan Call use_lchan early in allocate_loc_updating_req, do not directly call rsl_chan_release but go through channel alloc to take the use_count into account.
2009-01-01 03:46:11 +00:00
return 1;
}
Revert "bts: extend bts_chan_load to allow counting tch only" This reverts commit 308cb0719dca3ba8eed1eff2a2124d44f34d9a28. Problems in this commit: openbsc/src/libbsc/chan_alloc.c:523: case GSM_PCHAN_TCH_F_PDCH: This is actually wrong, GSM_PCHAN_TCH_F_PDCH use ts->flags, not ts->dyn below (due to historical reasons and could be unified). 560: if (only_count_tch && !chan_is_tch(ts)) This has exactly one effect: it excludes GSM_PCHAN_TCH_F_PDCH when in PDCH mode, because for all other PDCH (plain PDCH and TCH/F_TCH/H_PDCH in PDCH mode) below ts_subslots() returns 0 and skips the for() loop. I consider this a bug in TCH/F_PDCH, to be fixed in an upcoming commit. I don't see why we need the only_count_tch argument, because this should normally only count TCH, weren't it for the TCH/F_PDCH bug. If dyn TS should be counted differently, we should do this in a different way. Change-Id: I34dbbaf53a800115e3d03bd44028cad675f3b525
2016-09-25 15:01:20 +00:00
void bts_chan_load(struct pchan_load *cl, const struct gsm_bts *bts)
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
{
struct gsm_bts_trx *trx;
llist_for_each_entry(trx, &bts->trx_list, list) {
int i;
/* skip administratively deactivated tranxsceivers */
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
if (!nm_is_running(&trx->mo.nm_state) ||
!nm_is_running(&trx->bb_transc.mo.nm_state))
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
continue;
for (i = 0; i < ARRAY_SIZE(trx->ts); i++) {
struct gsm_bts_trx_ts *ts = &trx->ts[i];
struct load_counter *pl = &cl->pchan[ts->pchan];
int j;
dyn TS: bts_chan_load: use correct nr of subslots for dyn ts For TCH/F_TCH/H_PDCH dynamic timeslots, the ts->pchan does not lead to a meaningful value from the subslots_per_pchan[] array. Use the ts_subslots() function instead, which checks for dyn pchan. Change-Id: I659acebca82dfb3e305433471be64e9d27439af8
2016-07-29 16:10:59 +00:00
int subslots;
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
/* skip administratively deactivated timeslots */
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
if (!nm_is_running(&ts->mo.nm_state))
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
continue;
dyn TS: bts_chan_load: use correct nr of subslots for dyn ts For TCH/F_TCH/H_PDCH dynamic timeslots, the ts->pchan does not lead to a meaningful value from the subslots_per_pchan[] array. Use the ts_subslots() function instead, which checks for dyn pchan. Change-Id: I659acebca82dfb3e305433471be64e9d27439af8
2016-07-29 16:10:59 +00:00
subslots = ts_subslots(ts);
for (j = 0; j < subslots; j++) {
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
struct gsm_lchan *lchan = &ts->lchan[j];
pl->total++;
switch (lchan->state) {
case LCHAN_S_NONE:
break;
default:
pl->used++;
break;
}
}
}
}
}
void network_chan_load(struct pchan_load *pl, struct gsm_network *net)
{
struct gsm_bts *bts;
memset(pl, 0, sizeof(*pl));
llist_for_each_entry(bts, &net->bts_list, list)
Revert "bts: extend bts_chan_load to allow counting tch only" This reverts commit 308cb0719dca3ba8eed1eff2a2124d44f34d9a28. Problems in this commit: openbsc/src/libbsc/chan_alloc.c:523: case GSM_PCHAN_TCH_F_PDCH: This is actually wrong, GSM_PCHAN_TCH_F_PDCH use ts->flags, not ts->dyn below (due to historical reasons and could be unified). 560: if (only_count_tch && !chan_is_tch(ts)) This has exactly one effect: it excludes GSM_PCHAN_TCH_F_PDCH when in PDCH mode, because for all other PDCH (plain PDCH and TCH/F_TCH/H_PDCH in PDCH mode) below ts_subslots() returns 0 and skips the for() loop. I consider this a bug in TCH/F_PDCH, to be fixed in an upcoming commit. I don't see why we need the only_count_tch argument, because this should normally only count TCH, weren't it for the TCH/F_PDCH bug. If dyn TS should be counted differently, we should do this in a different way. Change-Id: I34dbbaf53a800115e3d03bd44028cad675f3b525
2016-09-25 15:01:20 +00:00
bts_chan_load(pl, bts);
Display current channel usage load in 'show bts' and 'show network' This is just the load at one given instant. We definitely also want to see some averages and record the measurements in a database later.
2009-12-22 12:09:29 +00:00
}
Revert "bts: extend bts_chan_load to allow counting tch only" This reverts commit 308cb0719dca3ba8eed1eff2a2124d44f34d9a28. Problems in this commit: openbsc/src/libbsc/chan_alloc.c:523: case GSM_PCHAN_TCH_F_PDCH: This is actually wrong, GSM_PCHAN_TCH_F_PDCH use ts->flags, not ts->dyn below (due to historical reasons and could be unified). 560: if (only_count_tch && !chan_is_tch(ts)) This has exactly one effect: it excludes GSM_PCHAN_TCH_F_PDCH when in PDCH mode, because for all other PDCH (plain PDCH and TCH/F_TCH/H_PDCH in PDCH mode) below ts_subslots() returns 0 and skips the for() loop. I consider this a bug in TCH/F_PDCH, to be fixed in an upcoming commit. I don't see why we need the only_count_tch argument, because this should normally only count TCH, weren't it for the TCH/F_PDCH bug. If dyn TS should be counted differently, we should do this in a different way. Change-Id: I34dbbaf53a800115e3d03bd44028cad675f3b525
2016-09-25 15:01:20 +00:00