This is merely a historical archive of years 2008-2021, before the migration to mailman3.
A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/osmocom-commitlog@lists.osmocom.org/.
gitosis at osmocom.org gitosis at osmocom.orgThis is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "Osmocom BTS-side code (Abis, scheduling, ...)". The branch, neels/dyn_no_pcu has been created at 061e66121f107a3c3a00d4f0657effab0b81ee1d (commit) - Log ----------------------------------------------------------------- http://cgit.osmocom.org/osmo-bts/commit/?id=061e66121f107a3c3a00d4f0657effab0b81ee1d commit 061e66121f107a3c3a00d4f0657effab0b81ee1d Author: Neels Hofmeyr <nhofmeyr at sysmocom.de> Date: Mon Aug 22 23:07:09 2016 +0200 dyn TS: if PCU is not connected, allow operation as TCH Before this patch, Osmocom style TCH/F_TCH/H_PDCH dyn TS were paralyzed if no PCU was running. The state of the dyn TS would lock up in the PDCH activation phase since the PCU never completed the process. Make more robust, i.e. don't concern the BSC with PDCH activation failures. This matches the way plain PDCH TS work: besides declaring the TS as PDCH, the BSC is not involved and is not told about errors. During PDCH deactivation, still wait for the PCU to tear down the PDTCH SAPIs, but in case no PCU is connected, send a rel ack right away. Thus, the BSC will happily switch Osmocom style dynamic timeslots to and from PDCH mode, using the dyn TS as voice channels as needed, and not caring about possible PDCH failures. GPRS starts working right away as soon as a PCU connects, regardless of dyn TS having been used for voice any number of times, and without another switchover needed. In detail: In rsl_rx_chan_activ(), upon receiving a PDCH activation, send an RSL chan act ack right away, unconditionally (with an explaining comment). Do not concern the Abis link with PDCH activation failures. Since we're acking right away now, drop the chan act ack that would follow after the PCU activation: as before dyn TS, only send acks and nacks for rel_act_kind == LCHAN_REL_ACT_RSL (PDCH runs as LCHAN_REL_ACT_PCU). In dyn_ts_pdch_release, indicate that the PCU is not connected by means of returning 1. In rsl_rx_rf_chan_rel(), use this indicator to send a rel ack right away if the PCU is not connected. Change-Id: I2a0b9730197786b99ff3bc1f08c75f7d279cb1f7 http://cgit.osmocom.org/osmo-bts/commit/?id=244db5514fbdcc684ac64c19e4b3e268eea5d027 commit 244db5514fbdcc684ac64c19e4b3e268eea5d027 Author: Neels Hofmeyr <nhofmeyr at sysmocom.de> Date: Mon Aug 22 18:23:03 2016 +0200 common/rsl: move decision whether to chan act ack/nack to common function Prepare for a dyn TS patch that needs to call rsl_tx_chan_act_ack() directly without the rel_act_kind decision. Add function rsl_tx_chan_act_acknack() to wrap rsl_tx_chan_act_ack() and rsl_tx_chan_act_nack(). Move the decision whether to drop the ack/nack, based on lchan->rel_act_kind, to the new function, losing some code dup. Change all callers to use the new function; drop the two older ones from rsl.h and make them static. Note: for nack, the exception for dyn TS in PDCH mode was missing (rsl_tx_chan_act_nack() had only the rel_act_kind != LCHAN_REL_ACT_RSL condition, but should also have had the dyn TS exception as in rsl_tx_chan_act_ack()). I already know that this exception will again be removed in an upcoming commit, but for patch readability it logically makes sense to add it here. To easily include the nack case, drop the check for which pchan the dyn TS is operating as, because a rel_act_kind == LCHAN_REL_ACT_PCU implies that it is either already in or trying to become PDCH mode. Change-Id: I57ba60c670730c6d7877a6a9b96ece0a7679a0bb ----------------------------------------------------------------------- hooks/post-receive -- Osmocom BTS-side code (Abis, scheduling, ...)