Bug #4501
closedosmo-bts-{sysmo,oc2g,litecell15}: Packet Control Ack (in form of Access Bursts) arrives on PRACH?!?
100%
Description
As was reported by keith (originally in #1526), with 'gprs control-ack-type-rach' enabled in the BSC, osmo-pcu logs:
DL1IF <0001> ../../git/src/osmo-bts-sysmo/sysmo_l1_if.c:251 Rx PH-RA.ind for unknown L1 SAPI PRACH
this is a regression introduced in I482d60a46b9d253dfe0b16140eac9fea6420b30c:
Parent: 17954da5 (pcuif_proto.h: extend RACH.ind with TRX / TS numbers) Author: Vadim Yanitskiy <axilirator@gmail.com> AuthorDate: 2019-10-05 23:45:31 +0700 Commit: Vadim Yanitskiy <axilirator@gmail.com> CommitDate: 2019-11-23 17:42:45 +0700 PTCCH: properly handle RACH.ind for PCU_IF_SAPI_PTCCH Change-Id: I482d60a46b9d253dfe0b16140eac9fea6420b30c Related: OS#1545
I would expect Packet Control Ack on GsmL1_Sapi_Pdtch or rather on GsmL1_Sapi_Pacch, but apparently (for some magic reason) it arrives on GsmL1_Sapi_Prach.
According to 3GPP TS 45.002, section 3.3.3.2.1:
- PRACH stands for Packet Random Access CHannel,
- used to request assignment of one or several PDTCHs,
- belongs to Packet Common Control Channel (PCCCH)!
Therefore the Packet Control Ack has nothing to do with PRACH. Neither we support PCCCH (it has been deprecated by the specifications).
GsmL1_LogChComb_XIII ///< PDTCH/F + PACCH/F + PTCCH/F
This odd behaviour seems to be caused by the following code (see #4500):
static const struct sapi_dir pdtch_sapis[] = {
{ GsmL1_Sapi_Pdtch, GsmL1_Dir_TxDownlink },
{ GsmL1_Sapi_Pdtch, GsmL1_Dir_RxUplink },
{ GsmL1_Sapi_Ptcch, GsmL1_Dir_TxDownlink },
{ GsmL1_Sapi_Prach, GsmL1_Dir_RxUplink }, // <-- ?!?
#if 0
{ GsmL1_Sapi_Ptcch, GsmL1_Dir_RxUplink },
{ GsmL1_Sapi_Pacch, GsmL1_Dir_TxDownlink }, // <-- ?!?
#endif
};
where we enable GsmL1_Sapi_Prach on GsmL1_LogChComb_XIII (again, there is no PRACH in XIII). I guess we should enable GsmL1_Sapi_Pacch on direction GsmL1_Dir_RxUplink instead.
For now I submitted a blind fix: https://gerrit.osmocom.org/c/osmo-pcu/+/17669. Let's keep it WIP for now.
Related issues
Updated by fixeria over 4 years ago
- Related to Bug #4500: osmo-bts-{sysmo,oc2g,litecell15}: PTCCH/U is not enabled on PDCH timeslots added
Updated by fixeria over 4 years ago
As it turns out, there is a nice (documented) feature that allows the DSP to send debug messages to /dev/rtfifo/dsp_trace. Here is what I see when trying to enable GsmL1_Sapi_Pacch on GsmL1_Dir_RxUplink:
MphActivateReq => [ hLayer1 = 0x87080020, hLayer2 = 0x600BB, u8Tn = 6, sapi = 16, dir = 2, subCh = 31, bfi = -200.00 dBm cfgParam = 00000000 00000000 00000000 00000000 ] [ERROR] : LogChMng_GetTypeXIIIBufAddr() => Unsupported sapi in U/L [logChCombXIII, sapi = 16 ] [ERROR] : Device_ActivateLogCh()=> Unable to activate. [pPhyCh = 0x87089AB0, pLogChMngBuf = 0x870801FC, pi8Buffer = 0x87089F90] MphActivateCnf => [ hLayer3 = 0x600BB, hLayer1 = 0x87080020, TN = 6, sapi = 16, dir = 2, subCh = 31, status = -4 ]
same error when enabling on GsmL1_Dir_TxDownlink:
MphActivateReq => [ hLayer1 = 0x87080020, hLayer2 = 0x600BB, u8Tn = 6, sapi = 16, dir = 4, subCh = 31, bfi = -200.00 dBm cfgParam = 00000000 00000000 00000000 00000000 ] [ERROR] : LogChMng_GetTypeXIIIBufAddr() => Unsupported sapi in D/L [logChCombXIII, sapi = 16 ] [ERROR] : Device_ActivateLogCh()=> Unable to activate. [pPhyCh = 0x87089AB0, pLogChMngBuf = 0x870801FC, pi8Buffer = 0x87089F90] MphActivateCnf => [ hLayer3 = 0x600BB, hLayer1 = 0x87080020, TN = 6, sapi = 16, dir = 4, subCh = 31, status = -4 ]
This is odd and confusing. From 3GPP TS 45.002, section 6.4.1 "Permitted channel combinations onto a basic physical channel":
xiii) PDTCH/F + PACCH/F + PTCCH/F
Updated by fixeria almost 4 years ago
- Status changed from Stalled to Resolved
- % Done changed from 0 to 100
I would expect Packet Control Ack on GsmL1_Sapi_Pdtch or rather on GsmL1_Sapi_Pacch, but apparently (for some magic reason) it arrives on GsmL1_Sapi_Prach.
We have to deal with that, as we cannot change the DSP. Let's handle GsmL1_Sapi_Prach in handle_ph_ra_ind().
The patch (https://gerrit.osmocom.org/c/osmo-pcu/+/17669) has been merged meanwhile.