Changes

bts: fix pchan config in default mp_trx_pars[]

The timeslot 3 of transceiver 2 is configured to PDCH in osmo-bsc.cfg,
so the pchan config must be GSM_PCHAN_PDCH, not GSM_PCHAN_TCHF.

Change-Id: I7de4e2796ca91e90f9261be8355e97e60a9cc13f
Fixes: 42d8bd501 ("BTS_Tests: re-organize configuration of TRX/TS params")

bts: document the default FDMA/TDMA configuration

Change-Id: I988f783c1e91c916982609c0f30088b00d00674e

RAN_Emulation: Support Tx RESET retries in RANAP

As already done in BSSAP in the same module.

Change-Id: I84553ae7aee19efa007bae060d7f7ca218a5f306

BSSAP_LE_Emulation: Support Tx RESET retries

Have same code as we use in RAN_Emulation.ttcnpp.

Change-Id: I475f93ff0b8cbfa5aa4900d02ddcb2e97fa5ca6a

BSSAP(_LE)/RANAP: Proper counting of RESET retries

N retries means N+1 attempts.

Since M3UA_Emulation (in deps/, not in library) doesn't support pushing
DUNA/DAVA as primitive up the stack, we cannot implement the related
SCCP primitives which should end up sending in turn other primitives to
our BSSAP(_LE)/RANAP emulations.
This means we currently have no way  to figure out the status of the
underlying SCCP/MTP link, and our only way is implement a poor man's
solution: retransmit RESET until we get an answer.

Since before bssap_reset_retries was set to 1 everywhere and we were
only attempting once, it could happen that the other end (the IUT) was
not yet connected to the STP when we send the RESET, and hence it is
discarded the STP. In that scenario, we'd time out and fail the test.
Instead, now that we do N+1 attempts, we retry after 5 seconds, which
should be enough time for the peer to reconnect to the STP.

Change-Id: If02e424672ff6270aed323c3bb236e822d8a39a4