pcu: Fix race condition in TC_n3101_max_t3169 and TC_n3103_max_t3169
I cannot reproduce the issue in my local system with docker, probabluy
due to different performance from jenkins builder. It triggers almost
every time in jenkins nightly TTCN3 PCU_tests though.
A similar fix was already introduced recently for TC_n3105_max_t3195
since PCU started submitting empty (idle) blocks when there's no TBF
listening on that PDCH. We need to update the test to account for that,
since our TTCN3 timers are not necessarily exactly matching the one at
the PCU side.
Change-Id: Ia5344df15c612c70a6cdd7bb6f12dc7524a23bf4
diff --git a/pcu/PCU_Tests.ttcn b/pcu/PCU_Tests.ttcn
index c347761..5bc2959 100644
--- a/pcu/PCU_Tests.ttcn
+++ b/pcu/PCU_Tests.ttcn
@@ -1524,6 +1524,9 @@
block_nr := nr.blk_nr));
repeat;
}
+ /* We may already receive empty (idle) blocks before our own TTCN3 timer
+ * triggers due to the TBF being released. Keep going until our T_3169 triggers. */
+ [mp_osmo_pcu_newer_than_0_9_0 and n3101 == N3101_MAX + 1] as_pcuif_rx_ignore_empty(nr);
[] T_3169.timeout {
log("T_3169 expired");
/* Done in alt */
@@ -1622,6 +1625,7 @@
f_shutdown(__BFILE__, __LINE__);
}
[] as_ms_rx_ignore_dummy(ms, nr);
+ [] as_pcuif_rx_ignore_empty(nr);
[T_3169.running] T_3169.timeout {
log("T_3169 timeout");
/* Done in alt, wait for pending RTS initiated previously in