Attention is currently required from: laforge, neels, osmith.

fixeria has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/36772?usp=email )

Change subject: msc: add TC_gsup_mt_sms_lu_delay for X36 (LU delay) timer ......................................................................

Patch Set 4:

(1 comment)

File msc/MSC_Tests.ttcn:

https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/36772/comment/a578eb01_2be17... PS4, Line 3325: mtc.stop;

Why mtc. […]

Your understanding is correct. This is unfortunately the weak side of TTCN-3: there exists no mechanism that would allow to execute some clean-up code regardless of the testcase execution flow. The is not the only `mtc.stop` that can abort the MTC and leave the IUT in 'unclean' state. We may also run into this problem if, for instance, `Tguard` fires. Likewise, in `f_expect_clear()` we also do `mtc.stop` if we do not receive CLEAR COMMAND in time.

A more fundamental approach to avoid this would be resetting this timer in `f_init()` or `f_init_handler()`. But this timer is not yet in master, so resetting it there would break all testcases.

I can remove this `mtc.stop`, but this would not solve the problem globally/fundamentally. So I would just keep this as-is and not invest time into that.