Attention is currently required from: fixeria, laforge, pespin.

lynxis lazus has posted comments on this change by lynxis lazus. ( https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/38396?usp=email )

Change subject: SGSN: BSSGP_ConnHdlr: GMM Service Request: add support to expect authentication ......................................................................

Patch Set 5:

(1 comment)

File sgsn/BSSGP_ConnHdlr.ttcn:

https://gerrit.osmocom.org/c/osmo-ttcn3-hacks/+/38396/comment/d59f58a3_d2952... : PS4, Line 358: as_gmm_auth

I don't want to block here, but yet I find this problematic. […]

as_gmm_auth() is using f_gmm_auth_as() to not duplicate code. f_gmm_auth_as() is a private function and not exported. It is not a 3rd api.

as_gmm_auth() vs f_gmm_auth(): f_gmm_auth: - requires an authentication if it is global enabled - forces + requires the order of - SAI Req - SAI Resp - Auth Req - Auth Resp - CommonId (if Iu enabled) - out of order: answer as_mm_identity() requests

as_gmm_auth(): - ignores the order of requests - answers SAI Req (if enabled) - answers GMM Auth Req

The old (or latest) SGSN won't do auth for many steps, even I think it should be mandatory here, while the new SGSN will handle it much better and do an Auth Req. To handle both gracefully (and I think) it doesn't matter for the testcase PDP Request time out, if an auth happened or not, just do it. While for other test case, e.g. Attach Request, it matters much more.

I don't see how you can migrate all users to f_gmm_auth and I think adding order to the alt step, makes it much unreadable (not 100% sure if it is even possible, because we would use global state here).