osmo-hlr[master]: rewrite subscriber_update_notify() without calls into luop

This is merely a historical archive of years 2008-2021, before the migration to mailman3.

A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/gerrit-log@lists.osmocom.org/.

Neels Hofmeyr gerrit-no-reply at lists.osmocom.org
Tue Apr 17 21:34:17 UTC 2018


Patch Set 2:

Thinking back about this, I told you two things that might have been a bit confusing.

One suggested solution is to use the "SGSN" string to determine whether the conn is PS. If we do that, it should be evaluated as soon as the conn gets established, i.e. in osmo_gsup_server_accept_cb().

The other suggested solution is to use the cn_domain indicator we receive as an IE during a lu_op. In this solution, we don't parse the "SGSN" string at all, but record the PS-ness when we receive the first LU, assuming that each GSUP client is always either PS or CS. For this solution, lu_op.c is indeed the right place, but then we don't parse the string but use the cn_domain received in the GSUP message.

Extending on the second solution: we may want to allow a GSUP client to handle both PS and CS, so we could record PS-ness and CS-ness independently, so that a client can be both CS and PS at the same time. This is so far unused, but thinking of GSUP proxying or stuff like that which we might want to do at some point is worth considering. The implication so far is that a client that has never sent a PS lu_op simply doesn't get notified of APN changes, AFAICT. Seems sane enough.

-- 
To view, visit https://gerrit.osmocom.org/7743
To unsubscribe, visit https://gerrit.osmocom.org/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I06c43ece2b48dc63d599000eb6d6d51e08963067
Gerrit-PatchSet: 2
Gerrit-Project: osmo-hlr
Gerrit-Branch: master
Gerrit-Owner: Stefan Sperling <ssperling at sysmocom.de>
Gerrit-Reviewer: Harald Welte <laforge at gnumonks.org>
Gerrit-Reviewer: Jenkins Builder
Gerrit-Reviewer: Neels Hofmeyr <nhofmeyr at sysmocom.de>
Gerrit-Reviewer: Stefan Sperling <ssperling at sysmocom.de>
Gerrit-Reviewer: neels <nhofmeyr at sysmocom.de>
Gerrit-HasComments: No



More information about the gerrit-log mailing list