On 04 Jan 2016, at 18:43, Jacob Erlbeck jerlbeck@sysmocom.de wrote:
Dear Jacob,
diff --git a/openbsc/src/gprs/gprs_gmm.c b/openbsc/src/gprs/gprs_gmm.c index 6e7e5f1..212c7d7 100644 --- a/openbsc/src/gprs/gprs_gmm.c +++ b/openbsc/src/gprs/gprs_gmm.c @@ -1172,13 +1172,33 @@ static int gsm48_rx_gmm_ra_upd_req(struct sgsn_mm_ctx *mmctx, struct msgb *msg, * if the TLLI matches foreign_tlli (P-TMSI). Note that this * is an optimization to avoid the RA reject (impl detached) * below, which will cause a new attach cycle. */
- }
the todo above reads:
/* TODO: Check if there is an MM CTX with old_ra_id and * the P-TMSI (if given, reguired for UMTS) or as last resort * if the TLLI matches foreign_tlli (P-TMSI). Note that this * is an optimization to avoid the RA reject (impl detached) * below, which will cause a new attach cycle. */
I think this todo is addressed with sgsn_mm_ctx_by_tlli_and_ptmsi? Can the comment be removed?
What about the test? Do we have one that gets the "XID RESET" we expect?
holger