* HACK: Don't retransmit InitialContextSetupReq
Related: #256
* HACK: Don't use buggy sa1p_copy() in eNBConfigTransfer
Related: #257
* mme: don't reject with 'IMSI is unknown in HLR' (permanent reject)
* MME: Implement S6a result -> EMM cause code mapping
Closes: #263
* Spencer: modification to Haralds fix because macros are now renamed
* MME: don't assert on MAC failures of uplink NAS frames
Closes: #267
* MME: Avoid ogs_assert() in many situations
We don't want to crash the entire program just because a message
received from an external entity didn't match some of our expectations.
* compiles fine, checked DIFFs and only difference is the ogs_assert -> ogs_expect
The check for whether the AVP exists in the ULA was wrong, resulting in
the MME refusing to proceed without this AVP. However, it is not a
mandatory AVP.
At least in version 11.8.0 Release 11 of 3GPP TS 29.272,
Subscribed-Periodic-RAU-TAU-Timer is only permitted inside
Subscription-Data. Before this patch, the MME expects it outside, i.e.
on the "main" level of the DIAMETER AVP tree.
Closes: #238
Sukchan Lee
Spencer Sevilla
Harald Welte