Attention is currently required from: neels.
4 comments:
File src/libosmo-mgcp/mgcp_network.c:
Patch Set #4, Line 838: osmo_sockaddr_port(&conn->end.addr.u.sa) == 0) {
the most concise restriction would be: […]
We need to keep it this way I submitted, at least for address, since the configured address at this point may be wrong (the hnb may actually be using another IP address, what osmo-mgw was submitted was just a guess, and it cannot be considered confirmed until IuUP Initialization happens).
So if IP address is set but port is not, we cannot validate IP address of received packet against the configured one, because it is "a guess", we must accept it anyway, at least to carry on with IuUP initialization.
Patch Set #4, Line 841: announce
"not all hNodeB", but some are happily sending RAB Assignment success before IuUP Initialization. […]
that's why I didn't write "not all hNodeb", to leave it undetailed. I can add "some", but since we didn't play with many others I preferred leaving it this way. I think even in the diagrams in specs it shows up this way (IuUP Initialization happening before RAB-Ass-Resp).
Do you know of any HNB not doing it this way?
(RAB Assignment Response at HNBGW)
Ack
Writing "MDCX" is too specific: it is up to the client to do CRCX with SDP all-in-one or adjust with […]
Ack
To view, visit change 35176. To unsubscribe, or for help writing mail filters, visit settings.