Bug #3018
closedOsmoBTS silently ignores PCU_IF_MSG_PAG_REQ
0%
Description
In pcu_socket.c we have the following "interesting" code snippet in pcu_rx_data_req():
»·······switch (data_req->sapi) { »·······case PCU_IF_SAPI_PCH: »·······»·······if (msg_type == PCU_IF_MSG_PAG_REQ) { »·······»·······»·······/* FIXME: Add function to schedule paging request. »·······»·······»······· * This might not be required, if PCU_IF_MSG_DATA_REQ »·······»·······»······· * is used instead. */ »·······»·······} else { »·······»·······»·······struct gsm_bts_role_bts *btsb = bts->role; »·······»·······»·······paging_add_imm_ass(btsb->paging_state, data_req->data, »·······»·······»·······»·······data_req->len); »·······»·······}
which means that any PCU_IF_MSG_PAG_REQ on PCU_IF_SAPI_PCH are silently discarded without ever beeing transmitted or any log message about it.
Updated by stsp almost 6 years ago
The PCU_IF_MSG_PAG_REQ message type is not sent by osmo-pcu at present.
This message type is only used during the pcu_sock Rx path by osmo-pcu.
What the comment alludes to is the fact that paging requests
are sent by osmo-pcu inside PCU_IF_MSG_DATA_REQ messages.
I agree logging the event would be worthwhile in case someone
tries to send PCU_IF_MSG_PAG_REQ to osmo-bts in the future.
But there is no real problem here. The comment is just vague.
issue #2404 documents an actual problem which prevents paging via
osmo-pcu from working (disagreement about the format of
mobdileIdenity1 in type1 paging messages).
Updated by stsp almost 6 years ago
Patch which adds logging: https://gerrit.osmocom.org/#/c/osmo-bts/+/9727/
Updated by stsp almost 6 years ago
- Status changed from New to Resolved
Above patch has been merged. Closing this ticket as per above explanation of current behaviour.