Attention is currently required from: pespin.

dexter has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-bsc/+/34192 )

Change subject: pcuif_proto: check confirm flag in struct gsm_pcu_if_pch ......................................................................

Patch Set 5:

(4 comments)

File src/osmo-bsc/abis_rsl.c:

https://gerrit.osmocom.org/c/osmo-bsc/+/34192/comment/0f8af513_5a38409b PS4, Line 1013: msgb_put_u32(msg, msg_id);

Just curious, did you check if Ericsson expects this ERIC_MOBILE_ID to actually be the TLLI? Or it's […]

I have checked this back, when I get everything right, then it is just an identifier that is loosely connected to a specific MS.

To be 100% sure that it is really just a number and not required to be the TLLI i did an experimet with osmo_swab32() in abis.rsl. It still works the same.

File src/osmo-bsc/pcu_sock.c:

https://gerrit.osmocom.org/c/osmo-bsc/+/34192/comment/d4e2763f_b09aaf9b PS4, Line 557: } else {

...

else is not generally useful after a break or return […]

Done

https://gerrit.osmocom.org/c/osmo-bsc/+/34192/comment/61005bb8_5db563e9 PS4, Line 558: if (pcu_rx_rr_paging_pch(bts, pag_grp, pch))

...

code indent should use tabs where possible […]

Done

https://gerrit.osmocom.org/c/osmo-bsc/+/34192/comment/1cb18da4_508905ad PS4, Line 561: return pcu_tx_pch_confirm(bts, pch->msg_id);

why are you confirming it here? You should be confirming it when scheduling it / sending it to the B […]

When pcu_rx_rr_paging_pch() returns, the message is sent out via RSL (in theory, in practice everything should happen in the next mainloop cycle). The BTS however, may have of course a paging queue which may introduce a delay, so it is not as accurate as in osmo-bts.