Attention is currently required from: jolly, pespin.
falconia has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-bts/+/35132?usp=email )
Change subject: Transmit invalid AMR speech blocks instead of dummy FACCH ......................................................................
Patch Set 2:
(5 comments)
Commit Message:
https://gerrit.osmocom.org/c/osmo-bts/+/35132/comment/501505b4_b36add36 PS2, Line 12: TCH/HS TCH/AHS, not TCH/HS.
File src/osmo-bts-trx/sched_lchan_tchf.c:
https://gerrit.osmocom.org/c/osmo-bts/+/35132/comment/4d3c04d4_04757578 PS2, Line 541: AMR Maybe say TCH/AFS rather than just AMR, for consistency.
https://gerrit.osmocom.org/c/osmo-bts/+/35132/comment/1d489c6e_ac89dd64 PS2, Line 541: * - If the channel mode is AMR, transmit a dummy with speech
I'd document here the rationale about "FACCH displaces two speech frames rather than one".
This C module and function are for TCH/F only, hence the TCH/H-specific problem you are referring to does not apply here. We could keep doing dummy FACCH for all TCH/F modes, but the idea is to be consistent with what we do in TCH/H.
File src/osmo-bts-trx/sched_lchan_tchh.c:
https://gerrit.osmocom.org/c/osmo-bts/+/35132/comment/89f63bda_a3d1555f PS2, Line 453: /* - If the channel mode is TCH/HS or TCH/EFS, transmit a dummy The overall structure of the comment after your change no longer makes sense. Either cover both TCH/HS and TCH/AHS cases in one clause (with "CRC3 or CRC6" wording), or leave the non-AMR TCH/HS clause alone (like you did in TCH/F) and add a new clause for TCH/AHS as in TCH/H version of AMR.
https://gerrit.osmocom.org/c/osmo-bts/+/35132/comment/193d7373_6cc69db0 PS2, Line 453: TCH/EFS If we are keeping the whole comment structure as-is (see my other comment), this part should read TCH/AHS and not TCH/EFS.