Patch Set 4: (2 comments)
> (2 comments) https://gerrit.osmocom.org/#/c/5157/1/include/osmocom/bsc/osmo_bsc_mgcp.h File include/osmocom/bsc/osmo_bsc_mgcp.h: Line 47: enum gsm48_chan_mode chan_mode; > (makes me think ... this requires that we only have a single MGCP transacti Yes, since we execute them one after another. This will also be the case in the future. I think this won't hurt in performance and we benefit from a clear structured FSM. https://gerrit.osmocom.org/#/c/5157/1/src/osmo-bsc/osmo_bsc_mgcp.c File src/osmo-bsc/osmo_bsc_mgcp.c: Line 825: mgcp_client_cancel(mgcp, mgcp_ctx->mgw_pending_trans); > Can it happen that no mgw_pending_trans was set? If yes, we might cancel an It should not because we do this only when an MGW transaction timed out, so that should be ok. Maybe we can define a TRANS_ID_INVALID for mgcp_trans_id_t so that we can initalize the variable before we start, but I think that is not necessary. -- To view, visit https://gerrit.osmocom.org/5157 To unsubscribe, visit https://gerrit.osmocom.org/settings Gerrit-MessageType: comment Gerrit-Change-Id: I40794dff7d10e2b6a96863a2da7e9fbd5662a1bf Gerrit-PatchSet: 4 Gerrit-Project: osmo-bsc Gerrit-Branch: master Gerrit-Owner: dexter <[email protected]> Gerrit-Reviewer: Harald Welte <[email protected]> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: Neels Hofmeyr <[email protected]> Gerrit-Reviewer: dexter <[email protected]> Gerrit-HasComments: Yes
