Attention is currently required from: arehbein.

pespin has posted comments on this change. ( https://gerrit.osmocom.org/c/osmo-bsc/+/31880?usp=email )

Change subject: Make NSE timing data configurable ......................................................................

Patch Set 2:

(5 comments)

Patchset:

PS2: Again this should be per BTS.

File include/osmocom/bsc/bts.h:

https://gerrit.osmocom.org/c/osmo-bsc/+/31880/comment/81c13d98_bee9d723 PS2, Line 326: #define GSM_BTS_TDEF_ID_TNS_BLOCK_NEG29 (-29) Better leave some space in the number series so that we can later on add similar timers next to the existing ones.

File src/osmo-bsc/bts_ipaccess_nanobts_omlattr.c:

https://gerrit.osmocom.org/c/osmo-bsc/+/31880/comment/cb9190bc_b2e5811e PS2, Line 187: struct osmo_tdef *tdefs = bts->network->T_defs; Again, this needs to be per BTS.

File src/osmo-bsc/bts_vty.c:

https://gerrit.osmocom.org/c/osmo-bsc/+/31880/comment/bf555dbd_2e2108c9 PS2, Line 76: static const int gprs_ns_timer_tdef_data[][2] = { AFAICT this is only used in cfg_bts_gprs_ns_timer_cmd, so it can be moved there.

https://gerrit.osmocom.org/c/osmo-bsc/+/31880/comment/63253556_aa03e90e PS2, Line 1627: int idx = get_string_value(gprs_ns_timer_strs, argv[0]); so you are converting to an index here once, and in line 1637 again. This looks ptoentially wrong. Why isn't this get_string_value providing the interesting idx directly?