Attention is currently required from: neels, laforge, fixeria, pespin.

View Change

10 comments:

• Patchset:

  • Patch Set #16:

    Hi all,

    Thanks again for the feedback. I think we're getting much closer. This version still needs additional testing on my side to confirm that I can still accurately see both per-lchan instance duration and total duration per lchan type.

    I will keep testing this and await further feedback.

    Regards,
    -Michael

• File src/osmo-bsc/lchan_fsm.c:

  • Patch Set #14, Line 465: struct rate_ctr *active_ms_rate_ctr;

    You could simply set it to NULL here and keep the code unchanged, but not critical.

    Done

• File src/osmo-bsc/lchan_fsm.c:

  • Patch Set #15, Line 208: if (lchan->active_ms.cfg.rate_ctr) {

    I understand now: the rate_ctr is set only for SDCCH and TCH type.

    Done

  • Patch Set #15, Line 209: osmo_time_cc_cleanup(&lchan->active_ms);

    osmo_time_cc_cleanup() is not necessary, the forget_sum configuration does this implicitly

    Done

  • Patch Set #15, Line 465: uint

    I like to use "uint" in private projects, but there i typedef it manually as unsigned int. […]

    Done

  • Patch Set #15, Line 478: counter_id = -1;

    could, yes, but totally fine here IMHO. […]

    Done

  • Patch Set #15, Line 479: switch (lchan->type) {

    the lchan->type changes during operation of osmo-bsc, it is not known at the time of lchan_fsm_alloc […]

    Done

  • Patch Set #15, Line 493: .gran_usec = 1*1000,

    Sorry for my earlier comment, I've come to realize what exactly I recommended by mentioning 1 ms gra […]

    Done

  • Patch Set #15, Line 498: .T_forget_sum = -18,

    timers X16, X17, X18 are defined to configure a different counter (the all_allocated counters). […]

    Done

  • Patch Set #15, Line 562: /* Stop the timekeeper, which triggers a report */

    the comment suggests that the report is triggered only now. […]

    Done

To view, visit change 27081. To unsubscribe, or for help writing mail filters, visit settings.