Commit e13abda4 authored by Mike Perry's avatar Mike Perry
Browse files

Bug 6647: Use correct scale constant and prevent rounding error 

We were effectively resetting our counts, and the rounding error
leads to incorrect log messages.
parent 4950618b

src/or/circuitbuild.c

+7 −2
Changes for src/or/circuitbuild.c: 7 added lines, 2 removed lines.
Original line number Diff line number Diff line
@@ -2601,12 +2601,12 @@ pathbias_get_scale_threshold(const or_options_t *options) 
static int

pathbias_get_scale_factor(const or_options_t *options)

{

#define DFLT_PATH_BIAS_SCALE_FACTOR 4

#define DFLT_PATH_BIAS_SCALE_FACTOR 2

  if (options->PathBiasScaleFactor >= 1)

    return options->PathBiasScaleFactor;

  else

    return networkstatus_get_param(NULL, "pb_scalefactor",

                                DFLT_PATH_BIAS_SCALE_THRESHOLD, 1, INT32_MAX);

                                DFLT_PATH_BIAS_SCALE_FACTOR, 1, INT32_MAX);

}



static const char *
@@ -2859,6 +2859,10 @@ entry_guard_inc_first_hop_count(entry_guard_t *guard) 
  /* If we get a ton of circuits, just scale everything down */

  if (guard->first_hops > (unsigned)pathbias_get_scale_threshold(options)) {

    const int scale_factor = pathbias_get_scale_factor(options);

    /* For now, only scale if there will be no rounding error...

     * XXX024: We want to switch to a real moving average for 0.2.4. */

    if ((guard->first_hops % scale_factor) == 0 &&

        (guard->circuit_successes % scale_factor) == 0) {

      log_info(LD_PROTOCOL,

               "Scaling pathbias counts to (%u/%u)/%d for guard %s=%s",

               guard->circuit_successes, guard->first_hops,
@@ -2867,6 +2871,7 @@ entry_guard_inc_first_hop_count(entry_guard_t *guard) 
      guard->first_hops /= scale_factor;

      guard->circuit_successes /= scale_factor;

    }

  }

  guard->first_hops++;

  log_info(LD_PROTOCOL, "Got success count %u/%u for guard %s=%s",

           guard->circuit_successes, guard->first_hops, guard->nickname,