diff options
| author | Mike Perry <mikeperry-git@fscked.org> | 2010-07-06 08:49:50 -0700 |
|---|---|---|
| committer | Mike Perry <mikeperry-git@fscked.org> | 2010-07-06 12:11:13 -0700 |
| commit | 7bbdf71a82bb064b2ca0eb3a296b4abab6b5ff2b (patch) | |
| tree | bde6a589e599d25866d7d6c7a357c33601d1bfcb /src/or/or.h | |
| parent | 6bb8747ffbd18e08f9724bb8e4f2a16f22ed6612 (diff) | |
| download | tor-7bbdf71a82bb064b2ca0eb3a296b4abab6b5ff2b.tar.gz tor-7bbdf71a82bb064b2ca0eb3a296b4abab6b5ff2b.zip | |
Fix unittest failure in bug 1660.
We now record large times as abandoned, to prevent a filter step from
happening and skewing our results.
Also, issue a warn for a rare case that can happen for funky values of Xm or
too many abandoned circuits. Can happen (very rarely) during unit tests, but
should not be possble during live operation, due to network liveness filters
and discard logic.
Diffstat (limited to 'src/or/or.h')
| -rw-r--r-- | src/or/or.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/or/or.h b/src/or/or.h index 5393b27d4d..60987698b4 100644 --- a/src/or/or.h +++ b/src/or/or.h @@ -3188,7 +3188,7 @@ build_time_t circuit_build_times_generate_sample(circuit_build_times_t *cbt, double q_lo, double q_hi); void circuit_build_times_initial_alpha(circuit_build_times_t *cbt, double quantile, double time_ms); -void circuit_build_times_update_alpha(circuit_build_times_t *cbt); +int circuit_build_times_update_alpha(circuit_build_times_t *cbt); double circuit_build_times_cdf(circuit_build_times_t *cbt, double x); void circuit_build_times_add_timeout_worker(circuit_build_times_t *cbt, double quantile_cutoff); |