diff options
| author |  Christian Couder <christian.couder@gmail.com> | 2018-05-04 14:36:36 +0200 |
|---|---|---|
| committer |  Junio C Hamano <gitster@pobox.com> | 2018-05-06 13:04:54 +0900 |
| commit | d9ea451ab6c11d548d3a11183fdbce49385f2e56 (patch) | |
| tree | 1c818fa7fe49c5d0319e97141a930b4fdb982f6f /t/perf/bisect_run_script | |
| parent | t/perf: add scripts to bisect performance regressions (diff) | |
| download | tgif-d9ea451ab6c11d548d3a11183fdbce49385f2e56.tar.xz | |
perf/bisect_run_script: disable codespeed
When bisecting a performance regression using a config file,
`./bisect_regression --config my_perf.conf` for example, the
config file can contain Codespeed configuration which would
instruct the 'aggregate.perl' script called by the 'run'
script to output results in the Codespeed format and maybe
to try to send this output to a Codespeed server.
This is unfortunate because the 'bisect_run_script' relies
on the regular output from 'aggregate.perl' to mesure
performance, so let's disable Codespeed output and sending
results to a Codespeed server.
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Diffstat (limited to 't/perf/bisect_run_script')
| -rwxr-xr-x | t/perf/bisect_run_script | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/t/perf/bisect_run_script b/t/perf/bisect_run_script index 038255df4b..3ebaf15521 100755 --- a/t/perf/bisect_run_script +++ b/t/perf/bisect_run_script @@ -24,6 +24,12 @@ result_file="$info_dir/perf_${script_number}_${bisect_head}_results.txt" GIT_PERF_DIRS_OR_REVS="$bisect_head" export GIT_PERF_DIRS_OR_REVS +# Don't use codespeed +GIT_PERF_CODESPEED_OUTPUT= +GIT_PERF_SEND_TO_CODESPEED= +export GIT_PERF_CODESPEED_OUTPUT +export GIT_PERF_SEND_TO_CODESPEED + ./run "$script" >"$result_file" 2>&1 || die "Failed to run perf test '$script'" rtime=$(sed -n "s/^$script_number\.$test_number:.*\([0-9]\+\.[0-9]\+\)(.*).*\$/\1/p" "$result_file") |