All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=757993

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=04c87f2c825bd43a2057adc358a9579512e58a482a01836c68ee5b5c1b90673d


Bot(s) for this bug's original alert(s):

chromium-rel-mac12-mini-8gb

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970522724186265712

=== BISECT JOB RESULTS ===
Bisect was unable to run to completion

Please try rerunning the bisect.


If failures persist contact the team (see below) and report the error.


Bisect Details
  Configuration: mac_10_12_mini_8gb_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstMeaningfulPaint_avg/pcv1-cold/HTML5Rocks

Revision             Result                  N
chromium@494316      70.9484 +- 30.0245      14      good
chromium@494969      1218.75 +- 6245.02      14      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=HTML5Rocks loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970522724186265712


For feedback, file a bug with component Speed>Bisection

+simonhatch: this timed out after 24 hours--what's the easiest way to see how far it narrowed down?

+kouhei: looking at random bisect results, this seems really bimodal: https://luci-logdog.appspot.com/v/?s=chrome%2Fbb%2Ftryserver.chromium.perf%2Fmac_10_12_mini_8gb_perf_bisect%2F299%2F%2B%2Frecipes%2Fsteps%2FBisecting_revision__32_%2F0%2Fsteps%2FBisecting_revision.Compare_samples__264_%2F0%2Flogs%2Fjson.output%2F0

Any idea what's going on?

Unfortunately it only managed to run the good/bad reference values. None of the subsequent steps generated the metric. The test seems to have run fine, but looking at the chartjson there were errors:

"none_value_reason": "Merging values containing a None value results in a None value. None values: [ScalarValue(HTML5Rocks, timeToFirstMeaningfulPaint_max, ms, None, important=False, description=time to first meaningful paint, tir_label=pcv1-cold, improvement_direction=down, grouping_keys={'cache_temperature': 'pcv1-cold'}]",
 

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970369622219256528

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970369610180432816

Assigning to kouhei, test owner, since the test doesn't produce data when bisecting.

I tried kicking off some bisects on other pages, so we can get kernel some data about whether his CL caused this regression on mac 10.12

=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Greg Kerr
  Commit : 9705b7aaabf4c81d4e9cb13174338df9bde61cc4
  Date   : Tue Aug 15 19:42:42 2017
  Subject: Enable Finch Trial Test for MacV2Sandbox.

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstMeaningfulPaint_avg/pcv1-cold/Ynet
  Change       : 15.71% | 436.176285714 -> 496.564333333

Revision             Result                  N
chromium@494401      436.176 +- 129.573      14      good
chromium@494466      439.298 +- 141.779      14      good
chromium@494498      428.761 +- 102.826      9       good
chromium@494500      445.183 +- 138.775      14      good
chromium@494501      413.684 +- 30.6769      6       good
chromium@494502      504.69 +- 110.016       9       bad       <--
chromium@494506      489.188 +- 119.107      14      bad
chromium@494514      487.986 +- 102.633      9       bad
chromium@494530      480.554 +- 97.5343      14      bad
chromium@494658      496.564 +- 63.1611      6       bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Ynet loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970369610180432816


For feedback, file a bug with component Speed>Bisection

Whoops, somehow ran a bisect on mac retina, which is 10.11, and the question is whether this regresses 10.12. Re-opening to kick off a few more bisects on 10.12

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970347240446051568

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970347153058223904

I reverted commit 9705b7aaabf4c81d4e9cb13174338df9bde61cc4, is it too late to run this bisect now?

The bisect still runs. It's completely unaware of what happened outside the range it's bisecting. Hopefully one of these will come back with some clear results.

you can also try perf trybots: https://chromium.googlesource.com/chromium/src/+/master/docs/speed/perf_trybots.md

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: mac_10_12_mini_8gb_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstContentfulPaint_avg/pcv1-cold/REI

Revision             Result                  N
chromium@494360      970.455 +- 2456.49      21      good
chromium@494538      982.639 +- 4332.92      18      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=REI loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970347240446051568


For feedback, file a bug with component Speed>Bisection

=== BISECT JOB RESULTS ===
Bisect was unable to run to completion

Error: INFRA_FAILURE

The bisect was able to narrow the range, you can try running with:
  good_revision: b77a942fbeaf003eb19e19372a6ae8f1d0d2f33d
  bad_revision : 677e791073d90814ce47b4e285e0b68d8d110ac3

If failures persist contact the team (see below) and report the error.


Bisect Details
  Configuration: mac_10_12_mini_8gb_perf_bisect
  Benchmark    : loading.desktop
  Metric       : cpuTimeToFirstMeaningfulPaint_avg/pcv1-cold/Bhaskar

Revision             Result                  N
chromium@494360      271.278 +- 106.376      9      good
chromium@494449      272.465 +- 85.1918      9      good
chromium@494494      271.245 +- 72.3742      9      good
chromium@494516      295.944 +- 73.1425      9      bad
chromium@494538      301.316 +- 86.0324      7      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Bhaskar loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970347153058223904


For feedback, file a bug with component Speed>Bisection

=== BISECT JOB RESULTS ===
Bisect was unable to run to completion

Please try rerunning the bisect.


If failures persist contact the team (see below) and report the error.


Bisect Details
  Configuration: mac_10_12_perf_bisect
  Benchmark    : loading.desktop
  Metric       : cpuTimeToFirstMeaningfulPaint_avg/pcv1-cold/PremierLeague

Revision             Result                  N
chromium@494315      34.811 +- 7.84858       14      good
chromium@494854      76.1177 +- 73.6844      14      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=PremierLeague loading.desktop

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970369622219256528


For feedback, file a bug with component Speed>Bisection

Almost all the regressions have been triaged out of this bug, and the one that is left doesn't repro.