All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=734642

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtpK47ggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtsWpowoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtrHn-QoM


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

chromium-rel-mac11
chromium-rel-mac11-pro
chromium-rel-mac12-mini-8gb

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976338530774470080

=== Auto-CCing suspected CL author binji@chromium.org ===

Hi binji@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


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

Suspected Commit
  Author : Ben Smith
  Commit : 284a4804f2f2cab172d2b4d96048b5d8986c72cd
  Date   : Tue Jun 13 23:45:44 2017
  Subject: [SAB] Move creation of SharedArrayBuffer/Atomics to InitializeGlobal

Bisect Details
  Configuration: mac_pro_perf_bisect
  Benchmark    : memory.desktop
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/TrivialAnimationPageSharedPageState
  Change       : 10.23% | 3191050.66667 -> 3517422.66667

Revision                           Result                  N
chromium@479381                    3191051 +- 608.281      6      good
chromium@479385                    3190971 +- 187.503      6      good
chromium@479385,v8@aea68b5a26      3188501 +- 203.227      6      good
chromium@479385,v8@5d7039eac3      3188540 +- 159.9        6      good
chromium@479385,v8@284a4804f2      3546041 +- 200614       6      bad       <--
chromium@479385,v8@7e6ed62071      3582695 +- 253602       6      bad
chromium@479386                    3655923 +- 253646       6      bad
chromium@479387                    3619333 +- 269085       6      bad
chromium@479388                    3582759 +- 253658       6      bad
chromium@479395                    3619295 +- 268996       6      bad
chromium@479409                    3546101 +- 200548       6      bad
chromium@479437                    3517417 +- 152.385      6      bad
chromium@479493                    3517423 +- 84.2219      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=TrivialAnimationPageSharedPageState memory.desktop

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976338530774470080

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5332995105357824


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection.  Thank you!

 Issue 734640  has been merged into this issue.

I took a look at some of these graphs. It's interesting, since for some of the graphs it's just returning the memory usage to the same level it was before the previous change (https://chromium.googlesource.com/chromium/src/+/e94f5343100836bcdc838dc4da4fd13d4c69a950 (position 471385) or the reland https://chromium.googlesource.com/chromium/src/+/ac2ac7b771c7d5838b4e2e35523bf38d215ac926 (position 471460)).

This makes me wonder if the values are just a measurement error: the previous change shouldn't have improved memory usage (it made it worse in most other examples).