Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8975750954318089296

=== BISECT JOB RESULTS ===
Perf regression found but unable to narrow commit range

Build failures prevented the bisect from narrowing the range further.


Bisect Details
  Configuration: win_perf_bisect
  Benchmark    : v8.runtimestats.browsing_desktop
  Metric       : GC:duration_avg/browse_social/browse_social_twitter
  Change       : 26.06% | 54.2968888889 -> 68.4487777778

Suspected Commit Range
  2 commits in range
  https://chromium.googlesource.com/v8/v8.git/+log/6681949808bb89675ab64c95a237ae76955e191b..5d8840e24e0e98136c081ee22441d808412d06e2


Revision                           Result                  N
chromium@481512                    54.2969 +- 29.8465      9        good
chromium@481562                    51.9867 +- 14.3226      6        good
chromium@481575                    55.0522 +- 11.2153      9        good
chromium@481581                    56.8193 +- 47.7507      14       good
chromium@481584                    58.0513 +- 23.6242      9        good
chromium@481586                    54.4176 +- 43.2022      14       good
chromium@481586,v8@6681949808      57.1357 +- 40.211       14       good
chromium@481586,v8@79ec067252      ---                     ---      build failure
chromium@481586,v8@5d8840e24e      69.0916 +- 33.0261      14       bad
chromium@481587                    74.2516 +- 33.3132      9        bad
chromium@481611                    68.4488 +- 18.1263      9        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=browse.social.twitter v8.runtimestats.browsing_desktop

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8975750954318089296

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5794580458373120


| 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!

Bisection didn't succeed, but this looks similar to the other regressions due to the recent GC changes. Can you take a look please?

Assigning to memory sheriff: https://rotation.googleplex.com/index.html#rotation?id=4838401396178944

Also, the bisection shows that it's supposedly one of Ulan's commits as far as I can see.

The regression range seems bogus. Do you remember what windows machine regressed? (there are no links to graphs).

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8975556244509881776

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

Suspected Commit
  Author : Ulan Degenbaev
  Commit : 5d8840e24e0e98136c081ee22441d808412d06e2
  Date   : Thu Jun 22 12:37:26 2017
  Subject: [runtime] Copy descriptors on prototype map transition.

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : v8.runtimestats.browsing_desktop
  Metric       : GC:duration_avg/browse_social/browse_social_twitter
  Change       : 44.26% | 65.002 -> 93.7733333333

Revision                           Result                  N
chromium@481508                    65.002 +- 4.25062       6      good
chromium@481560                    61.5718 +- 29.503       6      good
chromium@481586                    65.632 +- 6.44047       6      good
chromium@481586,v8@6681949808      58.6585 +- 31.9676      6      good
chromium@481586,v8@79ec067252      66.4587 +- 9.20809      6      good
chromium@481586,v8@5d8840e24e      93.4513 +- 5.18235      6      bad       <--
chromium@481587                    92.247 +- 6.30064       6      bad
chromium@481588                    88.6485 +- 10.749       6      bad
chromium@481590                    90.1962 +- 10.2282      6      bad
chromium@481593                    92.7442 +- 4.98309      6      bad
chromium@481599                    91.8987 +- 7.12443      6      bad
chromium@481611                    93.7733 +- 8.40455      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=browse.social.twitter v8.runtimestats.browsing_desktop

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8975556244509881776

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6466792232321024


| 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!