New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 717402 link

Starred by 4 users

Issue metadata

Status: Fixed
Owner:
Closed: May 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

14.3%-16.9% regression in v8.runtimestats.browsing_desktop_classic at 467899:468029 (JSArray::length handling in megamorphic KeyedLoadIC handler?)

Project Member Reported by bmeurer@google.com, May 2 2017

Issue description

See the link to graphs below.
 

Comment 1 by bmeurer@google.com, May 2 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=717402

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsr3fqAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg8pTcvgkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsqKtvAkM


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

chromium-rel-mac11-air
chromium-rel-mac12
chromium-rel-mac12-mini-8gb
Cc: delph...@chromium.org
Owner: delph...@chromium.org

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

Hi delphick@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 : delphick
  Commit : c41d3c785c5179a38e54c6c80ea63d56cc7cfd12
  Date   : Fri Apr 28 11:50:09 2017
  Subject: Restore one millisecond minimum timeout for setTimeout.

Bisect Details
  Configuration: mac_10_12_perf_bisect
  Benchmark    : v8.runtimestats.browsing_desktop
  Metric       : IC:duration_avg/browse_news/browse_news_flipboard
  Change       : 18.94% | 236.9385 -> 281.812833333

Revision             Result                  N
chromium@467898      236.938 +- 18.0803      6      good
chromium@467930      241.661 +- 9.23904      6      good
chromium@467946      240.74 +- 20.0826       6      good
chromium@467954      239.146 +- 6.05591      6      good
chromium@467958      246.476 +- 28.3403      6      good
chromium@467960      251.254 +- 24.746       6      good
chromium@467961      245.962 +- 16.508       6      good
chromium@467962      281.813 +- 17.2319      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.news.flipboard v8.runtimestats.browsing_desktop

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980733978633102464

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5043307605917696


| 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 717041 has been merged into this issue.
Cc: nzolghadr@google.com
 Issue 717656  has been merged into this issue.
Cc: kraynov@google.com
 Issue 718750  has been merged into this issue.
 Issue 718750  has been merged into this issue.
Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, May 17 2017

Cc: land...@opera.com
 Issue 717227  has been merged into this issue.
Status: Fixed (was: Assigned)
This was a rollback of a change that had both positive and negative outcomes so these regressions are expected.

Sign in to add a comment