New issue
Advanced search Search tips

Issue 675551 link

Starred by 1 user

Issue metadata

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



Sign in to add a comment

75% regression in v8.browsing_mobile at 439253:439327

Project Member Reported by ulan@google.com, Dec 19 2016

Issue description

See the link to graphs below.
 

Comment 1 by ulan@google.com, Dec 19 2016

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=675551

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgz87b6woM


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

android-nexus5
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Dec 19 2016

Cc: u...@chromium.org
Owner: u...@chromium.org

=== PERF REGRESSION ===


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

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


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : [heap] Reland "Use RAIL mode for initial heap sizing".
Author  : ulan
Commit description:
  
The original patch was reverted because of performance
regressions caused by removal of old heap sizing heuristics.

This patch keeps the old heuristics and adds RAIL mode.

BUG= chromium:613518 

Review-Url: https://codereview.chromium.org/2576543002
Cr-Commit-Position: refs/heads/master@{#41773}
Commit  : 7bdb906946b94e9ba10de13e104a39bbcabedbd6
Date    : Fri Dec 16 16:59:44 2016


===== TESTED REVISIONS =====
Revision                       Mean     Std Dev  N  Good?
chromium@439252                44.761   19.9349  6  good
chromium@439290                43.742   13.682   6  good
chromium@439309                41.2035  18.9502  6  good
chromium@439309,v8@3f5a60ac17  37.094   20.6695  6  good
chromium@439309,v8@b2b5d0ab5b  42.9535  16.0085  6  good
chromium@439309,v8@7bdb906946  59.1005  8.70031  6  bad    <--
chromium@439310                63.293   11.1377  6  bad
chromium@439311                60.9538  4.51209  6  bad
chromium@439312                61.3572  4.56638  6  bad
chromium@439314                58.5423  6.66376  6  bad
chromium@439318                60.9993  4.71169  6  bad
chromium@439327                61.32    2.99331  6  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 675551

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=browse.news.reddit v8.browsing_mobile
Test Metric: v8-gc-full-mark-compactor_max/browse_news/browse_news_reddit
Relative Change: 36.99%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/4470
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8992857646460666352


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5893371418116096

| 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 Tests>AutoBisect.  Thank you!

Comment 4 by u...@chromium.org, May 12 2017

Labels: Performance-Sheriff-V8

Comment 5 by u...@chromium.org, Jul 11 2017

Status: Fixed (was: Untriaged)

Sign in to add a comment