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

Issue 747000 link

Starred by 2 users

Issue metadata

Status: WontFix
Owner:
User never visited
Closed: Jul 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 1
Type: ----



Sign in to add a comment

thread_times.polymer failing on Android Perf bots

Project Member Reported by zh...@chromium.org, Jul 20 2017

Issue description

Project Member

Comment 2 by 42576172...@developer.gserviceaccount.com, Jul 20 2017

Cc: hayato@chromium.org
Owner: hayato@chromium.org

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

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


=== BISECT JOB RESULTS ===
Test failure found with culprit

Suspected Commit
  Author : Hayato Ito
  Commit : b160073dd51148f56b6586a9902b8a24674886c4
  Date   : Wed Jul 19 04:33:31 2017
  Subject: Reland: Make /deep/ as no-op and remove ::shadow in dynamic profile

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : thread_times.polymer
  Metric       : benchmark_duration/benchmark_duration

Revision             Exit Code      N
chromium@487678      0 +- N/A       5      good
chromium@487728      0 +- N/A       5      good
chromium@487741      0 +- N/A       5      good
chromium@487744      0 +- N/A       5      good
chromium@487745      1 +- N/A       5      bad       <--
chromium@487746      1 +- N/A       5      bad
chromium@487747      1 +- N/A       5      bad
chromium@487753      1 +- N/A       5      bad
chromium@487777      1 +- N/A       5      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests thread_times.polymer

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8973534588980613392


For feedback, file a bug with component Speed>Bisection

Comment 3 by hayato@chromium.org, Jul 21 2017

Owner: ykyyip@chromium.org
Status: Assigned (was: Available)
It looks the benchmark uses PolymerPageSet:

https://cs.chromium.org/chromium/src/tools/perf/page_sets/polymer.py?q=PolymerPageSet&sq=package:chromium&l=222

I guess these pages are using /deep/ or ::shadow.

ykikyip@, could you have a chance to look?




Cc: nedngu...@google.com charliea@chromium.org
ykyyip doesn't work on polymer anymore. Ned, Charlie, should we just remove this benchamrk?
A benchmark is only as good as we have people to maintain it. I will remove this.
Project Member

Comment 6 by bugdroid1@chromium.org, Jul 21 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/435188ccd7fc5a58e35f78928270954d0aede221

commit 435188ccd7fc5a58e35f78928270954d0aede221
Author: nednguyen <nednguyen@google.com>
Date: Fri Jul 21 16:59:04 2017

Remove thread_times.polymer benchmark


This benchmark is broken, and benchmark owner no longer works on the Chromium
project.

Bug:  747000 
Change-Id: Ideddacfdf10b4628eb660eb37bee270d4f339020
Reviewed-on: https://chromium-review.googlesource.com/581417
Reviewed-by: Annie Sullivan <sullivan@chromium.org>
Commit-Queue: Annie Sullivan <sullivan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#488684}
[modify] https://crrev.com/435188ccd7fc5a58e35f78928270954d0aede221/testing/buildbot/chromium.perf.json
[modify] https://crrev.com/435188ccd7fc5a58e35f78928270954d0aede221/tools/perf/benchmark.csv
[modify] https://crrev.com/435188ccd7fc5a58e35f78928270954d0aede221/tools/perf/benchmarks/thread_times.py

Status: WontFix (was: Assigned)
Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, Jul 21 2017

Cc: benhenry@google.com
 Issue 747580  has been merged into this issue.
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Jul 22 2017

Cc: mikec...@chromium.org
 Issue 746729  has been merged into this issue.
Thanks for doing this Ned!

Sign in to add a comment