New issue
Advanced search Search tips

Issue 717372 link

Starred by 1 user

Issue metadata

Status: Duplicate
Merged: issue 717078
Owner:
Closed: May 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 1
Type: Bug-Regression



Sign in to add a comment

51.7%-61.3% regression in v8.infinite_scroll-classic_tbmv2 at 467618:467688

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

Issue description

These infinite_scroll* benchmarks each increase by 10 minutes (12 min --> 22 mins).
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=717372

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgspDq0QgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgspSftgoM


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

chromium-rel-win7-dual
chromium-rel-win7-x64-dual
Labels: -Pri-2 Pri-1
Owner: u...@chromium.org
Status: Assigned (was: Untriaged)
ulan@: we are having a crisis in term of bot runtime (issue 713345). A 10 minutes cycle time increase without new tests is not great. Can you take a look at this?
Cc: martiniss@chromium.org
Components: -Speed>Telemetry Speed>Benchmarks

Comment 5 by u...@chromium.org, May 2 2017

> Can you take a look at this?
Sure. Nothing stands out in the regression range. I hope the bisect will pinpoint the culprit.
Mergedinto: 717078
Status: Duplicate (was: Assigned)

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

Suspected Commit
  Author : altimin
  Commit : c18764a33a595baab6ba35a1447e51c7c74b6e30
  Date   : Thu Apr 27 10:39:19 2017
  Subject: [scheduler] Move some task types to suspendable task runner.

Bisect Details
  Configuration: win_x64_perf_bisect
  Benchmark    : v8.infinite_scroll_tbmv2
  Metric       : benchmark_duration/benchmark_duration
  Change       : 47.24% | 12.6872833318 -> 18.6807222221

Revision             Result                   N
chromium@467617      12.6873 +- 0.38887       6      good
chromium@467627      12.5636 +- 0.154834      6      good
chromium@467632      12.5582 +- 0.135781      6      good
chromium@467633      18.2466 +- 3.41853       6      bad       <--
chromium@467634      17.953 +- 4.83109        6      bad
chromium@467635      18.2096 +- 3.3936        6      bad
chromium@467637      17.4595 +- 4.01568       6      bad
chromium@467657      18.6807 +- 4.51232       6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests v8.infinite_scroll_tbmv2

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8980743801698539008

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5556956300836864


| 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!

Sign in to add a comment