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

Issue 762253 link

Starred by 1 user
Status: WontFix
Owner: ----
Closed: Sep 2017
Cc:
█ briander...@chromium.org
michae...@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

5.4%-7.7% regression in smoothness.tough_filters_cases at 499355:499384

Project Member Reported by briander...@chromium.org, Sep 5 2017 
See the link to graphs below.
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Sep 5 2017 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=762253

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=a331ba72d7f4cc1c574d29bfd89384a3887212e122132b54f07aedbe1802273b


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

chromium-rel-win10
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Sep 6 2017

Cc: michae...@chromium.org
Owner: michae...@chromium.org
Status: Assigned (was: Untriaged)

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

Hi michaelpg@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 : Michael Giuffrida
  Commit : 2dbce0d12e269610dad1b6846f2b5278a1d77f59
  Date   : Sat Sep 02 03:30:59 2017
  Subject: Componentize KeepAliveRegistry

Bisect Details
  Configuration: winx64_10_perf_bisect
  Benchmark    : smoothness.tough_filters_cases
  Metric       : frame_times/MotionMark Focus
  Change       : 1.36% | 59.1338605486 -> 60.0453020758

Revision             Result                  N
chromium@499354      59.1339 +- 2.39661      14      good
chromium@499370      59.2842 +- 2.35257      21      good
chromium@499377      58.9161 +- 2.53414      9       good
chromium@499381      59.3201 +- 3.24832      14      good
chromium@499383      58.989 +- 1.42558       6       good
chromium@499384      60.0453 +- 3.27361      14      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 --story-filter=MotionMark.Focus smoothness.tough_filters_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969253229454024928


For feedback, file a bug with component Speed>Bisection

Comment 4 by michae...@chromium.org, Sep 6 2017

Owner: ----
Status: Unconfirmed (was: Assigned)
I'm confused, what is this?

According to this bisect, the metric went from
  59.1339 +- 2.39661
to
  60.0453 +- 3.27361

so the difference is less than the margin of error. Meaning my CL could actually have improved whatever metric this is...

Anywho, my CL had nothing to do with drawing/painting stuff. I moved some files. The only non-mechanical change is compiled out of Release builds (ie DCHECK-off builds).

Per http://g.co/ChromePerformanceRegressions I'll start another bisect.

Comment 6 by michae...@chromium.org, Sep 6 2017 

This looks more relevant, though the CL description says the functionality it removed was unused: https://chromium-review.googlesource.com/648096

also a v8 roll, https://chromium-review.googlesource.com/647912, and some CSS stuff, https://chromium-review.googlesource.com/611802
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Sep 6 2017 


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: winx64_10_perf_bisect
  Benchmark    : smoothness.tough_filters_cases
  Metric       : frame_times/IE PirateMark

Revision             Result                   N
chromium@499354      17.864 +- 1.26676        21      good
chromium@499384      17.7531 +- 0.539878      21      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 --story-filter=IE.PirateMark smoothness.tough_filters_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969235614694371632


For feedback, file a bug with component Speed>Bisection

Comment 9 by briander...@chromium.org, Sep 6 2017

Status: WontFix (was: Unconfirmed)
Closing. The ref bots show a regression at the same time.
Project Member

Comment 10 by 42576172...@developer.gserviceaccount.com, Sep 6 2017 


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: winx64_10_perf_bisect
  Benchmark    : smoothness.tough_filters_cases
  Metric       : frame_times/frame_times

Revision             Result                   N
chromium@499354      23.0036 +- 0.605748      21      good
chromium@499384      23.0647 +- 0.695115      21      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 smoothness.tough_filters_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969178067698219904


For feedback, file a bug with component Speed>Bisection
►

Sign in to add a comment