New issue
Advanced search Search tips

Issue 795783 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Dec 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

1.9% regression in thread_times.tough_scrolling_cases at 523380:523430

Project Member Reported by chiniforooshan@chromium.org, Dec 18 2017

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=795783

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=101eb11e70426749002600b28fd48f13a03fc4ab1b783ee892c635dfd9a95164


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

android-one

Comment 3 Deleted

Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

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

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

Hi mtklein@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 : Mike Klein
  Commit : 376fd31ad485c3df35d934c56364ff0c34eacdfa
  Date   : Tue Dec 12 12:02:48 2017
  Subject: remove vfpv4 requirement for SkJumper on ARMv7

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : thread_times.tough_scrolling_cases
  Metric       : thread_raster_cpu_time_per_frame/text_15000_pixels_per_second
  Change       : 2.66% | 6.93683346776 -> 7.10937144857

Revision                             Result                    N
chromium@523379                      6.93683 +- 0.146471       9       good
chromium@523405                      6.98364 +- 0.345684       9       good
chromium@523412                      6.92566 +- 0.0979257      6       good
chromium@523415                      6.95386 +- 0.200813       9       good
chromium@523416                      6.98823 +- 0.325332       14      good
chromium@523416,skia@376fd31ad4      7.05456 +- 0.209538       14      bad       <--
chromium@523417                      7.0904 +- 0.3735          9       bad
chromium@523418                      7.04959 +- 0.103626       9       bad
chromium@523430                      7.10937 +- 0.453429       9       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 --story-filter=text.15000.pixels.per.second thread_times.tough_scrolling_cases

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8959854286705080368


For feedback, file a bug with component Speed>Bisection
Status: WontFix (was: Assigned)
Good to know, but probably not going to do anything specific to address this.  That CL changed a lot of ARMv7 codegen slightly, and these slight perf changes are expected.  

It's this sort of refactoring that makes it easier to make the more interesting big sort performance improvements that you can see on the chart back on October 25.

Sign in to add a comment