All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=768550

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=8a4bac740aec1814070a7dc3c38473b2dbc2c22ffa574c15796fed8fe5dfad49


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

chromium-rel-win7-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8967445130358314624

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

Hi fgorski@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 : Filip Gorski
  Commit : 0da94d96c133a17aefee61ae35c826a69d1f1ddc
  Date   : Fri Sep 22 23:31:10 2017
  Subject: [Offline pages] Implement additional get pages functionality

Bisect Details
  Configuration: win_perf_bisect
  Benchmark    : thread_times.tough_scrolling_cases
  Metric       : thread_total_all_cpu_time_per_frame/text_constant_full_page_raster_10000_pixels_per_second
  Change       : 2.80% | 33.1523580572 -> 32.2233875168

Revision             Result                   N
chromium@503869      33.1524 +- 2.60551       14      good
chromium@503886      33.0756 +- 2.62885       14      good
chromium@503894      33.3788 +- 1.95933       6       good
chromium@503898      33.6112 +- 0.882994      6       good
chromium@503900      33.1699 +- 1.16228       6       good
chromium@503901      33.579 +- 0.313899       6       good
chromium@503902      32.2449 +- 0.701728      9       bad       <--
chromium@503935      32.2234 +- 2.3345        14      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=text.constant.full.page.raster.10000.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/8967445130358314624


For feedback, file a bug with component Speed>Bisection

The code in question is not compiled for Windows to my knowledge. It is Android specific.

Based on:
https://cs.chromium.org/chromium/src/chrome/common/BUILD.gn?q=offline_pages/features&l=240&dr=C
And:
https://cs.chromium.org/chromium/src/components/offline_pages/features/features.gni

Is there a chance that the bisect is incorrect.

It's possible. Kicking again.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8967355889979184688

fgorski - just to be sure, though, please prove that the files changed in your CL did not affect the Windows build, which you can do by performing two builds, one before and one after and comparing the results. According to dpranke@: "there is not a 1:1 mapping between 'this feature is enabled' and 'these files are compiled', life is way more complicated."

more specifically, even though the feature flag is not enabled, it looks to me that those files are still being compiled and linked into chrome on windows. I didn't look at the details of the CL or the code beyond the list of filenames, and so I couldn't say whether the change itself affected the chrome binary, but it at least looks like it might've.

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

Bisect Details
  Configuration: win_perf_bisect
  Benchmark    : thread_times.tough_scrolling_cases
  Metric       : thread_total_all_cpu_time_per_frame/text_constant_full_page_raster_10000_pixels_per_second

Revision             Result                  N
chromium@503869      57.3996 +- 5.15354      21      good
chromium@503935      57.4851 +- 7.8834       21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=text.constant.full.page.raster.10000.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/8967355889979184688


For feedback, file a bug with component Speed>Bisection

Can we reasonably presume there is nothing to fix here?