All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=781652

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=5eab54d60cb47f7ee0a2f480fe44b637ac109ec95baf3ea7b8f628deb61efde3


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

chromium-rel-win7-x64-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8963712957607546224

=== BISECT JOB RESULTS ===
Perf regression found but unable to continue

Bisect was stopped because a commit couldn't be classified as either
good or bad.


Bisect Details
  Configuration: win_x64_perf_bisect
  Benchmark    : media.desktop
  Metric       : cpu_time_percentage_avg/video.html?src_tulip2.ogg_type_audio

Revision             Result                       N
chromium@512425      0.027895 +- 0.00976983       14      good
chromium@512430      0.0275127 +- 0.00480307      14      good
chromium@512433      0.0279187 +- 0.0101091       21      good
chromium@512434      0.0286997 +- 0.0121297       21      unknown
chromium@512435      0.0305397 +- 0.00584637      21      bad
chromium@512445      0.0311743 +- 0.00412631      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=video.html.src.tulip2.ogg.type.audio media.desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8963712957607546224


For feedback, file a bug with component Speed>Bisection

Bisect narrowed down the regression to one of these two CLs as root cause. 
http://test-results.appspot.com/revision_range?start=512434&end=512435

My guess is #512435: "Call UpdateGPUInfo and UpdateGpuFeatureInfo always on GPU process launch."

zmo@, Could your change have caused ~9.4% regression in cpu usage during a video playback scenario? You can test this locally using the following command:

src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=video.html.src.tulip2.ogg.type.audio media.desktop

(Btw, if you look at https://chromeperf.appspot.com/group_report?bug_id=781652 , it seems like it's just returning to an earlier state)

My CL was actually reverted in https://chromium-review.googlesource.com/743741

Does the regression recover after the revert? If not, there must be another culprit.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8963468324105066848

Thanks zmo@. I'm starting another bisect over a wider range.

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

Suspected Commit
  Author : Zhenyao Mo
  Commit : 39d8a7f229e1330e72724895eb4730b7108e47a0
  Date   : Mon Oct 30 13:45:42 2017
  Subject: Revert "One step further into moving GPU feature decisions to GPU process"

Bisect Details
  Configuration: win_x64_perf_bisect
  Benchmark    : media.desktop
  Metric       : cpu_time_percentage_avg/video.html?src_tulip2.ogg_type_audio
  Change       : 25.76% | 0.0273144378468 -> 0.0343512258289

Revision             Result                       N
chromium@512418      0.0273144 +- 0.00990034      9      good
chromium@512467      0.0316791 +- 0.00435623      9      good
chromium@512479      0.0311157 +- 0.00430066      9      good
chromium@512482      0.0320574 +- 0.00156512      9      good
chromium@512484      0.0309527 +- 0.00538927      9      good
chromium@512485      0.0348306 +- 0.00506777      9      bad       <--
chromium@512491      0.0355683 +- 0.00728682      9      bad
chromium@512515      0.0365359 +- 0.00279007      6      bad
chromium@512611      0.0338791 +- 0.00660567      6      bad
chromium@512803      0.0343512 +- 0.00403024      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 --story-filter=video.html.src.tulip2.ogg.type.audio media.desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8963468324105066848


For feedback, file a bug with component Speed>Bisection