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

Suspected Commit
  Author : Christopher Cameron
  Commit : 1894d423068be735560e0171922c833c4091b5d1
  Date   : Wed Jun 28 06:31:03 2017
  Subject: color: Enable color correct rendering by default

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : battor.trivial_pages
  Metric       : idle:power_avg/TrivialBlurAnimationPageSharedPageState
  Change       : 4.20% | 25.932229197 -> 27.0203993974

Revision             Result                   N
chromium@482822      25.9322 +- 0.685159      4       good
chromium@482865      25.9692 +- 0.563613      5       good
chromium@482886      26.0496 +- 0.621093      9       good
chromium@482897      26.2129 +- 1.37027       12      good
chromium@482902      26.3056 +- 1.20399       9       good
chromium@482903      26.2686 +- 0.636961      5       good
chromium@482904      27.2953 +- 0.202219      6       bad       <--
chromium@482905      27.2513 +- 0.624796      6       bad
chromium@482907      27.0204 +- 0.44994       9       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=TrivialBlurAnimationPageSharedPageState battor.trivial_pages

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8974314902013701760


For feedback, file a bug with component Speed>Bisection

Un-merge!

Perf sheriff checking in: Any update on this?

For the ones that are related to video, we're doing more work to do the right thing, and this is costing more.

For some of the others I'm still taking a look, but it appears that we need to set the exact same ICC profile as we were given from the display.

Another perf sheriff ping! Feel free to remove the Performance-Sheriff label so you don't get these pings.

Marking as fixed, although some of these are WontFix. The difference is:

Things that draw rasterized content have returned to the same power level (we indeed need to match the ICC profile byte-for-byte).

Things that draw canvas, webgl, or video are now using more power (but that's required for correctness).