All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=771767

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=17f8ed7acd62b94038a1aabf48f29bebdfca7bc3b3bc039b8f2a8ab9154f0fab


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

chromium-rel-mac-retina
chromium-rel-mac11-air

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966629874745175920

I blame myself:

commit	9e59e6960ddaffb6b83ab23153c5434dd60fc286	
author	Yuri Wiitala <miu@chromium.org>committer
GLHelper: Split non-readback part of ReadbackYUV into I420Converter.

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

Hi miu@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 : Yuri Wiitala
  Commit : 9e59e6960ddaffb6b83ab23153c5434dd60fc286
  Date   : Mon Oct 02 22:54:31 2017
  Subject: GLHelper: Split non-readback part of ReadbackYUV into I420Converter.

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : performance_browser_tests
  Metric       : CastV2Performance_gpu_novsync_24fps/capture_duration
  Change       : 10.25% | 6.94087916667 -> 7.65209583333

Revision             Result                    N
chromium@505725      6.94088 +- 0.109995       6      good
chromium@505804      7.06703 +- 0.181257       6      good
chromium@505824      7.0443 +- 0.160178        6      good
chromium@505829      7.02132 +- 0.176099       6      good
chromium@505831      7.27218 +- 0.278104       6      good
chromium@505832      7.59383 +- 0.0299131      6      bad       <--
chromium@505834      7.58073 +- 0.0925564      6      bad
chromium@505843      7.69671 +- 0.0787715      6      bad
chromium@505882      7.6521 +- 0.103177        6      bad

To Run This Test
  ./src/out/Release/performance_browser_tests --test-launcher-print-test-stdio=always --enable-gpu

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966629874745175920


For feedback, file a bug with component Speed>Bisection

As mentioned in  bug 768335 , a lot of things are moving around right now; and when that's done, the performance issues may self-resolve. So, marking this bug as blocked on 754872.

There's still work actively happening on the blocked bug, but at this point it doesn't seem helpful for perf sheriffs to ping, and unfortunately monorail doesn't have a way to exclude from our list while the blocked bug is open.

I'm removing the Performance-Sheriff label so we don't have this in our queue, but please re-add if you still want pings.

No longer using I420Converter for the new tab capture pipeline. That code will be deleted soon.