All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=773021

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=6293ae5903ee87b97ec67d677026f871f8646f14a8031baf279ab1eebdf2fbc8


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

android-nexus5
android-nexus5X
android-nexus6
android-nexus7v2

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966182464227185936

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

Hi khushalsagar@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 : Khushal
  Commit : 5c6f076df93c934c9955fe2ff20583a89a60b310
  Date   : Thu Oct 05 04:03:36 2017
  Subject: cc: Eliminate required for activation veto.

Bisect Details
  Configuration: android_nexus7_perf_bisect
  Benchmark    : thread_times.simple_mobile_sites
  Metric       : thread_other_cpu_time_per_frame/http___www.ebay.co.uk_
  Change       : 952.26% | 0.0970378383909 -> 1.0210918335

Revision             Result                      N
chromium@506627      0.0970378 +- 0.0156337      6      good
chromium@506637      0.102633 +- 0.0144905       6      good
chromium@506642      0.106708 +- 0.112012        6      good
chromium@506643      1.03611 +- 0.13559          6      bad       <--
chromium@506644      1.03522 +- 0.109331         6      bad
chromium@506645      1.05132 +- 0.145986         6      bad
chromium@506647      1.00301 +- 0.036653         6      bad
chromium@506666      1.04149 +- 0.145398         6      bad
chromium@506705      1.02109 +- 0.145533         6      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=http...www.ebay.co.uk. thread_times.simple_mobile_sites

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966182464227185936


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966073001853143696

The increase in thread_other_cpu_time_per_frame is expected. We moved some work from the raster thread to TaskScheduler background worker thread. The same range has an equivalent drop in thread_raster_cpu_timer_per_frame.

I'm a bit surprised to see the increase in thread_compositor_cpu_time_per_frame. Started a bisect for that to confirm its this change.

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

Suspected Commit
  Author : Khushal
  Commit : 5c6f076df93c934c9955fe2ff20583a89a60b310
  Date   : Thu Oct 05 04:03:36 2017
  Subject: cc: Eliminate required for activation veto.

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : thread_times.simple_mobile_sites
  Metric       : thread_renderer_compositor_cpu_time_per_frame/http___www.ebay.co.uk_
  Change       : 12.67% | 4.03599076554 -> 4.54749535801

Revision             Result                   N
chromium@506609      4.03599 +- 0.106507      6      good
chromium@506640      4.0709 +- 0.179936       6      good
chromium@506642      4.06426 +- 0.147005      6      good
chromium@506643      4.54253 +- 0.119645      6      bad       <--
chromium@506644      4.59566 +- 0.127863      6      bad
chromium@506648      4.59871 +- 0.189785      6      bad
chromium@506655      4.5553 +- 0.185948       6      bad
chromium@506670      4.5475 +- 0.108975       6      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=http...www.ebay.co.uk. thread_times.simple_mobile_sites

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966073001853143696


For feedback, file a bug with component Speed>Bisection

 Issue 773703  has been merged into this issue.

 Issue 773702  has been merged into this issue.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8965828125097517792

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8965828085459437840

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

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : smoothness.top_25_smooth
  Metric       : percentage_smooth/Twitter

Revision             Result                  N
chromium@506609      88.3561 +- 21.9163      21      good
chromium@506670      90.2408 +- 19.684       21      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=Twitter smoothness.top_25_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8965828125097517792


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8965823529017889088

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

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : smoothness.top_25_smooth
  Metric       : percentage_smooth/Twitter

Revision             Result                  N
chromium@506609      89.0903 +- 20.2105      21      good
chromium@506670      89.7683 +- 21.1491      21      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=Twitter smoothness.top_25_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8965823529017889088


For feedback, file a bug with component Speed>Bisection

Both regressions merged into this bug (skia:effective_size_avg and thread_compositor_cpu_time_per_frame) had actually shown an equivalent improvement when the patch that this change reverted (r498571) had landed. It was reverted since it was introducing excessive checkerboarding by forcing synchronous decoding of images on the active tree.

The memory regression seems just came back to the original ref value, so no real regression there. For the thread time increase on the compositor thread, now there are more cases where we async decode images and replacing them requires a pending tree which is work on the compositor thread. There are definitely cases where we can optimize and avoid using a pending tree for these invalidations. I've filed 774621 to track it.

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

Suspected Commit
  Author : Khushal
  Commit : 051a012c57a3fdfd6c933ec3df1c0e60d247a73a
  Date   : Fri Oct 06 04:42:46 2017
  Subject: perf: Enable field trial testing for image animations in cc.

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : loading.mobile
  Metric       : cpuTimeToFirstMeaningfulPaint_avg/Regular-3G/GSShop
  Change       : 32.28% | 9505.29066667 -> 6436.8365

Revision             Result                  N
chromium@506638      9505.29 +- 265.386      6      good
chromium@506832      9401.64 +- 586.964      6      good
chromium@506929      9673.93 +- 523.131      6      good
chromium@506953      9437.85 +- 924.001      6      good
chromium@506965      9598.52 +- 625.935      6      good
chromium@506968      9736.15 +- 656.55       6      good
chromium@506969      6467.93 +- 225.163      6      bad       <--
chromium@506970      6548.98 +- 261.719      6      bad
chromium@506971      6592.75 +- 483.148      6      bad
chromium@506977      6661.42 +- 274.536      6      bad
chromium@507025      6436.84 +- 336.912      6      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=GSShop loading.mobile

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8965828085459437840


For feedback, file a bug with component Speed>Bisection

This is not a duplicate of 772881. I had started the try job to see where the improvement in GSShop had come from.

 Issue 775950  has been merged into this issue.

 Issue 775950  has been merged into this issue.