I kicked off a bisect: https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3785

Good news: The bisect reproduced the issue :-)
Bad news: The bisect failed shortly afterwards :-(

I'll try to start another one...

https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3789

The second bisect failed for the exact same reason. I guess there's nothing we can do but wait for  issue 624734  to be fixed.

 Issue 624734  should be fixed now, so her's another one: https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3790

It seems that issue 604021 has been fixed (although it hasn't been marked as such), so here's another bisect: https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3817

It turns out that we're still blocked on issue 604021 :-(

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/9004253360962262560

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

Hi vmiura@chromium.org, the bisect results pointed to your CL below as possibly
causing a regression. Please have a look at this info and see whether
your CL be related.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Re-enable GPU Rasterization for content with any author defined viewport.
Author  : vmiura
Commit description:
  
BUG= 591179 

Review-Url: https://codereview.chromium.org/2097413003
Cr-Commit-Position: refs/heads/master@{#402702}
Commit  : a017b667a53b3ee5f8bc630be98c37ebf53a2339
Date    : Wed Jun 29 03:37:05 2016


===== TESTED REVISIONS =====
Revision         Mean      Std Dev  N  Good?
chromium@402670  23699456  0.0      5  good
chromium@402693  23699456  0.0      5  good
chromium@402699  23699456  0.0      5  good
chromium@402701  23699456  0.0      5  good
chromium@402702  25614064  0.0      5  bad    <--
chromium@402704  25614064  0.0      5  bad
chromium@402715  25614064  0.0      5  bad

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 624518

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests system_health.memory_mobile
Test Metric: load_social-memory:chrome:renderer_processes:reported_by_chrome:cc:effective_size_avg/load_social_pinterest
Relative Change: 8.08%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3991
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004253360962262560


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5885222453772288

| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Tests>AutoBisect.  Thank you!

This has since recovered to 20,761,500 bytes.