All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=629502

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgqr-roAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgqu-eqgoM


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

android-nexus5
android-nexus9

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/9006701054768348304

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

Hi picksi@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 : Reland of -enable GPU Rasterization for content with any author defined viewport. (patchset #1 id:1 of https://codereview.chromium.org/2156553002/ )
Author  : picksi
Commit description:
  
Reason for revert:
Reverting change now that we have collected data from telemetry.

Original issue's description:
> Revert of Re-enable GPU Rasterization for content with any author defined viewport. (patchset #2 id:20001 of https://codereview.chromium.org/2097413003/ )
>
> Reason for revert:
> This has caused an unexpectedly large regression in overall PSS (about 5MB, graph here: https://chromeperf.appspot.com/report?sid=0b072725c25637efb0d3a44383da1e24e42bbb3740a00505e4040af2173423b0&start_rev=402061&end_rev=402790).
> The owners are both OOO and this is blocking Android release. This has already been reverted in the release branch but our infrastructure cannot gather data from the branch to confirm that the revert has had the intended result.
>
> This revert will allow us to confirm that this CL was the cause of the regression. Once confirmed (or otherwise) via telemetry dashboards this revert will be re-reverted.
>
> Original issue's description:
> > Re-enable GPU Rasterization for content with any author defined viewport.
> >
> > BUG= 591179 
> >
> > Committed: https://crrev.com/a017b667a53b3ee5f8bc630be98c37ebf53a2339
> > Cr-Commit-Position: refs/heads/master@{#402702}
>
> TBR=aelias@chromium.org,chrishtr@chromium.org,vmiura@chromium.org
> # Not skipping CQ checks because original CL landed more than 1 days ago.
> BUG= 591179 
>
> Committed: https://crrev.com/10a77702aa0490dfaaa6e354151b24b738474fb4
> Cr-Commit-Position: refs/heads/master@{#405750}

TBR=aelias@chromium.org,chrishtr@chromium.org,vmiura@chromium.org,primiano@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG= 591179 

Review-Url: https://codereview.chromium.org/2154193002
Cr-Commit-Position: refs/heads/master@{#405974}
Commit  : 5f19720ded2369857bee408ab48e11e7b5d28f7b
Date    : Mon Jul 18 11:06:45 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@405970  10338.5  50.0029  8  good
chromium@405972  10417.5  158.391  8  good
chromium@405973  10424.5  121.917  8  good
chromium@405974  10729.0  64.2762  8  bad    <--

Bisect job ran on: android_nexus5_perf_bisect
Bug ID: 629502

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests webrtc.datachannel
Test Metric: vm_private_dirty_final_renderer/vm_private_dirty_final_renderer
Relative Change: 3.67%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5_perf_bisect/builds/3843
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9006701054768348304


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5777200595337216

| 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!

Working as expected. This is a re-land of a feature that eats RAM but gives performance benefits. It was reverted & relanded to determine how much RAM it is actually using.

Thanks. So this is a duplicate of  issue 624630 . As I pointed out the regressions are not limited to RAM but also impact input latency as well.