All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=651109

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgpZrs9AoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg5ZT2oAkM


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

linux-release

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/9000260845211341296

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

Hi vmpstr@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 : cc: Use medium filter quality for downscales.
Author  : vmpstr
Commit description:
  
This patch makes us use a medium filter quality if one of the dimensions
of an image is being downscaled. There are a couple of things to watch
out for:

1. This may cause worse quality on some images
2. This may regress memory, since mips are larger cache objects than
   specific smallest size.

R=chrishtr@chromium.org, enne@chromium.org
BUG= 649046 
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_precise_blink_rel

Review-Url: https://codereview.chromium.org/2361243002
Cr-Commit-Position: refs/heads/master@{#421375}
Commit  : 7a49649a96774df16e8f1401137e557e6877a4a1
Date    : Tue Sep 27 23:05:26 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev     N  Good?
chromium@421362  2.33866  0.00406484  5  good
chromium@421372  2.33272  0.00415607  5  good
chromium@421374  2.33511  0.00961433  5  good
chromium@421375  2.42553  0.00426991  5  bad    <--
chromium@421377  2.42113  0.00658151  5  bad
chromium@421381  2.41763  0.0119415   5  bad
chromium@421399  2.41847  0.00596659  5  bad

Bisect job ran on: linux_perf_bisect
Bug ID: 651109

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests rasterize_and_record_micro.top_25_smooth
Test Metric: rasterize_time/rasterize_time
Relative Change: 3.41%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/linux_perf_bisect/builds/6731
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9000260845211341296


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5851610222166016

| 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!

vmpstr@, it looks like your cl increased rasterize time. Could you take a look?

Any updates?

It looks like this is caused by the fact that we now have larger images that we sample from. For example, if we had a 100x100 image with a 0.26 scale, then before we'd scale this once and sample from a 26x26 image. Now, we scale it once to 50x50 and sample from that.

I think this is an acceptable regression, given that the image quality improves as a result. +chrishtr, +enne for opinions.

Yeah, seems reasonable for an image quality tradeoff.

Agreed.