All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=700867

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtPScrwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1LOm7wsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtKLLqgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1Ku6owoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDglMHZogsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgtMT4rQsM


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

android-nexus5
android-nexus6
chromium-rel-mac-retina
chromium-rel-mac11
chromium-rel-win7-gpu-nvidia
chromium-rel-win8-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985236944346410416

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

Hi eae@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 : eae
  Commit : 8683b364d376b33bed4cbfe6b46d2c39411b05cd
  Date   : Sat Mar 11 03:28:13 2017
  Subject: Limit optimized LayoutBlockFlowLine width computation to 8bit

Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : page_cycler_v2.intl_ko_th_vi
  Metric       : timeToFirstMeaningfulPaint_avg/pcv1-warm/http___pantip.com_
  Change       : 17.83% | 162.215777771 -> 194.31311112

Revision             Result                  N
chromium@456261      162.216 +- 14.0389      9      good
chromium@456270      166.755 +- 59.0228      9      good
chromium@456271      195.171 +- 12.0392      9      bad       <--
chromium@456272      192.315 +- 18.2567      9      bad
chromium@456274      200.463 +- 21.477       9      bad
chromium@456278      194.313 +- 11.5887      9      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...pantip.com. page_cycler_v2.intl_ko_th_vi

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8985236944346410416

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4910215767523328


| 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 Speed>Bisection.  Thank you!

This, unfortunately, was an intentional change that trades speed for correctness. We had an optimization that allowed some inline layout to reuse cached metrics however this produced incorrect results for non-latin content.

We'll fix this once and for all with LayoutNG but until then being correct is more important than being fast.