All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=792494

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=5fb876c2c1c9d0a8d4fe907410223d667cac902c272b5e8ead9dad8300bd30d4


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

android-webview-nexus6

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8960942712698963072

=== Auto-CCing suspected CL author herb@google.com ===

Hi herb@google.com, 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 : Herb Derby
  Commit : a08dcbd65621907b84f81f1e0aad2dd72534a08f
  Date   : Tue Dec 05 21:16:03 2017
  Subject: Use direct eval for small sigma. Try 2

Bisect Details
  Configuration: android_webview_nexus6_aosp_perf_bisect
  Benchmark    : rasterize_and_record_micro.top_25
  Metric       : rasterize_time/file___static_top_25_googleplus.html
  Change       : 14.15% | 43.6281666667 -> 49.8

Revision             Result                   N
chromium@521818      43.6282 +- 1.15846       6      good
chromium@521827      43.713 +- 0.742761       6      good
chromium@521830      43.8067 +- 0.355603      6      good
chromium@521831      43.9155 +- 0.686649      6      good
chromium@521832      50.624 +- 1.156          6      bad       <--
chromium@521836      50.4208 +- 0.545836      6      bad
chromium@521854      50.278 +- 0.389405       6      bad
chromium@521889      49.8 +- 0.693194         6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=file...static.top.25.googleplus.html rasterize_and_record_micro.top_25

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8960942712698963072


For feedback, file a bug with component Speed>Bisection

Issue 792956 has been merged into this issue.

herb: any update on the regressions here? +rasterize_and_record owners in case they have thoughts on if it's worth addressing a month after filing.

This works as intended. The quality of the blur < 2 sigma is much higher, and the performance is a wash when all machine types are compared.

But the main reason is that the previous blur was garbage, and the new one is actually correct.

https://chromeperf.appspot.com/report?sid=13b13515c3af9b99a343aebd098cd7ef55e271c758c448fee5756a4b0d0506d3&start_rev=520469&end_rev=526855