All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=881399

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=14c7d893ab95d919e5e9db4179206c9c354c8b3bcc276e5bb458054ec2c73d0b


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

Android Nexus5 Perf
android-nexus5x-perf

rendering.mobile - Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/145d0f15640000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/145d0f15640000

Fix the error catching logic of should().throw() in audit.js by hongchan@chromium.org
https://chromium.googlesource.com/chromium/src/+/02a673df348e86553b50467a7dde1afce427b713
217.2 → 137.6 (-79.59)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Benchmark documentation link:
  https://bit.ly/rendering-benchmarks

The commit is nothing to do with the canvas rendering. It only touches layout tests for WebAudio. Are we sure this about the cause of this regression?

WDYT rtoy@?

Agree, the change for audit.js shouldn't be affecting the canvas test.  Perhaps the skia roll in https://chromium.googlesource.com/chromium/src/+/53e63388501b107ea5c7a2f60518300d91358541 is relevant?  It's in the regression range.

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/119a6487640000

Something went wrong with the bisect, it was supposed to detect a regression, instead it says your CL improved performance. I agree that neither is likely when it only touches layout tests. Re-running the bisect.

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/119a6487640000

Looks like this benchmark is just noisy.