All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=827998

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=3f7d178206a703646d86af1bc2bc0ec930f534c39e4ffb282f1d4a3ac14209b0


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

android-one

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14dfe77f440000

📍 Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/14dfe77f440000

[Liftoff][mips] Implement instruction for type conversions by sreten.kovacevic@mips.com
https://chromium.googlesource.com/v8/v8/+/53decc1a03f5659fe64c9956f3ba6b471c971bdd

Remove legacy C++ implementation of Array#slice by danno@chromium.org
https://chromium.googlesource.com/v8/v8/+/5020630808043615f2f6fbbfba1ea73342bebd69

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

This CL simply removed dead code, there was not change in functionality, so it's unlikely that it was actually this CL. There were a bunch of other bigger functionality changes in this VL roll that seem more likely culprits. However, since the memory graphs have long-since recovered, I'm going to close as WontFix.