All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=816522

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=b2ba38665793753b417b3833bc54166033602ff0a010af54e9b016075ce228e8


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

chromium-rel-win7-dual

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/148a8f70440000

📍 Found significant differences after each of 2 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/148a8f70440000

[turbofan] Introduce StringSubstring operator by sigurds@chromium.org
https://chromium.googlesource.com/v8/v8/+/bcb5d452102d268a7976a9f4da73864fc0369ff8

Remove unused runtime functions by jgruber@chromium.org
https://chromium.googlesource.com/v8/v8/+/1ee80ebab0ca075d56061311b2056e5dc85c56eb

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

Might be related to change in computing indices as word32 instead of Smi. Only affected platform is i32.

Can you provide details on why this was WontFixed?

More investigation shows this is related to GVN hoisting a call to String.p.substring out of a switch-case. Fix in flight.

The following revision refers to this bug:
  https://chromium.googlesource.com/v8/v8.git/+/20baf1bfb7a9faf987bbdf857d6764ce33c58dcc

commit 20baf1bfb7a9faf987bbdf857d6764ce33c58dcc
Author: Sigurd Schneider <sigurds@chromium.org>
Date: Tue Feb 27 17:45:24 2018

[turbofan] Put StringSubstring on the effect chain

This change ensures that GVN does not move StringSubstring out of
switches, which might introduce partial redundancies.

Bug:  chromium:816522 
Change-Id: I63b91edd995c84b68d756ed5de08fa13567f3d80
Reviewed-on: https://chromium-review.googlesource.com/939621
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51610}
[modify] https://crrev.com/20baf1bfb7a9faf987bbdf857d6764ce33c58dcc/src/compiler/js-call-reducer.cc
[modify] https://crrev.com/20baf1bfb7a9faf987bbdf857d6764ce33c58dcc/src/compiler/simplified-lowering.cc
[modify] https://crrev.com/20baf1bfb7a9faf987bbdf857d6764ce33c58dcc/src/compiler/simplified-operator.cc

If I interpret the graph correctly, this had been fixed. Can you verify?

Yup, looks recovered to me.