All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=613894

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgzLrapwoM


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

chromium-rel-win7-single

Moving this nonessential bug to the next milestone.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Trying another bisect.

This issue has been moved once and is lower than Pri-1. Removing the milestone.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

The issue still persist and suspecting CL: https://chromium.googlesource.com/chromium/src/+/2ea2924dc159391166a1b42008f576a501e42ea3

The other CL touching core/layout only affects table layout, which isn't relevant in the test: https://cs.chromium.org/chromium/src/third_party/WebKit/PerformanceTests/Layout/multicol/tall-content-short-columns-realistic.html?sq=package:chromium&dr

That CL makes sense to me. Chris, can you take a look?

I have a WIP that may affect this. Also still need to verify whether I can
reproduce the slowdown.

Reverting my patch did not make the test 25% faster. I don't think it was
my patch.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d74f43ae4a3e5cc9392c15728be21e3f2d14a1be

commit d74f43ae4a3e5cc9392c15728be21e3f2d14a1be
Author: chrishtr <chrishtr@chromium.org>
Date: Thu Jul 28 18:03:27 2016

Fix order of operations in LayoutView::mapAncestorToLocal

mapAncestorToLocal should first recurse, then apply each of the offsets/transforms
in the forward direction but inverse order.

The forward direction will be transformed to backwards direction by the TransformState state bit indicating "unapply inverse direction".
This was incorrectly fixed in https://codereview.chromium.org/1994203003.

Also, remove code to handle transforms on the LayoutView. It's impossible
to have such a transform from content. This code was added in:

https://chromium.googlesource.com/chromium/src/+/ce75f4f9824faae4464420cf9c96a6cf0fb85c77

without comment as to why.

BUG= 613894 

Review-Url: https://codereview.chromium.org/2141693002
Cr-Commit-Position: refs/heads/master@{#408434}

[modify] https://crrev.com/d74f43ae4a3e5cc9392c15728be21e3f2d14a1be/third_party/WebKit/Source/core/layout/LayoutView.cpp
[modify] https://crrev.com/d74f43ae4a3e5cc9392c15728be21e3f2d14a1be/third_party/WebKit/Source/core/layout/MapCoordinatesTest.cpp

I can't reproduce this on Linux or on Windows, despite trying hard.

Also, the charts haven't been updated since July 13, so I can't see if d74f43ae4a3e5cc9392c15728be21e3f2d14a1be helped. Finally, it's only reported
for obscure multicolumn tests.

So I'm going to just close this bug.