This is a result of the same known limitation in the quad-based hit testing that we do in the browser process -- the hit testing system assumes that when we receive a quad that covers an area, it can receive mouse/touch events.

Our existing plan was to wait until we can switch over to Viz hit testing ( issue 750755 ), but that work is still in progress. I don't know how prevalent it problem would be on the web, the example above was specifically constructed to illustrate the bug by flackr@.

Three options I can see:
1. Live with the problem until Viz hit testing can ship.
2. Figure out some kind of hack (this is not trivial, I know that somebody will suggest that we turn input event forwarding back on but I think that has unacceptable security problems)
3. Disable layer squashing

Hmm, I might amend option 3 to: 'Disable layer squashing when OOPIFs are enabled', excepting extension OOPIFs for which we haven't received any reports of problems so far.

Comment 1: Thanks for the background.  It seems like we do have some user-reported bugs on this, but maybe there's some flexibility in how quickly it's fixed.  A few questions below to figure out whether we need a short-term fix.

Comment 2: Isn't  issue 755515  about extension OOPIFs?  Seems like we want that case to work.

Do we when when Viz hit testing will be ready?  (fsamuel@?)

What are the benefits of layer squashing, and how bad would it be to disable it when a page has OOPIFs?

We plan to turn on Viz hit testing for M66.

Some form of viz hit-testing should be available for M66.

The viz hit-testing will be changed from its current implementation to use prop-trees instead, and so will require more work. Until then, the viz hit-testing should still be functionally correct, although may have some performance issues in a small number of edge cases.

Issue 793472 has been merged into this issue.

It's not really layer squashing per se that is the source of this class of bug.
I would describe it as "content in this layer does not necessarily handle events
on all pixels".

I agree, we have the same problem with overlays and pointer-events:none elements, but we have been breaking out the specific causes for the sake of considering whether we need to address any individually before Viz hit testing is ready to ship, which replaces the current flawed hit testing method. I am putting together a document summarizing all of this right now.

Passing to vollick@ who has a CL that fixes this problem in the short term, until a more comprehensive solution is landed.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/aff0a4488d81c9710ce45296ddaebae94c7c1556

commit aff0a4488d81c9710ce45296ddaebae94c7c1556
Author: Ian Vollick <vollick@chromium.org>
Date: Fri Dec 22 20:15:16 2017

Enable Layer Squashing Prevention

With this CL, a layer may preclude squashing by subsequent layers
in paint order. NB: it makes no attempt to handle cases where
layers change sort order due to relative position in a 3D rendering
context.

I plan to add the test separately. It's in progress here:
crrev.com/c/842946

Bug: 788807
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I817a7ddc6f446b7d65dde2c4b4d88a6ac8912549
Reviewed-on: https://chromium-review.googlesource.com/832668
Reviewed-by: Avi Drissman <avi@chromium.org>
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Ian Vollick <vollick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#526046}
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/chrome/browser/about_flags.cc
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/chrome/browser/flag_descriptions.cc
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/chrome/browser/flag_descriptions.h
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/content/child/runtime_features.cc
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/content/public/common/content_features.cc
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/content/public/common/content_features.h
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.cpp
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.h
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/Source/platform/exported/WebRuntimeFeatures.cpp
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.h
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/Source/platform/runtime_enabled_features.json5
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/third_party/WebKit/public/platform/WebRuntimeFeatures.h
[modify] https://crrev.com/aff0a4488d81c9710ce45296ddaebae94c7c1556/tools/metrics/histograms/enums.xml

Passing to gklassen@ who is, I believe, working on the more long-term solution.

Grabbing again to do the merge.

I've tested in canary on mac os (which was at 526108; this change is at 526046) and have confirmed that it works as expected and that the flag I've plumbed does indeed disable the feature. I think it's ok to merge at this point.

This bug requires manual review: M64 has already been promoted to the beta branch, so this requires manual review
Please contact the milestone owner if you have questions.
Owners: cmasso@(Android), cmasso@(iOS), kbleicher@(ChromeOS), abdulsyed@(Desktop)

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

Thanks for testing. Approving it for merge in M64. Branch:3282

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b2036273845fe40bc8327517e31bb2e2a3ad6070

commit b2036273845fe40bc8327517e31bb2e2a3ad6070
Author: Ian Vollick <vollick@chromium.org>
Date: Sun Dec 24 12:24:40 2017

Enable Layer Squashing Prevention

With this CL, a layer may preclude squashing by subsequent layers
in paint order. NB: it makes no attempt to handle cases where
layers change sort order due to relative position in a 3D rendering
context.

I plan to add the test separately. It's in progress here:
crrev.com/c/842946

Bug: 788807
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Change-Id: I817a7ddc6f446b7d65dde2c4b4d88a6ac8912549
Reviewed-on: https://chromium-review.googlesource.com/832668
Reviewed-by: Avi Drissman <avi@chromium.org>
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Ian Vollick <vollick@chromium.org>
Cr-Original-Original-Commit-Position: refs/heads/master@{#526046} (commit aff0a4488d81c9710ce45296ddaebae94c7c1556)
Reviewed-on: https://chromium-review.googlesource.com/843896
Reviewed-by: Ian Vollick <vollick@chromium.org>
Cr-Commit-Position: refs/branch-heads/3282@{#355}
Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/chrome/browser/about_flags.cc
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/chrome/browser/flag_descriptions.cc
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/chrome/browser/flag_descriptions.h
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/content/child/runtime_features.cc
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/content/public/common/content_features.cc
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/content/public/common/content_features.h
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.cpp
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.h
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/platform/exported/WebRuntimeFeatures.cpp
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.cpp
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.h
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/Source/platform/runtime_enabled_features.json5
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/third_party/WebKit/public/platform/WebRuntimeFeatures.h
[modify] https://crrev.com/b2036273845fe40bc8327517e31bb2e2a3ad6070/tools/metrics/histograms/enums.xml

Thanks vollick@!  I've confirmed the fix in 65.0.3305.0 as well.

Reassigning to gklassen@ for the long-term fix, per comment 13.  (Gary, feel free to mark this fixed and file a new bug for the long-term support, which might help keep the merge labels separate.)

Tested this issue on Windows,Mac 10.12.6 & Ubuntu 14.04 using chrome#64.0.3282.71 as per the steps mentioned in C#0.

Steps Followed:
--------------
(1) Launched Chrome with --site-per-process.
(2) Navigated to "http://output.jsbin.com/wabuwov" site
(3) Tried to click anywhere in the video within the rectangle defined by the blue boxes.

Observation:
-----------
Video is getting paused/resumed upon mouse click on video.
No action performed when we click on blue rectangular area.

Observed the same behavior on 65.0.3305.0 & 64.0.3282.71.

creis@, Please find the attached screencast for reference & confirm on the Fix on M64 as well.

Thanks in advance..!

Deleted: 788807.mp4 1.2 MB

	788807.mp4 1.2 MB View Download

The above fix is working as intended on latest M64 Beta#64.0.3282.71 for Win10, Mac OS X 10.12.6 & Linux Ubuntu 14.04.

jmukthavaram@, please try to replicate this issue on reported version of chrome (#64.0.3278.0), then it's easy to verify expected behavior.

Thank you!

M65 branch is on Thursday. Is there any reason to *not* revert this change before then? 'Slow-path' asynchronous hit testing has landed enough to be in a usable state, so I think this fix becomes redundant in 65.

I'm not aware of any reason. I think we should revert.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/23716e8311a3a114606437154cc68ebca8f76048

commit 23716e8311a3a114606437154cc68ebca8f76048
Author: Chris Harrelson <chrishtr@chromium.org>
Date: Tue Jan 16 23:23:23 2018

Revert "Enable Layer Squashing Prevention"

This reverts commit b2036273845fe40bc8327517e31bb2e2a3ad6070.

Reason for revert: this workaround is no longer needed, because
all of the correct input code is now implemented.

Original change's description:
> Enable Layer Squashing Prevention
> 
> With this CL, a layer may preclude squashing by subsequent layers
> in paint order. NB: it makes no attempt to handle cases where
> layers change sort order due to relative position in a 3D rendering
> context.
> 
> I plan to add the test separately. It's in progress here:
> crrev.com/c/842946
> 
> Bug: 788807
> Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
> Change-Id: I817a7ddc6f446b7d65dde2c4b4d88a6ac8912549
> Reviewed-on: https://chromium-review.googlesource.com/832668
> Reviewed-by: Avi Drissman <avi@chromium.org>
> Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
> Commit-Queue: Ian Vollick <vollick@chromium.org>
> Cr-Original-Original-Commit-Position: refs/heads/master@{#526046} (commit aff0a4488d81c9710ce45296ddaebae94c7c1556)
> Reviewed-on: https://chromium-review.googlesource.com/843896
> Reviewed-by: Ian Vollick <vollick@chromium.org>
> Cr-Commit-Position: refs/branch-heads/3282@{#355}
> Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}

TBR=avi@chromium.org,vollick@chromium.org,chrishtr@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug: 788807
Change-Id: I0939fe19ee3354a8dad057e5cd98edb1c790ec9b
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Reviewed-on: https://chromium-review.googlesource.com/869054
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/branch-heads/3282@{#516}
Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/chrome/browser/about_flags.cc
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/chrome/browser/flag_descriptions.cc
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/chrome/browser/flag_descriptions.h
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/content/child/runtime_features.cc
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/content/public/common/content_features.cc
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/content/public/common/content_features.h
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.cpp
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.h
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/platform/exported/WebRuntimeFeatures.cpp
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.cpp
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.h
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/Source/platform/runtime_enabled_features.json5
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/third_party/WebKit/public/platform/WebRuntimeFeatures.h
[modify] https://crrev.com/23716e8311a3a114606437154cc68ebca8f76048/tools/metrics/histograms/enums.xml

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/684779464c0077516831fce7b0d796ed8cbbb6e8

commit 684779464c0077516831fce7b0d796ed8cbbb6e8
Author: Chris Harrelson <chrishtr@chromium.org>
Date: Tue Jan 16 23:30:28 2018

Revert "Revert "Enable Layer Squashing Prevention""

This reverts commit 23716e8311a3a114606437154cc68ebca8f76048.

Reason for revert: Incorrectly reverted on the branch.

Original change's description:
> Revert "Enable Layer Squashing Prevention"
> 
> This reverts commit b2036273845fe40bc8327517e31bb2e2a3ad6070.
> 
> Reason for revert: this workaround is no longer needed, because
> all of the correct input code is now implemented.
> 
> Original change's description:
> > Enable Layer Squashing Prevention
> > 
> > With this CL, a layer may preclude squashing by subsequent layers
> > in paint order. NB: it makes no attempt to handle cases where
> > layers change sort order due to relative position in a 3D rendering
> > context.
> > 
> > I plan to add the test separately. It's in progress here:
> > crrev.com/c/842946
> > 
> > Bug: 788807
> > Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
> > Change-Id: I817a7ddc6f446b7d65dde2c4b4d88a6ac8912549
> > Reviewed-on: https://chromium-review.googlesource.com/832668
> > Reviewed-by: Avi Drissman <avi@chromium.org>
> > Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
> > Commit-Queue: Ian Vollick <vollick@chromium.org>
> > Cr-Original-Original-Commit-Position: refs/heads/master@{#526046} (commit aff0a4488d81c9710ce45296ddaebae94c7c1556)
> > Reviewed-on: https://chromium-review.googlesource.com/843896
> > Reviewed-by: Ian Vollick <vollick@chromium.org>
> > Cr-Commit-Position: refs/branch-heads/3282@{#355}
> > Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}
> 
> TBR=avi@chromium.org,vollick@chromium.org,chrishtr@chromium.org
> 
> # Not skipping CQ checks because original CL landed > 1 day ago.
> 
> Bug: 788807
> Change-Id: I0939fe19ee3354a8dad057e5cd98edb1c790ec9b
> Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
> Reviewed-on: https://chromium-review.googlesource.com/869054
> Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
> Cr-Commit-Position: refs/branch-heads/3282@{#516}
> Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}

TBR=avi@chromium.org,vollick@chromium.org,chrishtr@chromium.org

Change-Id: Ieda24a5f799dc4c254aecbdfbc9202fa18b3207d
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 788807
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Reviewed-on: https://chromium-review.googlesource.com/868647
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/branch-heads/3282@{#517}
Cr-Branched-From: 5fdc0fab22ce7efd32532ee989b223fa12f8171e-refs/heads/master@{#520840}
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/chrome/browser/about_flags.cc
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/chrome/browser/flag_descriptions.cc
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/chrome/browser/flag_descriptions.h
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/content/child/runtime_features.cc
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/content/public/common/content_features.cc
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/content/public/common/content_features.h
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.cpp
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.h
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/platform/exported/WebRuntimeFeatures.cpp
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.cpp
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.h
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/Source/platform/runtime_enabled_features.json5
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/third_party/WebKit/public/platform/WebRuntimeFeatures.h
[modify] https://crrev.com/684779464c0077516831fce7b0d796ed8cbbb6e8/tools/metrics/histograms/enums.xml

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a0d5a910c303e9243d30ecaff5227b9f6b65d785

commit a0d5a910c303e9243d30ecaff5227b9f6b65d785
Author: Chris Harrelson <chrishtr@chromium.org>
Date: Wed Jan 17 20:38:14 2018

Revert "Enable Layer Squashing Prevention"

This reverts commit aff0a4488d81c9710ce45296ddaebae94c7c1556.

Reason for revert: this workaround is no longer needed, as the
correct input processing paths have been implemented.

Original change's description:
> Enable Layer Squashing Prevention
> 
> With this CL, a layer may preclude squashing by subsequent layers
> in paint order. NB: it makes no attempt to handle cases where
> layers change sort order due to relative position in a 3D rendering
> context.
> 
> I plan to add the test separately. It's in progress here:
> crrev.com/c/842946
> 
> Bug: 788807
> Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
> Change-Id: I817a7ddc6f446b7d65dde2c4b4d88a6ac8912549
> Reviewed-on: https://chromium-review.googlesource.com/832668
> Reviewed-by: Avi Drissman <avi@chromium.org>
> Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
> Commit-Queue: Ian Vollick <vollick@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#526046}

TBR=avi@chromium.org,vollick@chromium.org,chrishtr@chromium.org

# Not skipping CQ checks because original CL landed > 1 day ago.

Bug: 788807
Change-Id: If31c2b1a702d85c0dfa737e91c792e6db7be6174
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2
Reviewed-on: https://chromium-review.googlesource.com/868648
Reviewed-by: Chris Harrelson <chrishtr@chromium.org>
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#529867}
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/chrome/browser/about_flags.cc
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/chrome/browser/flag_descriptions.cc
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/chrome/browser/flag_descriptions.h
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/content/child/runtime_features.cc
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/content/public/common/content_features.cc
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/content/public/common/content_features.h
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.cpp
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/Source/core/paint/compositing/CompositingLayerAssigner.h
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/Source/platform/exported/WebRuntimeFeatures.cpp
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/Source/platform/graphics/SquashingDisallowedReasons.h
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/Source/platform/runtime_enabled_features.json5
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/third_party/WebKit/public/platform/WebRuntimeFeatures.h
[modify] https://crrev.com/a0d5a910c303e9243d30ecaff5227b9f6b65d785/tools/metrics/histograms/enums.xml

Code is now reverted.

The remaining task on this bug is to make sure that the use cases in question have
good tests in place to prevent regression. Gary, could you triage to someone who can
find out?

I'm seeing the symptom in issue 793472 but only after changing tabs back and forth. Maybe this is a KI but I thought the original bug happened all the time and not just after tabbing back/forth. Win10 / 66.0.3359.26

Re comment 28: Please file a new bug with steps to reproduce the issue, and cc me.

Comment 28: From the discussion on  issue 823405 , it sounds like the changing tabs back/forth problem has been fixed in  issue 822442 .