Moving this nonessential bug to the next milestone.

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

Adding RB label as this is a recent regression.

Not observing this on Canary (53.0.2761.2) or Beta (52.0.2743.24) on retina Macbook Pro. Trying a non-retina machine.

Able to reproduce this issue on MacBook Air 10.11.5 using chrome version 53.0.2768.0.

Able to reproduce on Macbook Air locally.

Still able to reproduce the issue on 0.11.5 using chrome version 53.0.2774.2

ccameron@ Could you please look into this issue.

Thanks,

Still able to reproduce the issue on Mac 10.11.5 using 53.0.2780.0.

ccameron@ Any update on this issue.

Unable to repro this issue on MAC (10.11.5) for Google Chrome Canary Version - 53.0.2784.1

Screen-recording is attached.

@jshanbal: Could you please confirm the same.

Thank you.

Deleted: 618287.mov 3.8 MB

618287.mov 3.8 MB Download

With response to comment #8, above issue is reproducible on latest canary version 54.0.2787.0 on Mac 10.11.4.
Please refer the attached screencast.

Deleted: Actual_video.mov 3.9 MB

Actual_video.mov 3.9 MB Download

This seems specific to Macbook Air -- it might be specific to a particular GPU

Able to reproduce the issue on MacBook Air 10.11.5(non-touch & Non-retina) using latest canary 54.0.2795.0. Issue is seen only at the beginning and once its started played for  4-5 secs and then hovered the seek-bar and the media icons are looking fine.

@ccameron: Hey, would you mind providing an update on the above issue?

I really appreciate your help.

Thank you!

@ccameron: Gentle Ping.

Oh dear ... here I thought this was an Apple issue, and it's my own dumb fault.

Fix is easy, fortunately.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4eff049bf6baef15d02a9c2d019a38e4a50c9731

commit 4eff049bf6baef15d02a9c2d019a38e4a50c9731
Author: ccameron <ccameron@chromium.org>
Date: Mon Jul 25 20:22:20 2016

Mac: Correct CALayer reuse logic

This is a bug in how we construct the CALayer tree. The code
originally assumed that we would always re-use the old CALayer
if it existed. This is no longer the case now that we sometimes
use AVSampleBufferDisplayLayers.

With this patch, we ensure that we always put the new CALayer
in the position of the old CALayer.

A unit test for this is coming in another patch, but this one is
being merged back to M52, and the unit test files have changed
dramatically.

BUG= 618287 

Review-Url: https://codereview.chromium.org/2181713004
Cr-Commit-Position: refs/heads/master@{#407561}

[modify] https://crrev.com/4eff049bf6baef15d02a9c2d019a38e4a50c9731/ui/accelerated_widget_mac/ca_renderer_layer_tree.mm

Adding merge request for M52

And M53

Before we approve merge to M53, Could you please confirm whether this change is baked/verified in Canary or ToT and safe to merge?

The change is very simple and small. I've verified that it fixes the issue in local Chromium builds without any regressions.

I understand that the M52 bar is very high at this point, but I believe this issue should be merged at this point.

Approving merge to M53 branch 2785 based on comment #21. Please merge ASAP so we can take it for tomorrow's M53 last Dev release. Thank you.

Also approving merge to M52 branch 2743. Please make sure to test/verify the fix in tonight's canary and revert the change from M52/53 branches if things are NOT looking good in canary.

#23 SGTM

Thank you very much ccameron@. Please update the bug with Canary test result tomorrow morning. Truly appreciate you time and help.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e71bcfb49d00bd77ed4bd22aa059f6233f56a789

commit e71bcfb49d00bd77ed4bd22aa059f6233f56a789
Author: Christopher Cameron <ccameron@chromium.org>
Date: Mon Jul 25 21:39:59 2016

Mac: Correct CALayer reuse logic

This is a bug in how we construct the CALayer tree. The code
originally assumed that we would always re-use the old CALayer
if it existed. This is no longer the case now that we sometimes
use AVSampleBufferDisplayLayers.

With this patch, we ensure that we always put the new CALayer
in the position of the old CALayer.

A unit test for this is coming in another patch, but this one is
being merged back to M52, and the unit test files have changed
dramatically.

BUG= 618287 

Review-Url: https://codereview.chromium.org/2181713004
Cr-Commit-Position: refs/heads/master@{#407561}
(cherry picked from commit 4eff049bf6baef15d02a9c2d019a38e4a50c9731)

Review URL: https://codereview.chromium.org/2184463002 .

Cr-Commit-Position: refs/branch-heads/2785@{#345}
Cr-Branched-From: 68623971be0cfc492a2cb0427d7f478e7b214c24-refs/heads/master@{#403382}

[modify] https://crrev.com/e71bcfb49d00bd77ed4bd22aa059f6233f56a789/ui/accelerated_widget_mac/ca_renderer_layer_tree.mm

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/09bcd556fc055ba67b6ea65d708d7c1ac3091ae1

commit 09bcd556fc055ba67b6ea65d708d7c1ac3091ae1
Author: Christopher Cameron <ccameron@chromium.org>
Date: Mon Jul 25 21:47:28 2016

Mac: Correct CALayer reuse logic

This is a bug in how we construct the CALayer tree. The code
originally assumed that we would always re-use the old CALayer
if it existed. This is no longer the case now that we sometimes
use AVSampleBufferDisplayLayers.

With this patch, we ensure that we always put the new CALayer
in the position of the old CALayer.

A unit test for this is coming in another patch, but this one is
being merged back to M52, and the unit test files have changed
dramatically.

BUG= 618287 

Review-Url: https://codereview.chromium.org/2181713004
Cr-Commit-Position: refs/heads/master@{#407561}
(cherry picked from commit 4eff049bf6baef15d02a9c2d019a38e4a50c9731)

Review URL: https://codereview.chromium.org/2176113003 .

Cr-Commit-Position: refs/branch-heads/2743@{#699}
Cr-Branched-From: 2b3ae3b8090361f8af5a611712fc1a5ab2de53cb-refs/heads/master@{#394939}

[modify] https://crrev.com/09bcd556fc055ba67b6ea65d708d7c1ac3091ae1/ui/accelerated_widget_mac/ca_renderer_layer_tree.mm

Marking as fixed. I will verify with Canary tomorrow morning.

Thank you very much ccameron@.

Verified the issue on Latest Dev# 53.0.2785.21 and Latest Canary# 54.0.2808.0 on Mac OS X 10.11.5, the issue has been fixed and is working as intended.
Hence adding TE-Verified Labels.
Thank You.

Also to add, verified the issue on Dev# 53.0.2785.30 and Stable# 52.0.2743.91 and is also working as intended without any issue.
Also adding a screen cast for reference.
Thank You.

Deleted: 618287.mov 7.1 MB

618287.mov 7.1 MB Download

Also verified in Canary.

Verified the issue on Stable Version# 52.0.2743.91 on Mac OS X 10.11.5 and the issue has been fixed and is working as intended. Hence adding TE-Verified Labels.
Thank You everyone.

 Issue 627666  has been merged into this issue.