New issue
Advanced search Search tips

Issue 598432 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Apr 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 1
Type: Bug-Regression



Sign in to add a comment

[Regression] Touch scrolling stops scrolling when there is a related DOM mutation

Reported by dsmlo...@gmail.com, Mar 28 2016

Issue description

Example URL:
http://jsbin.com/liliveyolu

Steps to reproduce the problem:
1. Open http://jsbin.com/liliveyolu on an Android device with Chrome Beta (50.0.2661.49) or Chrome Dev (51.0.2690.0)
2. Put your finger on the text Item 6
3. Move your finger upward to start scrolling downward
4. Continue moving your finger upward 400px

What is the expected behavior?
The page should continue scrolling with your finger

What went wrong?
The page stops scrolling

Does it occur on multiple sites: Yes

Is it a problem with a plugin? No 

Did this work before? Yes 49.0.2623.105

Does this work in other browsers? Yes 

Chrome version: 50.0.2661.49  Channel: beta
OS Version: 6.0.1
Flash Version: 

I believe this is caused by a DOM mutation happening.

Here's a video of me reproducing it. You can see that it only is an issue when the touch is on the text itself, not on the background or the space between the text. Since this is using a "virtual list" library the other rows are being swapped in and out as it is scrolled. Near the end of the video I show the same in the latest Chrome stable (49.0.2623.105) which is working as expected. https://www.youtube.com/watch?v=ISlkwX-5O68

This has a major impact on our app so I really hope you can get this fixed before releasing this to Chrome stable.

Thanks!
 
Cc: bokan@chromium.org dtapu...@chromium.org tdres...@chromium.org
Components: -Blink Blink>Scroll
bokan@ is this something possibly caused by one of your recent changes?
Summary: [Regression] Touch scrolling stops scrolling when there is a related DOM mutation (was: Touch scrolling stops scrolling when there is a related DOM mutation)
Labels: -Type-Bug Type-Bug-Regression

Comment 4 by bokan@chromium.org, Mar 29 2016

Owner: bokan@chromium.org
Status: Assigned (was: Unconfirmed)
I don't think my changes have hit beta yet but I'll take a look.

Comment 5 by bokan@chromium.org, Apr 5 2016

Cc: -dtapu...@chromium.org
Labels: -Pri-2 Hotlist-Input-Dev Pri-1
Owner: dtapu...@chromium.org
Bisect points to: https://chromium.googlesource.com/chromium/src/+log/11ed0249bc070ad5541385aa8ee9a5704fe2c04f..0c7ea64a8b460a71fecc7cd0bc7d04d8dea7a843

Most likely this patch:

Move have_wheel_event_handlers to WebLayerTreeView.
Review URL: https://codereview.chromium.org/1639363002

Dave, does it make sense this patch could have caused this? Upping priority since this is in the current beta.
This is actually fixed with change https://codereview.chromium.org/1843813002

With wheel gestures combined that change did expose an  issue 599682  in the compositor. However M50 doesn't have wheel gestures enabled so if we merge this change into M50 I think it will be safe.

I will build M50 and try my patch on it before requesting a merge.
Status: Fixed (was: Assigned)
Merged change into branch M50. 


commit 8a65242bf4cadef0413ced54175c075532c2ef27
Author: Dave Tapuska <dtapuska@chromium.org>
Date:   Tue Apr 5 14:07:15 2016 -0400

    Fix issue with hit testing on the layer tree.
    
    When wheel event listeners were moved to the layer tree instead of
    on individual layers in https://codereview.chromium.org/1639363002
    
    We should have stopped hit testing on whether there was a wheel
    event listener or not. These properties are checked on the general
    page before processing the event.
    
    BUG= 597913 
    CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
    
    Review URL: https://codereview.chromium.org/1843813002
    
    Cr-Commit-Position: refs/heads/master@{#383872}
    (cherry picked from commit 913f48327c7c50fa79df5af74d68fc7af7d3e2f4)
    
    Review URL: https://codereview.chromium.org/1864633002 .
    
    Cr-Commit-Position: refs/branch-heads/2661@{#494}
    Cr-Branched-From: ef6f6ae5e4c96622286b563658d5cd62a6cf1197-refs/heads/master@{#378081}


Sign in to add a comment