New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 640639 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
Closed: Oct 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux , Windows , Mac
Pri: 1
Type: Bug-Regression



Sign in to add a comment

absolute position not recalculated on resize inside fixed container

Reported by diamondo...@gmail.com, Aug 24 2016

Issue description

Chrome Version       : 52.0.2743.116 (64-bit)
URLs (if applicable) : http://jsbin.com/pogiwitena/edit?html,css,output
Other browsers tested:
     Safari: OK 9.1.2
    Firefox: OK 48.0.1

What steps will reproduce the problem?
(1) Load http://jsbin.com/pogiwitena/edit?html,css,output
(2) Resize the browser window so the height changes
(3) It only works if you grab the bottom edge.  Grabbing the corner so the width changes as well triggers a recalc.

What is the expected result?
light blue footer should move to stay at the bottom of the page

What happens instead?
The footer stays "stuck" wherever it currently is.

Please provide any additional information below. Attach a screenshot if
possible.

 
Screen Shot 2016-08-24 at 11.06.46 AM.png
15.1 KB View Download
Screen Shot 2016-08-24 at 11.06.53 AM.png
16.9 KB View Download
Components: Blink>CSS
Components: -Blink>CSS Blink>Layout>Flexbox
Labels: -Pri-3 Pri-2
Status: Available (was: Unconfirmed)
Able to repro in Chrome 53. The footer randomly updates its vertical position as the viewport is resized.
In Chrome 54 it's better but not perfect. The footer updates its vertical position when the width changes but not when the height changes.
Labels: Needs-Bisect
Cc: cbiesin...@chromium.org
Components: -Blink>Layout>Flexbox Blink>Layout
This doesn't actually need any flexbox -- the attached testcase has the issue, too (showing up slightly differently -- when resizing smaller, the footer stays outside of the viewport)
x.html
506 bytes View Download
Cc: krajshree@chromium.org
Labels: -Type-Bug -Pri-2 -Needs-Bisect M-55 has-Bisect OS-Linux OS-Mac OS-Windows Pri-1 Type-Bug-Regression
Owner: yukishiino@chromium.org
Status: Assigned (was: Available)
Able to reproduce on Windows 10, Ubuntu 14.04 and Mac OS 10.11.6 using chrome stable #53.0.2785.116 and the reported chrome version #52.0.2743.116.

Bisect Information:
=====================
Good build: 45.0.2430.0
Bad Build : 45.0.2433.0

Change Log URL: https://chromium.googlesource.com/chromium/src/+log/76af67653a03212a3d9104e3a94a44d773917214..7ec6a579b7e47f84b384b98b6f04ee65beb55758

From the above change log suspecting below change
Review URL: 
https://codereview.chromium.org/1174343003

yukishiino@ - Could you please check whether this is caused with respect to your change, if not please help us in assigning it to the right owner.

Labels: -Pri-1 Needs-Bisect Pri-2
Owner: krajshree@chromium.org
I'm quite sure that my CL mentioned has nothing to do with the issue.  I guess that the issue was caused by Blink.

krajshree@, may I ask you to do a more detailed bisect?

https://chromium.googlesource.com/chromium/src/+log/76af67653a03212a3d9104e3a94a44d773917214..7ec6a579b7e47f84b384b98b6f04ee65beb55758
where we can see "Roll src/third_party/WebKit 65258a2:55df775 (svn 197117:197124)" for example.  Can you bisect which roll of WebKit caused the issue?  Then, I think you can find a better owner.

Labels: -Pri-2 -Needs-Bisect Pri-1
Owner: skobes@chromium.org
Able to reproduce the issue on Windows 10, Ubuntu 14.04, Mac OS 10.11.6 using chrome stable #53.0.2785.116, reported chrome version #52.0.2743.116 and latest canary #55.0.2872.0.

Bisect Information:
=====================
Good build: 45.0.2432.0	
Bad Build : 45.0.2433.0

Change Log URL: https://chromium.googlesource.com/chromium/src/+log/d9418440b42c22114448813f8d8341d22c9c0a51..fcbc256d04f63d5ef61b481e3a984a2c65e238ce

Blink changelog:
https://chromium.googlesource.com/chromium/blink/+log/97081fe..ddd1e8f

From the above change log suspecting below change

Review URL: https://codereview.chromium.org/1172153004

skobes@ - Could you please check whether this is caused with respect to your change, if not please help us in assigning it to the right owner.

Comment 8 by e...@chromium.org, Oct 18 2016

Steve, could you take a look at this when you get a chance? Looks like it was caused by your change https://codereview.chromium.org/1172153004

Comment 9 by skobes@chromium.org, Oct 18 2016

Status: Started (was: Assigned)
Project Member

Comment 11 by bugdroid1@chromium.org, Oct 20 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1abcf254f3e55be39879ca76d268192faf64a254

commit 1abcf254f3e55be39879ca76d268192faf64a254
Author: skobes <skobes@chromium.org>
Date: Thu Oct 20 17:17:17 2016

Don't update height in LayoutBlock::markFixedPositionObjectForLayoutIfNeeded.

This prevented LayoutBlockFlow::layoutBlockFlow from seeing that the height had
changed.  As a result, it failed to set relayoutChildren = true.

The bug affects fixed-pos objects in abs-pos containers, so it got worse with
http://crrev.com/1172153004 but existed before.

An equivalent patch for width was made in http://crrev.com/171463003.

Additionally, update the early return for the case where the abs-pos container
is the LayoutView.  (This is just an optimization.)

BUG= 640639 

Review-Url: https://chromiumcodereview.appspot.com/2429113002
Cr-Commit-Position: refs/heads/master@{#426515}

[add] https://crrev.com/1abcf254f3e55be39879ca76d268192faf64a254/third_party/WebKit/LayoutTests/fast/block/positioning/fixed-in-abs-height-change.html
[modify] https://crrev.com/1abcf254f3e55be39879ca76d268192faf64a254/third_party/WebKit/Source/core/layout/LayoutBlock.cpp
[modify] https://crrev.com/1abcf254f3e55be39879ca76d268192faf64a254/third_party/WebKit/Source/core/layout/LayoutBox.cpp
[modify] https://crrev.com/1abcf254f3e55be39879ca76d268192faf64a254/third_party/WebKit/Source/core/layout/LayoutBox.h

Status: Fixed (was: Started)
Verified in canary (56.0.2897.0).

Sign in to add a comment