Bug 225442 – will-change: contain should create a containing block and a stacking context

• Status: RESOLVED
• Resolution: FIXED
• Component: Layout and Rendering

[Tim Nguyen (:ntim)]

https://github.com/web-platform-tests/wpt/blob/master/css/css-will-change/will-change-fixpos-cb-contain-1.html

[Rob Buis]

This is possibly already done in https://bugs.webkit.org/show_bug.cgi?id=224742 but I am fine if this bug covers it (and in fact my change may only be a subset of this bug).

[Radar WebKit Bug Importer]

<rdar://problem/77958378>

[Tim Nguyen (:ntim)]

Rob, I'm tentatively assigning this to you. 

Relevant code is here:

https://webkit-search.igalia.com/webkit/rev/7c76a6e054902182b30fd3ce7f3a08f507dc1f1f/Source/WebCore/rendering/style/WillChangeData.cpp#63-78


Adding the property here will cause it to create a containing block for both position: fixed & position: absolute cases.

Please let me know if you have any questions.

[Rob Buis]

Created attachment 428784 [details]
Patch

[Rob Buis]

Created attachment 428789 [details]
Patch

[Rob Buis]

Created attachment 428803 [details]
Patch

[Tim Nguyen (:ntim)]

Comment on attachment 428803 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=428803&action=review

> Source/WebCore/ChangeLog:3
> +        will-change: contain should create a containing block

I think the CSS stacking context change is important to mention in the main message and in LayoutTests/ChangeLog as well.

> LayoutTests/TestExpectations:4926
> +webkit.org/b/208988 imported/w3c/web-platform-tests/webxr/xrWebGLLayer_opaque_framebuffer_stencil.https.html [ Skip ]

nit: unrelated change

[Tim Nguyen (:ntim)]

You'll probably need to rebase on top of bug 225443 btw (sorry for the conflict!).

[Rob Buis]

Created attachment 428820 [details]
Patch

[Rob Buis]

Comment on attachment 428803 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=428803&action=review

>> Source/WebCore/ChangeLog:3
>> +        will-change: contain should create a containing block
> 
> I think the CSS stacking context change is important to mention in the main message and in LayoutTests/ChangeLog as well.

Done.

>> LayoutTests/TestExpectations:4926
>> +webkit.org/b/208988 imported/w3c/web-platform-tests/webxr/xrWebGLLayer_opaque_framebuffer_stencil.https.html [ Skip ]
> 
> nit: unrelated change

This is a vi(m) setting problem that comes up from time to time, I think I figured it out. Done!

[EWS]

Committed r277580 (237807@main): <https://commits.webkit.org/237807@main>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 428820 [details].

[Tim Nguyen (:ntim)]

(In reply to Rob Buis from comment #10)
> Comment on attachment 428803 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=428803&action=review
> 
> >> Source/WebCore/ChangeLog:3
> >> +        will-change: contain should create a containing block
> > 
> > I think the CSS stacking context change is important to mention in the main message and in LayoutTests/ChangeLog as well.
> 
> Done.

It doesn't matter too much since things are in the details, but that part wasn't done.

The main commit message should have been "will-change: contain should create a containing block & a stacking context".

Anyway, I'll rename this bug for clarity.