1920525 - content-visibility: auto causes object-fit to no longer apply once image element has left and re-entered the viewport

Status: NEW

(Core :: Layout, defect)

Description

[Lewis Lane]

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36

Steps to reproduce:

Create a wrapping element with a fixed aspect ratio. Add an image with a visibly different aspect ratio to it with width and height set to 100%, object-fit set to contain and content-visibility set to auto. Observe that object-fit keeps the visible image proportions correct. Scroll the page so the image has fully left the viewport. Scroll the image back into view.

Bare bones CodePen example here: https://codepen.io/enigmalewislane/pen/VwowpNm

Actual results:

The image reappears without object-fit applied. Trying to change the value of object-fit or enabling/disabling it in dev tools has no effect, and the image remains stretched/squashed.

Expected results:

The image should reappear looking the same as it did when it left the viewport.

Comment 1

[BugBot [:suhaib / :marco/ :calixte]]

The Bugbug bot thinks this bug should belong to the 'Core::Layout' component, and is moving the bug to that component. Please correct in case you think the bot is wrong.

Comment 2

[Alice0775 White]

Regression window:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=f89991d1b2cdd13120304730efba7878d4cfac5b&tochange=af2192d1c5378cb3d6d9334ddcf52500f94bc385

Comment 3

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1804761

Comment 4

[Dianna Smith [:diannaS]]

:daniel do you think you can take a look and assign a priority/severity here?

Comment 5

[Daniel Holbert [:dholbert]]

Attachment: testcase 1 (no interaction needed)

Thanks for the bug report!

Here's a version of the testcase that automatically scrolls down+up after 500ms and triggers the bug (for me at least).

Comment 6

[Daniel Holbert [:dholbert]]

Triaging as S3 (regular-severity) given we've been shipping with this for a while. Seems like it's probably not too complex of a fix, though. Leaving ni=me to investigate further (probably can't today but hope to do so soon).