1983756 - Abspos table or start-aligned block with non-auto insets and auto margins is laid out at a huge negative position

Status: RESOLVED FIXED

(Core :: Layout: Positioned, defect, P3)

Description

[Oriol Brufau [:Oriol]]

Attachment: testcase

The table is not visible, it's laid out at -8947788

Bisection: https://hg-edge.mozilla.org/integration/autoland/pushloghtml?fromchange=bb710b7475eaecf3e8ed868af7f1ae062e7e0b0a&tochange=889105cc0018e8d2aa77407ef90f68adf878933f

See https://github.com/w3c/csswg-drafts/issues/11478 for some extra context.

Comment 1

[Oriol Brufau [:Oriol]]

Or instead of display: table, align-self: start.

Comment 2

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

:dshin, since you are the author of the regressor, bug 1920160, could you take a look? Also, could you set the severity field?

For more information, please visit BugBot documentation.

Comment 3

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

Set release status flags based on info from the regressing bug 1920160

Comment 4

[Emilio Cobos Álvarez [:emilio]]

Seems worth addressing sooner rather than later since it can cause missing content... But it's probably rare in practice?

Comment 5

[David Shin[:dshin]]

nsTableFrame::ComputeAutoSize returns unconstrained block size no matter what. This is because for tables, desired block size is computed after its children are reflowed.
That calls down from ReflowInput::InitAbsoluteConstraints where it's soon used for computation of auto margin in block direction.

Addressing this fundamentally probably requires a lot of table work. OTOH, it's probably a good idea to constrain the computed auto sizes to the abs-cb size so the content at least doesn't get hidden like this.

Comment 6

[Emilio Cobos Álvarez [:emilio]]

But per the above it's not specific to table, right? <div style="align-self: start"> also triggers this.

Comment 7

[Oriol Brufau [:Oriol]]

Basically any align-self different than auto, normal, stretch or center is broken.

The containing block of an abspos should be completely definite, and you resolve stretch correctly, so it's weird that you don't align correctly.

Comment 8

[David Shin[:dshin]]

(Argh, comment 1, investigating further)

Comment 9

[David Shin[:dshin]]

Right. When automatic block size becomes fit-content size size, we need to defer computation of auto margins to after layout, like we already do if we specify height: fit-content.

Comment 10

[Oriol Brufau [:Oriol]]

Wouldn't it be easier to always defer the computation of auto margins to after layout? That's how we do it in Servo.
https://github.com/servo/servo/blob/87fe202ded0fe9b0d61acda7ac81ab1714eb2389/components/layout/positioned.rs#L582-L600

Comment 11

[David Shin[:dshin]]

May be worth pursuing - our abs offset/margin resolution is not the easiest to reason with because of this. Though that's (Unfortunately) a sizeable refactoring that should be worked as a separate bug.

Comment 12

[David Shin[:dshin]]

Attachment: Bug 1983756: Properly compute deferred auto margin sizes when abspos element's auto-size is fit-content size. r=#layout

Comment 13

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/8a50d94240fb
https://hg.mozilla.org/integration/autoland/rev/af222f790798
Properly compute deferred auto margin sizes when abspos element's auto-size is fit-content size. r=layout-reviewers,emilio

Comment 14

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/54634 for changes under testing/web-platform/tests

Comment 15

[Cosmin Sabou [:CosminS]]

https://hg.mozilla.org/mozilla-central/rev/af222f790798

Comment 16

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot

Comment 17

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

The patch landed in nightly and beta is affected.
:dshin, is this bug important enough to require an uplift?

• If yes, please nominate the patch for beta approval.
  • See https://wiki.mozilla.org/Release_Management/Requesting_an_Uplift for documentation on how to request an uplift.
• If no, please set status-firefox143 to wontfix.

For more information, please visit BugBot documentation.

Comment 18

[David Shin[:dshin]]

It's a pretty low-risk fix, but OTOH, we're pretty late into the beta phase, and I think this case (content-fit + auto margins + non-auto insets) won't be hugely common (The regressing bug 1920160 was closed 10 months ago).