1809764 - Webpage causes tab to consume ever-growing amounts of memory and does not render

Status: VERIFIED FIXED

(Core :: Layout: Columns, defect, P2)

Description

[Mark A. Hershberger (hexmode)]

I initially filed this as https://github.com/webcompat/web-bugs/issues/116682 but I'm pretty sure this is something in Firefox, not a web page that just isn't working right.

The page at https://ballotpedia.org/Arkansas_Supreme_Court does not work in Firefox, but works in other browsers.

In firefox, the browser spins and spins. Opening about:processes shows that memory usage is growing continuously.

Other pages on the site function ok.

Comment 1

[Mike Conley (:mconley) (:⚙️)]

Looks like we're in some kind of degenerate layout case:

https://share.firefox.dev/3QCnMkK

Comment 2

[Mark A. Hershberger (hexmode)]

I Just tried to load the page in 111.0a1 (2023-01-17) and it worked.

Has there been a change in the code that could have targeted this?

Comment 3

[Emilio Cobos Álvarez (:emilio)]

I can repro, there's a bunch of multicol reflow on the stack, I think we're getting stuck in column balancing... Ting-Yu, any chance you can take a look?

Comment 4

[Emilio Cobos Álvarez (:emilio)]

Attachment: Arkansas Supreme Court - Ballotpedia (1_26_2023 2_14_22 PM).html

File that seems to repro the issue locally.

Comment 5

[Emilio Cobos Álvarez (:emilio)]

I think this is a bug in the multicol code since virtually all time is spent under FindBestBalanceBSize. I tried to get a regression range but it's weird because it sometimes works... I can try later.

Comment 6

[Alice0775 White]

The browser hang up.
Regression window:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=7cae93f348a7f209c35b545a3b2a18dacc0a8477&tochange=cad5fc248dfe0d6116682b8a1d7ee9cbaeab1763

Comment 7

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

Set release status flags based on info from the regressing bug 793686

Comment 8

[Ting-Yu Lin [:TYLin] (PST, UTC-8)]

I think we're getting stuck in column balancing... Ting-Yu, any chance you can take a look?

Yes, we are stuck in column balancing loop with a nested multi-column layout where the inner multi-column contains a tall break-inside:avoid element. (Keep the NI as a reminder to look deeper into the problem.)

Comment 9

[Ting-Yu Lin [:TYLin] (PST, UTC-8)]

Attachment: Bug 1809764 Part 1 - Add more COLUMN_SET_LOG to debug nested multicols. r?emilio

Comment 10

[Ting-Yu Lin [:TYLin] (PST, UTC-8)]

Attachment: Bug 1809764 Part 2 - Ignore "break-inside:avoid" for content in a nested top-of-page multicol in last balancing reflow. r?emilio

The only difference between multicol-nested-030.html and
multicol-nested-031.html is the column-fill property on the outer column.

WARNING: loading 030.html and 031.html without this patch can trap Firefox in
infinite loops.

Depends on D169004

Comment 11

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/156cb3db4033
Part 1 - Add more COLUMN_SET_LOG to debug nested multicols. r=emilio
https://hg.mozilla.org/integration/autoland/rev/cdd9f798f751
Part 2 - Ignore "break-inside:avoid" for content in a nested top-of-page multicol in last balancing reflow. r=emilio

Comment 12

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

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/38386 for changes under testing/web-platform/tests

Comment 13

[Cristian Tuns]

https://hg.mozilla.org/mozilla-central/rev/156cb3db4033
https://hg.mozilla.org/mozilla-central/rev/cdd9f798f751

Comment 14

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

Upstream PR merged by moz-wptsync-bot

Comment 17

[Ardelean Oana, Desktop Test Engineering [:oardelean]]

Reproducible on a 2023-01-20 Nightly build on macOS 12.
Verified as fixed on Firefox 111.0b5(build ID: 20230223185944) and Nightly 112.0a1(build ID: 20230223172038) on macOS 12, Windows 11, Ubuntu 22.