2083404 - Subtracting the result of hypot in calc in CSS causes addition instead

Status: RESOLVED FIXED

(Core :: CSS Parsing and Computation, defect)

Description

[Auberon López]

Attachment: Demo HTML file reproducing the calc error

Steps to reproduce

When hypot is subtracted from a value inside of calc in CSS, addition is performed instead of subtraction. For example, calc(100px - hypot(30px, 40px)) should evaluate to 50px, but it actually evaluates to 150px. The attached demo file visualizes this issue.

Steps to reproduce:

1. Start Firefox by clicking on the desktop icon.
2. Open the attached file.

Tested on
Ubuntu 26.0.4.1, with Firefox 154.0 AND Nightly 160.0a1 (2026-10-17)
MacOS 26.6.2 Firefox 157.0.1

AI Disclaimer
The example code was generated using AI but checked by hand. The bug report was written by hand.

Actual results

Multiple calc values were incorrectly evaluated, causing the test page to show red bars representing the delta. In each case except for the control case the value of hypot was added instead of subtracted.

Expected results

calc should have been correctly evaluated, with hypot being subtracted. This would have resulted in all green bars with no red.

Comment 1

[BugBot [:suhaib / :marco]]

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

Comment 2

[BugBot [:suhaib / :marco]]

The severity field is not set for this bug.
:tlouw, could you have a look please?

For more information, please visit BugBot documentation.

Comment 3

[Daniel Holbert [:dholbert]]

Thanks for the bug report!

It looks like this dates back to when we added support for hypot(), in bug 1814469 -- I can reproduce this bug all the way back to Nightly 2023-03-05, the first build with that bug's patch.

One clue -- Bug 1814469 comment 2 mentioned what-sounds-like this exact issue (an inability to negate hypot() as well as abs() at the time).

And then later on, it looks like we fixed the analogous bug for abs() in bug 1863779. But we never fixed it for hypot(). So: likely we need toapply the same approach as in bug 1863779, but now for hypot().

Leaving ni=tlouw open since I suspect he's among the best-prepped to work on this, having worked on bug 1863779. :)

Comment 4

[Daniel Holbert [:dholbert]]

Actually this looks easy enough; I'll just take it.

Comment 5

[Daniel Holbert [:dholbert]]

Attachment: Bug 2083404: Remove invalid special-case for negating hypot() expressions in calc(). r?tlouw,#style

Before this patch, we mishandled negation of 'hypot()' in a special-case that
propagated the negation to the children. This was ineffective and wrong; it
still produced a positive result, due to the squaring arithmetic that happens
in a hypot() operation.

This patch makes hypot() share the same codpath that we use for many other math
functions, implementing negation using wrap_self_in_negate().

Comment 6

[Daniel Holbert [:dholbert]]

https://treeherder.mozilla.org/jobs?repo=try&landoInstance=lando-prod-2025&landoCommitID=110787

Comment 7

[Pulsebot]

Pushed by [e-mail]:
https://github.com/mozilla-firefox/firefox/commit/a362b3e9a326
https://hg.mozilla.org/integration/autoland/rev/e7324d6ddff6
Remove invalid special-case for negating hypot() expressions in calc(). r=firefox-style-system-reviewers,emilio

Comment 8

[Iulian Moraru]

https://hg.mozilla.org/mozilla-central/rev/e7324d6ddff6

Comment 10

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

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/64703 for changes under testing/web-platform/tests

Comment 11

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

Upstream PR merged by moz-wptsync-bot