1977839 - 7.43 - 4.37% stylebench Descendant and child combinators/Mutating attributes - 2/Sync / stylebench Descendant and child combinators/Removing leaf elements - 4 + 6 more (Windows) regression on Tue July 15 2025

Status: RESOLVED FIXED

(Core :: Layout: Form Controls, defect)

Description

[Treeherder Bug Filer]

Perfherder has detected a browsertime performance regression from push 197645ffd8065cb1c8bc1137f804091695494284. As author of one of the patches included in that push, we need your help to address this regression.

Please acknowledge, and begin investigating this alert within 3 business days, or the patch(es) may be backed out in accordance with our regression policy. Our guide to handling regression bugs has information about how you can proceed with this investigation.

If you have any questions or need any help with the investigation, please reach out to [e-mail]. Alternatively, you can find help on Slack by joining #perf-help, and on Matrix you can find help by joining #perftest.

Regressions:

Ratio	Test	Platform	Options	Absolute values (old vs new)	Performance Profiles
7%	stylebench Descendant and child combinators/Mutating attributes - 2/Sync	windows11-64-24h2-shippable	fission webrender	2.06 -> 2.22	Before/After
7%	stylebench Descendant and child combinators/Removing leaf elements - 4/Sync	windows11-64-24h2-shippable	fission webrender	3.59 -> 3.84	Before/After
6%	stylebench Descendant and child combinators/Mutating attributes - 2/Sync	windows11-64-24h2-shippable	fission webrender	2.08 -> 2.22	Before/After
6%	stylebench Descendant and child combinators/Removing leaf elements - 4/Sync	windows11-64-24h2-shippable	fission webrender	3.62 -> 3.82	Before/After
5%	stylebench Descendant and child combinators/Removing leaf elements - 4	windows11-64-24h2-shippable	fission webrender	4.59 -> 4.84	Before/After
5%	stylebench Descendant and child combinators/Mutating attributes - 2	windows11-64-24h2-shippable	fission webrender	3.06 -> 3.22	Before/After
4%	stylebench Descendant and child combinators/Mutating attributes - 2	windows11-64-24h2-shippable	fission webrender	3.08 -> 3.22	Before/After
4%	stylebench Descendant and child combinators/Removing leaf elements - 4	windows11-64-24h2-shippable	fission webrender	4.61 -> 4.82	Before/After

Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests.

If you need the profiling jobs you can trigger them yourself from treeherder job view or ask [e-mail] to do that for you.

You can run all of these tests on try with ./mach try perf --alert 45983

The following documentation link provides more information about this command.

Comment 1

[Emilio Cobos Álvarez [:emilio]]

I think bug 1977669 might've had something to do with it. Otherwise I'll start investigating / trying to make it faster. I'm guessing the button checks could be faster...

Comment 2

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

Set release status flags based on info from the regressing bug 1873301

Comment 3

[Emilio Cobos Álvarez [:emilio]]

Attachment: WIP: Bug 1977839 - Speed up IsButtonLike.

WIP because it's pretty tentative, I'm not sure why this would be the
case but it's the only thing from the regressing patch that makes sense
to me?

Try: https://perf.compare/compare-lando-results?baseLando=143175&newLando=143176&baseRepo=try&newRepo=try&framework=13

Comment 4

[Emilio Cobos Álvarez [:emilio]]

Attachment: WIP: Bug 1977839 - Try: Less overflow hidden overhead.

Just another hunch at this point: https://perf.compare/compare-lando-results?baseLando=143204&newLando=143220&baseRepo=try&newRepo=try&framework=13

Comment 5

[Emilio Cobos Álvarez [:emilio]]

Attachment: WIP: Bug 1977839 - Speed up IsReplaced().

https://perf.compare/compare-lando-results?baseLando=143337&newLando=143338&baseRepo=try&newRepo=try&framework=13

Comment 6

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

Set release status flags based on info from the regressing bug 1873301

Comment 7

[Emilio Cobos Álvarez [:emilio]]

So I'm a bit confused at this regression, but I realized that, when I look at the alert summary now, there's a bunch of other stuff, including improvements in similar stylebench subtests.

So I'm a bit confused over-all, could I get a good overview of where my patch ended up performance wise? I just don't want to end up wasting a lot of time chasing stuff if the answer is "stuff just moved around", as it was a rather significant change.

Comment 8

[Andra Esanu (needinfo me)]

Hi! This graph can be an example, it's for stylebench on windows.
And this graph is for booking FirstVisualChange on android.

Does this help?

Comment 9

[Emilio Cobos Álvarez [:emilio]]

Well, what I meant was: I can see my patch regressed some stylebench subtests, but looking at the alerts it also improved others in roughly the same amount, correct?

So I think that was mostly performance neutral for stylebench, and I should try to dig into the amazon/booking regressions. Would you agree with that assessment? Or am I misreading the alert page?

Comment 10

[Andra Esanu (needinfo me)]

Yes, it would be a good idea to dig into the amazon/booking regressions.

:sparky How would you interpret the stylebench performance?

Comment 11

[Greg Mierzwinski [:sparky]]

I agree with your assessment :emilio regarding ignoring stylebench and focusing on the large pageload changes. Looking at the stylebench results in perfcompare, the overall score didn't change at all: https://perf.compare/compare-results?baseRev=24935e432860eb5f2c5c2c7ffab8342dac8a1461&baseRepo=autoland&newRev=197645ffd8065cb1c8bc1137f804091695494284&newRepo=autoland&framework=13&search=stylebench

See the graph here too: https://treeherder.mozilla.org/perfherder/graphs?highlightAlerts=1&highlightChangelogData=1&highlightCommonAlerts=0&replicates=0&series=autoland,5259859,1,13&timerange=2592000&zoom=1752194091406,1753234218814,469.3876466349294,520.2598016944887

That said, something to keep in mind is that I don't know the specifics of what those subtests are testing so it's possible the regressing ones test something more important than what the ones that improved test. It could be useful to double-check this just in case.

Comment 12

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

It has been over 7 days with no activity on this performance regression.

:emilio, since you are the author of the regressor, bug 1873301, which triggered this performance alert, could you please provide a progress update?

If this regression is something that fixes a bug, changes the baseline of the regression metrics, or otherwise will not be fixed, please consider closing it as WONTFIX. See this documentation for more information on how to handle regressions.

For additional information/help, please needinfo the performance sheriff who filed this alert (they can be found in comment #0), or reach out in #perftest, or #perfsheriffs on Element.

For more information, please visit BugBot documentation.

Comment 13

[Emilio Cobos Álvarez [:emilio]]

I still think it's worth digging into though I'm don't have any other interesting idea there... I should try the wip patches again to look at whether they move the pageload numbers but...

Comment 14

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

It has been over 7 days with no activity on this performance regression.

:emilio, since you are the author of the regressor, bug 1873301, which triggered this performance alert, could you please provide a progress update?

If this regression is something that fixes a bug, changes the baseline of the regression metrics, or otherwise will not be fixed, please consider closing it as WONTFIX. See this documentation for more information on how to handle regressions.

For additional information/help, please needinfo the performance sheriff who filed this alert (they can be found in comment #0), or reach out in #perftest, or #perfsheriffs on Element.

For more information, please visit BugBot documentation.

Comment 15

[Emilio Cobos Álvarez [:emilio]]

See above?

Comment 16

[Emilio Cobos Álvarez [:emilio]]

I think bug 1984884 fixed this, looking at the graph.

Comment 17

[Greg Mierzwinski [:sparky]]

(In reply to Emilio Cobos Álvarez (:emilio) from comment #15)

See above?

:emilio, in the future, if this is a bug you want to keep open but are unlikely to get to resolving soon, then you can add the backlog-deferred keyword to stop the bot from asking for updates. Please only use it sparingly though if possible.

Comment 18

[Mathew Hodson]

Are we sure this was fixed on version 142 and 143? There were two patches in bug 1984884 but only one was uplifted.

Comment 19

[Emilio Cobos Álvarez [:emilio]]

Yes