1864158 - 4.76 - 2.46% Base Content Resident Unique Memory / Base Content Resident Unique Memory + 3 more (Linux, Windows) regression on Tue November 7 2023

Status: RESOLVED FIXED

(Core :: Internationalization, defect)

Description

[Acasandrei Beatrice (needinfo me)]

Perfherder has detected a awsy performance regression from push 35ff07f4726ad465baa0d222c0c55d48a1082600. As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

Ratio	Test	Platform	Options	Absolute values (old vs new)
5%	Base Content Resident Unique Memory	windows11-64-2009-shippable-qr	fission	10,365,952.00 -> 10,859,008.00
4%	Base Content Explicit	linux1804-64-shippable-qr	fission	9,782,288.00 -> 10,218,000.00
4%	Base Content Explicit	windows11-64-2009-shippable-qr	fission	10,526,053.33 -> 10,979,173.33
4%	Base Content Explicit	linux1804-64-shippable-qr	fission	9,785,360.00 -> 10,198,032.00
2%	Base Content Resident Unique Memory	linux1804-64-shippable-qr	fission	14,966,528.00 -> 15,334,229.33

Details of the alert can be found in the alert summary, including links to graphs and comparisons for each of the affected tests. Please follow our guide to handling regression bugs and let us know your plans within 3 business days, or the patch(es) may be backed out in accordance with our regression policy.

If you need the profiling jobs you can trigger them yourself from treeherder job view or ask a sheriff to do that for you.

For more information on performance sheriffing please see our FAQ.

Comment 1

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

Set release status flags based on info from the regressing bug 1859752

Comment 2

[Ryan VanderMeulen [:RyanVM]]

This seems like a pretty significant hit on memory usage. We probably need a better of understand of what's causing the regression here.

Comment 3

[André Bargull [:anba]]

Base revision: https://treeherder.mozilla.org/jobs?repo=try&revision=c8f3e51797d8c63ceab5f1be380108d613566b25
ICU reverted to tag cldr/2023-09-27: https://treeherder.mozilla.org/jobs?repo=try&revision=027104f61d82dd8142da7a2c7b309e4e765b7906
ICU reverted to tag cldr/2023-08-22: https://treeherder.mozilla.org/jobs?repo=try&revision=b5521d7f2323ecc3ce24be8d28a1ace59d8357af
ICU reverted to tag cldr/2023-07-19: https://treeherder.mozilla.org/jobs?repo=try&revision=1921644071abb4a5c25f9c7246fe7cab593b0109

No change for tag "cldr/2023-09-27":
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=c8f3e51797d8c63ceab5f1be380108d613566b25&newProject=try&newRevision=027104f61d82dd8142da7a2c7b309e4e765b7906&framework=4&page=1

No change for tag "cldr/2023-08-22":
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=c8f3e51797d8c63ceab5f1be380108d613566b25&newProject=try&newRevision=b5521d7f2323ecc3ce24be8d28a1ace59d8357af&framework=4&page=1

Changes are visible for tag "2023-07-19":
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=c8f3e51797d8c63ceab5f1be380108d613566b25&newProject=try&newRevision=1921644071abb4a5c25f9c7246fe7cab593b0109&framework=4&page=1

Comment 4

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

The bug is marked as tracked for firefox121 (nightly). We have limited time to fix this, the soft freeze is in 2 days. However, the bug still isn't assigned.

:Amir, could you please find an assignee for this tracked bug? Given that it is a regression and we know the cause, we could also simply backout the regressor. If you disagree with the tracking decision, please talk with the release managers.

For more information, please visit BugBot documentation.

Comment 5

[André Bargull [:anba]]

(In reply to BugBot [:suhaib / :marco/ :calixte] from comment #4)

The bug is marked as tracked for firefox121 (nightly). We have limited time to fix this, the soft freeze is in 2 days. However, the bug still isn't assigned.

Bug 1859752 was backed-out because of bug 1864612, so we no longer have to worry about the soft freeze.

Comment 6

[André Bargull [:anba]]

The memory regression is introduced by the new likely-subtags implementation. It seems like the new implementation keeps the likely-subtags in memory, which would explain the memory regression.

BAD: ffc449de624 "ICU-20777 Merge the likelySubtags implemention"
https://github.com/unicode-org/icu/commit/ffc449de6243b17857b5d67854bdec723bf06f21
https://treeherder.mozilla.org/jobs?repo=try&revision=7e747c38081c699a7cb2fdd965555a6ceabb1e66
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=027104f61d82dd8142da7a2c7b309e4e765b7906&newProject=try&newRevision=7e747c38081c699a7cb2fdd965555a6ceabb1e66&framework=4&page=1

GOOD: 27181e36a6b "ICU-22435 Add C API for Locale"
https://github.com/unicode-org/icu/commit/27181e36a6bed372482e5a11c25009fa4dd971f1
https://treeherder.mozilla.org/jobs?repo=try&revision=4c7562666d71b0bd40bb1a132de77537ea53cd52
https://treeherder.mozilla.org/perfherder/compare?originalProject=try&originalRevision=027104f61d82dd8142da7a2c7b309e4e765b7906&newProject=try&newRevision=4c7562666d71b0bd40bb1a132de77537ea53cd52&framework=4&page=1

Comment 7

[André Bargull [:anba]]

Upstream bug report to investigate the heap memory increase: https://unicode-org.atlassian.net/jira/software/c/projects/ICU/issues/ICU-22647