1952917 - 7.6 - 3.47% cpstartup content-process-startup / sessionrestore_no_auto_restore + 1 more (Linux) regression on Wed March 5 2025

Status: RESOLVED WORKSFORME

(Core :: IPC, defect)

Description

[Treeherder Bug Filer]

Perfherder has detected a talos performance regression from push 7a43eb20fb7c112e190821dec3c58fb8727e7b25. 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)
8%	cpstartup content-process-startup	linux1804-64-shippable-qr	e10s fission stylo webrender	55.92 -> 60.17
5%	cpstartup content-process-startup	linux1804-64-shippable-qr	e10s fission stylo webrender-sw	54.54 -> 57.25
3%	sessionrestore_no_auto_restore	linux1804-64-shippable-qr	e10s fission stylo webrender	936.21 -> 968.67

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
3%	tabpaint	linux1804-64-shippable-qr	e10s fission stylo webrender	55.16 -> 53.38
3%	tabpaint	linux1804-64-shippable-qr	e10s fission stylo webrender-sw	56.20 -> 54.61

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.

You can run all of these tests on try with ./mach try perf --alert 44218

The following documentation link provides more information about this command.

For more information on performance sheriffing please see our FAQ.

If you have any questions, please do not hesitate to reach out to [e-mail].

Comment 1

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

Set release status flags based on info from the regressing bug 1942129

Comment 2

[Alex Franchuk [:afranchuk]]

This is potentially caused by the same cause of bug 1951925. I'm trying out a fix and I'll look at these metrics, too.

Comment 3

[Alex Franchuk [:afranchuk]]

On a try run it appears that the fix for bug 1951925 returns the regressing values to normal.

Comment 4

[Dianna Smith [:diannaS]]

should we mark this as fixed for 138 or do we need to give it a little more time?

Comment 5

[Alex Franchuk [:afranchuk]]

Maybe wait a little longer; the metrics are a bit noisy.

Comment 6

[Mayank Bansal]

Curious why would there be improvements to tabpaint metric

Comment 7

[Alex Franchuk [:afranchuk]]

It appears that most of these metrics have returned to previous values (including the tabpaint improvements, possibly). sessionrestore_no_auto_restore is very noisy so it's hard to see whether this has returned, however based on the others going back I think we can close this. It's possible some other change is responsible for sessionrestore_no_auto_restore, too. I can try to run it with only my patches applied to the prior state to see whether my patches may be the cause (though again, it's such a small and noisy change that it may be difficult to measure well).

Comment 8

[Alex Franchuk [:afranchuk]]

Using perf.compare, I see no regression for sessionrestore_no_auto_restore when comparing 017c9f9793ecaed3435ad4fd2674fb1d6a9e6358 (the fix for bug 1951925) with revisions prior to 7a43eb20fb7c112e190821dec3c58fb8727e7b25. Closing this as WORKSFORME based on the fix in bug 1951925.