1928180 - 9.51% tresize (Linux) regression on Mon October 21 2024

Status: RESOLVED FIXED

(Firefox :: Address Bar, defect, P2)

Description

[Treeherder Bug Filer]

Perfherder has detected a talos performance regression from push 42d3134758c767b29889b2a3184fc869cc184d63. 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)
10%	tresize	linux1804-64-shippable-qr	e10s fission stylo webrender-sw	20.64 -> 22.60

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
7%	tart	windows11-64-shippable-qr	e10s fission stylo webrender-sw	1.79 -> 1.67
6%	tart	macosx1015-64-shippable-qr	e10s fission stylo webrender-sw	1.78 -> 1.67
5%	tart	windows11-64-shippable-qr	e10s fission stylo webrender	1.84 -> 1.75

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 2610

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 1921811

Comment 2

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

:sfoster, since you are the author of the regressor, bug 1921811, could you take a look? Also, could you set the severity field?

For more information, please visit BugBot documentation.

Comment 3

[Sam Foster [:sfoster] (he/him)]

try push for perf --alert 2610

Comment 4

[Christoph Kerschbaumer [:ckerschb]]

Hey Sam, I am the REO for Fx133 and I am checking in because this bug is marked affecting Fx133. I see you did a try push and you set the priority and severity of the bug. Did you gain some insights from the TRY push, or still not critical to fix in Fx133? Please let me know. Thank you.

Comment 5

[Sarah Clements [:sclements]]

There's been subsequent changes to the regressing bug, see bug 1927111 and bug 1928255 (which we'll very likely uplift to 133) so its unclear if this is still going to be an issue. Perhaps Emilio can help with your question?

Comment 6

[Emilio Cobos Álvarez (:emilio)]

I'm looking at the tresize graph, and yes, it seems like there was a regression from bug 1921811, but then a subsequent improvement from bug 1927111, in a way such that we ended up with better performance than before. Andrej, Mayank, could you confirm my read of the graph? If so, I think we can dupe this to bug 1927111.

Comment 7

[Mayank Bansal]

improvement from bug 1917458 (25.64ms) [16 Sep]
improvement from bug 1921257 (23.20ms) (28sep)
Regression on Backout of bug 1921257 (24.13ms) (2oct) - And looks like this bug never relanded.
Improvement from bug 1922546 (23.4ms) [3 Oct]
Improvement from bug 1923334 (20.8ms) [10Oct]
regression from bug 1921811 (22.6ms) [21Oct]
improvement from bug 1927111 (19ms) [30Oct]

(ignoring the land-and-backout of the box-shadow work, and the X11 gpu process)

Does this help?

Comment 8

[Emilio Cobos Álvarez (:emilio)]

Yeah, seems like we're good here, lmk if anyone disagrees :)