1817762 - 6.44% perf_reftest_singletons getElementById-1.html (Windows) regression on Fri February 17 2023

Status: RESOLVED INVALID

(Core :: XUL, defect)

Description

[Acasandrei Beatrice (needinfo me)]

Perfherder has detected a talos performance regression from push 60ebe844b93bf6f069589075b6e4474483aa4a47. 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)
6%	perf_reftest_singletons getElementById-1.html	windows10-64-shippable-qr	e10s fission stylo webrender	47.23 -> 50.27

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 offending 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

[Emilio Cobos Álvarez [:emilio]]

Can we double-check the regression range? It doesn't make sense, that patch shouldn't even change the generated code.

Comment 2

[Timothy Nikkel (:tnikkel)]

Attachment: Screen Shot 2023-02-20 at 6.26.18 AM.png

The vertical line is Emilio's commit. If anything it looks like the next commit introduced an improvement in this test. Seems more like noise than signal here.

Comment 3

[Alexandru Ionescu (needinfo me) [:alexandrui]]

I retriggered on culprit, comeback and the revision before, up to 20 retriggers total. The result should confirm or infirm the noise.

Comment 4

[Acasandrei Beatrice (needinfo me)]

(In reply to Alexandru Ionescu (needinfo me) [:alexandrui] from comment #3)

I retriggered on culprit, comeback and the revision before, up to 20 retriggers total. The result should confirm or infirm the noise.

Based on the retriggers the regression looks valid.

Comment 5

[Timothy Nikkel (:tnikkel)]

Attachment: Screen Shot 2023-02-21 at 3.49.22 AM.png

Graph still looks the same to me. If there's a regression on Emilio's push then there is an improvement on then the next push.

Comment 6

[Alexandru Ionescu (needinfo me) [:alexandrui]]

If the regression was invalid, then we were expecting all the 3 retriggered revisions to fill with datapoints in the same interval. But the graph is noisier than that. The retriggers from 58ce88d4f5dab reveals that this regression is invalid, though.
Beatrice, can you please close this as invalid?