1775236 - 5.75% assorted-dom (OSX) regression on Thu June 9 2022

Status: RESOLVED WORKSFORME

(Core :: JavaScript Engine: JIT, defect, P3)

Description

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

Perfherder has detected a browsertime performance regression from push 774256e8b6a57ae4a5482f2581a2e66caf7c1878. 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%	assorted-dom	macosx1015-64-shippable-qr	fission webrender	17.33 -> 18.33

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 offending patch(es) will 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]]

Set release status flags based on info from the regressing bug 1770366

Comment 2

[Jan de Mooij [:jandem]]

Looking at the 30 day graph, this is actually better than it was before since June 10 or so.

Comment 3

[Nicolas B. Pierron [:nbp] {away}]

Jan, are the other changes related to JIT frames? Or is it that the test case became bi-modal and more runs are necessary to distinguish differences?

Comment 4

[Jan de Mooij [:jandem]]

Looking at the 30 day graph, this is a bit better than it was before. The getprop-performance-getter sub test that the alert shows as regressing the most, shows an improvement for me around the time bug 1770366 landed:

https://treeherder.mozilla.org/perfherder/graphs?highlightAlerts=1&highlightChangelogData=1&highlightCommonAlerts=0&highlightedRevisions=b5817546051a&highlightedRevisions=774256e8b6a5&series=autoland,3930354,1,13&timerange=2592000

It looks like bug 1770366 was an improvement, but it was backed out temporarily and then the graph went back to the worse numbers (maybe that triggered the alert) and then it improved again when the changes relanded.