1910128 - 12.8 - 2.01% speedometer3 NewsSite-Nuxt/NavigateToUS/Sync / speedometer3 NewsSite-Nuxt/total + 21 more (Linux, OSX, Windows) regression on Mon July 22 2024

Status: RESOLVED WONTFIX

(Testing :: Raptor, defect)

Description

[Florin Bilt]

Perfherder has detected a browsertime performance change from push bebaa09716b033fea32fdbc1223ac5c05a42c5ad.

Regressions:

Ratio	Test	Platform	Options	Absolute values (old vs new)	Performance Profiles
13%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	linux1804-64-nightlyasrelease-qr	fission webrender	33.35 -> 37.62	Before/After
12%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	linux1804-64-shippable-qr	fission webrender	32.41 -> 36.24	Before/After
8%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	macosx1015-64-shippable-qr	fission webrender	30.18 -> 32.60	Before/After
7%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	macosx1015-64-nightlyasrelease-qr	fission webrender	31.32 -> 33.66	Before/After
7%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	windows11-64-shippable-qr	fission webrender	18.17 -> 19.53	Before/After
7%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	windows11-64-nightlyasrelease-qr	fission webrender	17.89 -> 19.20	Before/After
7%	speedometer3 NewsSite-Nuxt/NavigateToUS/Sync	macosx1015-64-nightlyasrelease-qr	fission webrender	31.10 -> 33.33	Before/After
6%	speedometer3 NewsSite-Nuxt/NavigateToUS/total	linux1804-64-nightlyasrelease-qr	fission webrender	77.77 -> 82.72	Before/After
6%	speedometer3 Charts-observable-plot/Dotted/Async	linux1804-64-shippable-qr	fission webrender	8.02 -> 8.48	Before/After
6%	speedometer3 NewsSite-Nuxt/NavigateToUS/total	linux1804-64-shippable-qr	fission webrender	76.28 -> 80.48	Before/After
...	...	...	...	...	...
3%	speedometer3 NewsSite-Nuxt/NavigateToUS/total	windows11-64-nightlyasrelease-qr	fission webrender	38.58 -> 39.89	Before/After
3%	speedometer3 NewsSite-Nuxt/total	linux1804-64-nightlyasrelease-qr	fission webrender	219.50 -> 225.85	Before/After
2%	speedometer3 NewsSite-Nuxt/total	windows11-64-shippable-qr	fission webrender	106.83 -> 109.08	Before/After
2%	speedometer3 NewsSite-Nuxt/total	macosx1015-64-shippable-qr	fission webrender	183.32 -> 187.16	Before/After
2%	speedometer3 NewsSite-Nuxt/total	macosx1015-64-nightlyasrelease-qr	fission webrender	188.26 -> 192.05	Before/After

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)	Performance Profiles
4%	speedometer3 TodoMVC-jQuery/CompletingAllItems/Async	windows11-64-nightlyasrelease-qr	fission webrender	2.04 -> 1.95	Before/After
4%	speedometer3 TodoMVC-JavaScript-ES5/CompletingAllItems/Async	linux1804-64-shippable-qr	fission webrender	2.89 -> 2.77	Before/After
3%	speedometer3 TodoMVC-JavaScript-ES5/CompletingAllItems/Async	linux1804-64-nightlyasrelease-qr	fission webrender	2.92 -> 2.83	Before/After
3%	speedometer3 TodoMVC-React-Complex-DOM/CompletingAllItems/Async	linux1804-64-shippable-qr	fission webrender	6.03 -> 5.83	Before/After
3%	speedometer3 TodoMVC-JavaScript-ES6-Webpack-Complex-DOM/CompletingAllItems/Async	windows11-64-nightlyasrelease-qr	fission webrender	3.88 -> 3.77	Before/After
...	...	...	...	...	...
2%	speedometer3 NewsSite-Next/NavigateToWorld/total	macosx1400-64-shippable-qr	fission webrender	18.25 -> 17.86	Before/After

As author of one of the patches included in that push, we need your help to address this regression.
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 these tests on try with ./mach try perf --alert 1463

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 1867135

Comment 2

[Alex Finder]

Based on the retriggers I made on the graph before the culprit revision, it doesn't seem like this is an infra. It's weird that the changes would have caused these performance changes, considering this is just a patch to use the in-tree resources instead of cloning the speedometer repo before running the test. Also, running mach try perf did not report any performance changes on try. I'm proposing we close the defect as WONTFIX and add the "harness" tag to the summary. What do you think sparky ?

Comment 3

[Greg Mierzwinski [:sparky]]

I think what happened here is that we updated the speedometer3 code that the PGO stage is using and this caused some changes from there. I agree with closing this as WONTFIX but I'll mention this to the sp3 team so they're aware of it. I'm talking about the installer size change here: https://bugzilla.mozilla.org/show_bug.cgi?id=1867135#c5

Regarding the changes in this bug, I agree it's definitely a harness change.

Comment 4

[Alex Finder]

(In reply to Greg Mierzwinski [:sparky] from comment #3)

I think what happened here is that we updated the speedometer3 code that the PGO stage is using and this caused some changes from there. I agree with closing this as WONTFIX but I'll mention this to the sp3 team so they're aware of it. I'm talking about the installer size change here: https://bugzilla.mozilla.org/show_bug.cgi?id=1867135#c5

Regarding the changes in this bug, I agree it's definitely a harness change.

Thanks for replying sparky, closing the defect as WONTFIX.