1738894 - 8.92 - 2.82% cnn fnbpaint / cnn dcf + 1 more (Linux) regression on Wed October 27 2021

Status: RESOLVED WONTFIX

(Core :: Privacy: Anti-Tracking, defect)

Description

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

Perfherder has detected a browsertime performance regression from push 8c5791e2b256b3e5ecb9918faae745e47f58fd3d. 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)
9%	cnn fnbpaint	linux1804-64-shippable-qr	cold webrender	685.21 -> 746.33
6%	cnn dcf	linux1804-64-shippable-qr	cold webrender	1,195.33 -> 1,270.50
3%	cnn dcf	linux1804-64-shippable-qr	cold webrender	1,196.17 -> 1,229.88

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
9%	cnn SpeedIndex	linux1804-64-shippable-qr	warm webrender	1,447.33 -> 1,314.89
6%	cnn LastVisualChange	linux1804-64-shippable-qr	warm webrender	3,156.36 -> 2,971.76
6%	cnn PerceptualSpeedIndex	linux1804-64-shippable-qr	warm webrender	1,154.61 -> 1,087.12
5%	cnn loadtime	linux1804-64-shippable-qr	warm webrender	2,979.10 -> 2,834.21
5%	cnn ContentfulSpeedIndex	linux1804-64-shippable-qr	warm webrender	1,245.82 -> 1,185.64
3%	cnn LastVisualChange	windows10-64-shippable-qr	cold webrender	4,924.21 -> 4,791.92

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.

For more information on performance sheriffing please see our FAQ.

Comment 1

[Thomas Wisniewski [:twisniewski]]

:alexandrui, if this is true, then it's an expected perf hit that we will just have to live with.

What was happening is that instead of loading the real Amazon TAM ad/tracking scripts, we were loading a "shim" for them which is bundled with Firefox (as part of SmartBlock). That should of course be faster than loading the real scripts, but it was never intended to happen outside of Strict mode ETP or private browsing, because it causes unexpected site breakage.

And note that if this bug is really causing a regression, it should compensate for a false perf improvement we would begin seeing after https://bugzilla.mozilla.org/show_bug.cgi?id=1713698 landed.

Comment 2

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

Sounds reasonable to me.

Comment 3

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

Set release status flags based on info from the regressing bug 1737899