1714333 - 4.86 - 3.68% Images / Images + 1 more (Linux, Windows) regression on Mon May 31 2021

Status: RESOLVED INVALID

(Core :: Graphics: WebRender, defect)

Description

[Andra Esanu (needinfo me)]

Perfherder has detected a awsy performance regression from push 8f60f65d85d892a4a6195b185bb718f49f222808. As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
5%	Images		windows10-64-shippable-qr		7,070,208.14 -> 7,413,810.62
5%	Images		windows10-64-shippable-qr		7,075,175.74 -> 7,411,248.21
4%	Images		linux1804-64-shippable-qr		4,928,224.18 -> 5,109,409.46

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

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

Set release status flags based on info from the regressing bug 1713651

Comment 2

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

== Change summary for alert #30269 (as of Wed, 02 Jun 2021 10:39:01 GMT) ==

Regressions:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
11%	tp5o (docs)	responsiveness	linux1804-64-shippable-qr	e10s stylo webrender	2.47 -> 2.74

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=30269

Comment 3

[Florin Strugariu [:Bebe]]

== Change summary for alert #30308 (as of Mon, 07 Jun 2021 04:37:53 GMT) ==

Regressions:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
22%	stackoverflow	loadtime	android-hw-g5-7-0-arm7-api-16-shippable-qr	warm webrender	994.48 -> 1,214.29
20%	stackoverflow	loadtime	android-hw-g5-7-0-arm7-api-16-shippable-qr	warm webrender	1,009.65 -> 1,214.46
20%	stackoverflow	loadtime	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,032.79 -> 2,443.50
14%	youtube-watch	dcf	android-hw-g5-7-0-arm7-api-16-shippable-qr	warm webrender	1,250.79 -> 1,419.92
13%	youtube-watch	loadtime	android-hw-g5-7-0-arm7-api-16-shippable-qr	warm webrender	1,256.67 -> 1,425.54
11%	stackoverflow	LastVisualChange	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,863.83 -> 3,167.58
10%	stackoverflow	ContentfulSpeedIndex	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,724.25 -> 3,006.08
10%	stackoverflow	PerceptualSpeedIndex	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,730.00 -> 3,008.50
10%	stackoverflow	FirstVisualChange	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,720.25 -> 2,997.08
10%	stackoverflow	SpeedIndex	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	2,724.25 -> 3,001.58
8%	youtube-watch	dcf	android-hw-g5-7-0-arm7-api-16-shippable-qr	cold webrender	1,761.60 -> 1,903.46

Improvements:

Ratio	Suite	Test	Platform	Options	Absolute values (old vs new)
50%	google-slides	loadtime	linux1804-64-shippable-qr	cold webrender	3,794.08 -> 1,893.75
36%	google-slides	FirstVisualChange	windows10-64-shippable-qr	warm webrender	312.33 -> 201.33
27%	stackoverflow	ContentfulSpeedIndex	android-hw-p2-8-0-android-aarch64-shippable-qr	warm webrender	482.42 -> 354.08
27%	stackoverflow	FirstVisualChange	android-hw-p2-8-0-android-aarch64-shippable-qr	warm webrender	482.42 -> 354.08
27%	stackoverflow	LastVisualChange	android-hw-p2-8-0-android-aarch64-shippable-qr	warm webrender	482.42 -> 354.08
...	...	...	...	...	...
7%	google-slides	ContentfulSpeedIndex	linux1804-64-shippable-qr	warm webrender	702.75 -> 653.58

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=30308

Comment 4

[Florin Strugariu [:Bebe]]

aosmond can your commit cause android perf regressions?

Comment 5

[Marco Castelluccio [:marco]]

There were improvements from enabling the feature, and the disabling reverted the improvements. Effectively, we are back to the status before the feature was enabled.
When we'll enable it again, we'll see the improvements again.

We can either close this as INVALID, or wait for bug 1673653 to land.