1789612 - 38.41 - 20.78% tsvgx / tsvgx + 3 more (Linux, OSX) regression on Wed August 31 2022

Status: RESOLVED FIXED

(Core :: Graphics: WebRender, defect, P3)

Description

[Andra Esanu]

Perfherder has detected a talos performance regression from push 105858693e9a1ad58afa7309eb38f31ace500d2a. 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)
38%	tsvgx	macosx1015-64-shippable-qr	e10s fission stylo webrender	120.61 -> 166.93
34%	tsvgx	macosx1015-64-shippable-qr	e10s fission stylo webrender	124.62 -> 166.47
33%	tsvgx	macosx1015-64-shippable-qr	e10s fission stylo webrender-sw	145.87 -> 194.24
21%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender	186.13 -> 225.19
21%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender-sw	202.86 -> 245.02

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

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

Set release status flags based on info from the regressing bug 1787706

Comment 2

[Jeff Muizelaar [:jrmuizel]]

https://sql.telemetry.mozilla.org/dashboard/test_38?p_Build%201=20220830092750&p_Build%202=20220817091029&p_Days%20to%20Query=7&p_OS%20Filter=%20&p_Probe=payload.processes.gpu.histograms.wr_rasterize_blobs_time&p_Start%20Date=2022-08-30&p_Start%20Date%202=2022-08-17&p_Table=telemetry.main_nightly&p_channel=nightly also suggests that blob rasterization time got worse.

Comment 3

[Nicolas Silva [:nical]]

Sad. Probably means the tile size should be incresed only when the screen resolution is large (or the blob visible area). I'll back it out and experiment some more in a week when I come back from PTO.

Comment 4

[Nicolas Silva [:nical]]

Attachment: Bug 1789612 - Backed out changeset 105858693e9a for causing tsvgx regression. r=#gfx-reviewers

Comment 5

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/540dcf23e80e
Backed out changeset 105858693e9a for causing tsvgx regression. r=gfx-reviewers,jgilbert

Comment 6

[Marian-Vasile Laza]

https://hg.mozilla.org/mozilla-central/rev/540dcf23e80e

Comment 7

[Pascal Chevrel:pascalc]

Does it need to be backed out in beta as well?

Comment 8

[Nicolas Silva [:nical]]

Yes I had missed that, thanks

Comment 9

[Nicolas Silva [:nical]]

Comment on attachment 9294910 [details]
Bug 1789612 - Backed out changeset 105858693e9a for causing tsvgx regression. r=#gfx-reviewers

Beta/Release Uplift Approval Request

• User impact if declined: Some things get faster some things get slower, we didn't expect things to get slower so we want to be on the safe side.
• Is this code covered by automated tests?: No
• Has the fix been verified in Nightly?: No
• Needs manual test from QE?: No
• If yes, steps to reproduce:
• List of other uplifts needed: None
• Risk to taking this patch: Low
• Why is the change risky/not risky? (and alternatives if risky): backout of a simple and recent change.
• String changes made/needed:
• Is Android affected?: Yes

Comment 10

[Pascal Chevrel:pascalc]

Comment on attachment 9294910 [details]
Bug 1789612 - Backed out changeset 105858693e9a for causing tsvgx regression. r=#gfx-reviewers

Approved for 106.0b3, thanks.

Comment 11

[Pascal Chevrel:pascalc]

https://hg.mozilla.org/releases/mozilla-beta/rev/d77889bf8fff

Comment 12

[Acasandrei Beatrice (needinfo me)]

(In reply to Pulsebot from comment #5)

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/540dcf23e80e
Backed out changeset 105858693e9a for causing tsvgx regression.
r=gfx-reviewers,jgilbert

== Change summary for alert #35450 (as of Thu, 22 Sep 2022 10:31:08 GMT) ==

Improvements:

Ratio	Test	Platform	Options	Absolute values (old vs new)
27%	tsvgx	macosx1015-64-shippable-qr	e10s fission stylo webrender-sw	197.48 -> 144.76
26%	tsvgx	macosx1015-64-shippable-qr	e10s fission stylo webrender	167.90 -> 124.47
21%	tsvgx	windows10-64-shippable-qr	e10s fission stylo webrender	209.81 -> 165.19
21%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender	227.89 -> 180.86
20%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender-sw	247.09 -> 196.95
18%	tsvgx	linux1804-64-shippable-qr	e10s fission stylo webrender	226.83 -> 185.84

For up to date results, see: https://treeherder.mozilla.org/perfherder/alerts?id=35450