1586257 - 2% ts_paint (linux64-shippable-qr) regression on push d7f1f760cd4582ab452b04fdea0425a4072a0295 (Thu October 3 2019)

Status: RESOLVED WONTFIX

(Core :: Graphics: WebRender, defect)

Description

[Florin Strugariu [:Bebe]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=845c92fd34a67a3924abf7341224279c02fbab24

As author of one of the patches included in that push, we need your help to address this regression.

Regressions:

2% ts_paint linux64-shippable-qr opt e10s stylo 271.88 -> 277.67
2% ts_paint linux64-shippable-qr opt e10s stylo 272.46 -> 277.92

Improvements:

4% sessionrestore linux64-shippable-qr opt e10s stylo 1,038.83 -> 1,001.33
3% tp5o_scroll linux64-shippable-qr opt e10s stylo 2.32 -> 2.24
3% tp5o_scroll linux64-shippable-qr opt e10s stylo 2.32 -> 2.25
3% startup_about_home_paint_realworld_webextensions linux64-shippable-qr opt e10s stylo 1,286.92 -> 1,249.33
3% tscrollx linux64-shippable-qr opt e10s stylo 1.23 -> 1.20
3% startup_about_home_paint linux64-shippable-qr opt e10s stylo 1,269.58 -> 1,235.08

You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=23347

On the page above you can see an alert for each affected platform as well as a link to a graph showing the history of scores for this test. There is also a link to a treeherder page showing the Talos jobs in a pushlog format.

To learn more about the regressing test(s), please see: https://wiki.mozilla.org/TestEngineering/Performance/Talos

For information on reproducing and debugging the regression, either on try or locally, see: https://wiki.mozilla.org/TestEngineering/Performance/Talos/Running

*** Please let us know your plans within 3 business days, or the offending patch(es) will be backed out! ***

Our wiki page outlines the common responses and expectations: https://wiki.mozilla.org/TestEngineering/Performance/Talos/RegressionBugsHandling

Comment 1

[Glenn Watson [:gw]]

This is likely to be a small cost due to the compilation of a new shader. Given that, and the improvements in other tests this is probably a WONTFIX. But I'll take a closer look next week (Monday is a public holiday here).

Comment 2

[Glenn Watson [:gw]]

I investigated this test (it starts up the browser 20 times) and confirmed it's caused by the extra compile time for the new composite shader.

We don't currently have shader compile caching on Linux, so this is expected for now. It also resulted in a number of improvements on other tests due to the faster composite shader when drawing real content.

So I think it's fine to close this as WONTFIX. Does that sound reasonable?

Comment 3

[Florin Strugariu [:Bebe]]

sound good to me.