1443806 - stylo: 2.57 - 2.96% about_preferences_basic (linux64) regression on push a50ae6e8b614d0c1c1cc0196b691f7732def60f4 (Wed Mar 7 2018)

Status: RESOLVED WONTFIX

(Core :: CSS Parsing and Computation, defect, P2)

Description

[Ionuț Goldan [:igoldan]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=a50ae6e8b614d0c1c1cc0196b691f7732def60f4

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

Regressions:

  3%  about_preferences_basic linux64 opt e10s stylo     143.45 -> 147.70
  3%  about_preferences_basic linux64 pgo e10s stylo     130.70 -> 134.06


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=11951

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/Buildbot/Talos/Tests

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

Comment 1

[Ionuț Goldan [:igoldan]]

:xidorn This is a strange regression. It happens only on Linux *and* only on the initial landing branch, which is autoland.

Though there was a following merge with mozilla-inbound, this regression didn't reproduce. Or at least not yet.

What thoughts do you have on the validity of this perf alert?

Comment 2

[Bobby Holley (:bholley)]

Four milliseconds, linux-only, and autoland-only make me not too concerned about this particular regression. However, we expect this change to be a significant win on a number of talos tests, per bug 1375913 comment 10. Do you see such results?

Comment 3

[Xidorn Quan [:xidorn] UTC+10]

This is the same regression on m-i, isn't it? https://treeherder.mozilla.org/perf.html#/alerts?id=11969

I cannot really see how it can regress that, though...

Comment 4

[Ionuț Goldan [:igoldan]]

(In reply to Ionuț Goldan [:igoldan], Performance Sheriffing from comment #1)
> Though there was a following merge with mozilla-inbound, this regression
> didn't reproduce. Or at least not yet.

OK, it reproduced on mozilla-inbound as well.

Comment 5

[Ionuț Goldan [:igoldan]]

(In reply to Bobby Holley (:bholley) from comment #2)
> However, we expect this change to be a
> significant win on a number of talos tests, per bug 1375913 comment 10. Do
> you see such results?

Yes, I see sessionrestore wins:

== Change summary for alert #11951 (as of Wed, 07 Mar 2018 00:18:48 GMT) ==

Improvements:

  4%  sessionrestore_no_auto_restore windows7-32 opt e10s stylo     589.88 -> 567.50
  4%  sessionrestore windows7-32 opt e10s stylo                     506.42 -> 487.42
  3%  sessionrestore_many_windows windows10-64 pgo e10s stylo       1,367.54 -> 1,323.00
  2%  sessionrestore_many_windows windows10-64 opt e10s stylo       1,602.71 -> 1,565.33

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=11951

Comment 6

[Bobby Holley (:bholley)]

So we've got a 4ms linux-only regression on about:preferences, compared to a ~20ms improvement on window-opening on windows (tpaint is also affected by my measurements, curious why it didn't show up here). I'm comfortable wontfixing the linux issue.