1444361 - 216.36 - 534.99% tp5o_webext responsiveness (windows10-64, windows7-32) regression on push 22ba4f32a7af (Thu Mar 8 2018)

Status: RESOLVED FIXED

(Toolkit :: Crash Reporting, defect)

Description

[Ionuț Goldan [:igoldan]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=7daaf289c085e05e0e4a0327a13748002fb467a0&tochange=22ba4f32a7afe64038a38cd2e1fc16ede661f9ab

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

Regressions:

535%  tp5o_webext responsiveness windows7-32 opt e10s stylo     3.48 -> 22.08
216%  tp5o_webext responsiveness windows10-64 opt e10s stylo    3.17 -> 10.04

Improvements:

  2%  tp5o_webext osx-10-10 opt e10s stylo     397.37 -> 388.89


You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=12015

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]]

:aklotz I see bug 1434489 caused some big regressions on our Windows OPT builds. The PGO builds don't seem affected.

Can you please look over this performance drop?

Comment 2

[(No longer employed by Mozilla) Aaron Klotz]

I'm kind of surprised that Talos caught this given that the modules ping is so infrequent. Anyway, I see a problem that would affect perf, so I will fix that.

Comment 3

[(No longer employed by Mozilla) Aaron Klotz]

Attachment: Move gathering of cert info off main thread

Comment 4

[Chris H-C :chutten]

Comment on attachment 8957664 [details] [diff] [review]
Move gathering of cert info off main thread

Review of attachment 8957664 [details] [diff] [review]:
-----------------------------------------------------------------

Looks good. One question about maybe using a utility function.

::: toolkit/components/telemetry/Telemetry.cpp
@@ +908,5 @@
>  #if defined(XP_WIN)
>        // Cert Subject.
> +      nsString* subject = mCertSubjects.GetValue(info.GetModulePath());
> +      if (subject) {
> +        JS::RootedString jsOrg(cx, JS_NewUCStringCopyN(cx, subject->get(),

Can you use froydnj's new ToJSString (from TelemetryCommon) here?

Comment 5

[(No longer employed by Mozilla) Aaron Klotz]

https://hg.mozilla.org/integration/mozilla-inbound/rev/017dbfcb23698b9484a4467221e86a0dca980326
Bug 1444361: Don't obtain cert subject info on main thread; r=chutten

Comment 6

[Noemi Erli[:noemi_erli]]

https://hg.mozilla.org/mozilla-central/rev/017dbfcb2369

Comment 7

[Ionuț Goldan [:igoldan]]

I confirm this got fixed:

== Change summary for alert #12047 (as of Fri, 09 Mar 2018 18:25:13 GMT) ==

Improvements:

 74%  tp5o_webext responsiveness windows7-32 opt e10s stylo     12.98 -> 3.33
 67%  tp5o_webext responsiveness windows10-64 opt e10s stylo    9.65 -> 3.19

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=12047