1577137 - 27.92% raptor-tp6m-microsoft-support-geckoview-cold loadtime (android-hw-p2-8-0-android-aarch64) regression on push 3a3ff2f2324cea34a7f2db4ecf865e6ffd8ebbc5 (Thu August 15 2019)

Status: RESOLVED INVALID

(Firefox for Android Graveyard :: Profile Handling, defect)

Description

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

Raptor has detected a Firefox performance regression from push:

https://hg.mozilla.org/releases/mozilla-beta/pushloghtml?changeset=3a3ff2f2324cea34a7f2db4ecf865e6ffd8ebbc5

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

Regressions:

28% raptor-tp6m-microsoft-support-geckoview-cold loadtime android-hw-p2-8-0-android-aarch64 pgo 1,749.64 -> 2,238.08

You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=22647

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 Raptor jobs in a pushlog format.

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

*** 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

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

Not 100% sure. Did some retriggers to confirm.

Comment 2

[Nick Alexander :nalexander [he/him]]

This makes little sense. The offending commit

151bfae582afadf93c61e55e1fd4f980a716756f	Nick Alexander — Bug 1570690 - Synchronously extract system addons before Gecko startup in Fennec. r=VladBaicu,Grisha, a=RyanVM

will absolutely have regressed Fennec performance. It's a trade off, correctness for... incorrectness and speed. But this code doesn't impact GV. And if it did impact GV, every single coldload test should have been impacted, since it would impact startup uniformly (without regard to site under test).

I think this is just noise, and will leave it to somebody more familiar with the process to resolve it.

Comment 3

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

You're right, the retriggers reveald that not this commit caused the regression. Sorry!