1590045 - 2.3 - 4.2% raptor-tp6-linkedin-firefox / raptor-tp6-linkedin-firefox fcp (linux64-shippable, linux64-shippable-qr, windows7-32-shippable) regression on push 8b6f7d7158287a1175178e33deffdb18f0c5d2db (Mon October 7 2019)

Status: RESOLVED DUPLICATE of bug 1587638

(Core :: JavaScript Engine, defect)

Description

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

Raptor has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=8b6f7d7158287a1175178e33deffdb18f0c5d2db

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

Regressions:

4% raptor-tp6-linkedin-firefox fcp windows7-32-shippable opt 423.54 -> 441.33
4% raptor-tp6-linkedin-firefox windows7-32-shippable opt 533.99 -> 555.24
4% raptor-tp6-linkedin-firefox windows7-32-shippable opt 533.91 -> 554.87
4% raptor-tp6-linkedin-firefox fcp linux64-shippable opt 415.44 -> 430.75
3% raptor-tp6-linkedin-firefox fcp linux64-shippable-qr opt 441.98 -> 454.58
2% raptor-tp6-linkedin-firefox linux64-shippable-qr opt 568.28 -> 581.34

You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=23515

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

[Iain Ireland [:iain]]

The regression was the result of backing out a patch that had some GC-related bugs. An updated version of the patch is attached to bug 1587638. It will be landed after merge day, to give it more time to bake in Nightly. We should get the performance improvements back once that patch lands.

Comment 2

[Nicolas B. Pierron [:nbp]]

This regression is strange, to my knowledge none of our tp6 benchmarks were testing XDR serialization & deserialization.
At least, this was the case when the JavaScript Start-up Bytecode Cache landed, as no regressions nor improvement got reported.