1578386 - 2.72 - 11% raptor-wasm-misc-baseline-firefox / raptor-wasm-misc-firefox / raptor-wasm-misc-ion-firefox (linux64-shippable, linux64-shippable-qr) regression on push eee3b548e7cd4757d7611ec319aabcea8bf3c2df (Sat August 31 2019)

Status: RESOLVED FIXED

(Core :: JavaScript: WebAssembly, defect)

Description

[Florin Strugariu [:Bebe]]

Raptor has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/autoland/pushloghtml?changeset=eee3b548e7cd4757d7611ec319aabcea8bf3c2df

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

Regressions:

11% raptor-wasm-misc-ion-firefox linux64-shippable opt 39,595.71 -> 43,952.00
11% raptor-wasm-misc-firefox linux64-shippable-qr opt 39,709.53 -> 43,974.60
11% raptor-wasm-misc-firefox linux64-shippable opt 39,729.51 -> 43,977.49
3% raptor-wasm-misc-baseline-firefox linux64-shippable-qr opt 69,790.92 -> 71,689.90

You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=22918

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

[Lars T Hansen [:lth]]

Interesting, could indicate that the test machine gets the explicit bounds checking path?

Comment 2

[Ryan Hunt [:rhunt] (on leave until early May)]

I think the issue resolved by bug 1578031 is a more likely candidate. It would cause us to fold all offsets > 1, which seems like it would have a performance impact.

Comment 3

[Ryan Hunt [:rhunt] (on leave until early May)]

Oh, and bug 1507765 (the work to actually dynamically turn off huge memory with low VM limits) hasn't landed yet. So we definitely shouldn't unintentionally be getting explicit bounds checks on these machines.

Comment 4

[Florin Strugariu [:Bebe]]

== Change summary for alert #22942 (as of Tue, 03 Sep 2019 12:56:41 GMT) ==

Improvements:

10% raptor-wasm-misc-firefox linux64-shippable opt 44,064.13 -> 39,739.46
10% raptor-wasm-misc-ion-firefox linux64-shippable opt 43,917.87 -> 39,609.96
10% raptor-wasm-misc-ion-firefox linux64-shippable-qr opt 43,930.73 -> 39,677.87
9% raptor-wasm-misc-firefox linux64-shippable-qr opt 43,835.86 -> 39,999.05
2% raptor-wasm-misc-baseline-firefox linux64-shippable opt 71,504.26 -> 69,756.17
2% raptor-wasm-misc-baseline-firefox linux64-shippable-qr opt 71,604.80 -> 69,951.33

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=22942

Comment 5

[Liz Henry (:lizzard) (relman/hg->git project)]

What actually fixed this?

Comment 6

[Ryan Hunt [:rhunt] (on leave until early May)]

Bug 1578031 actually fixed this issue.