1572674 - 2.29% ts_paint (windows10-64-shippable-qr) regression on push 9d66a096d9e2872c33798b7633448b9144298885 (Sat August 3 2019)

Status: RESOLVED DUPLICATE of bug 1572682

(Firefox Build System :: Toolchains, defect)

Description

[Marian Raiciof [:marauder]]

Talos has detected a Firefox performance regression from push:

https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?changeset=9d66a096d9e2872c33798b7633448b9144298885

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

Regressions:

2% ts_paint windows10-64-shippable-qr opt e10s stylo 315.46 -> 322.67

You can find links to graphs and comparison views for each of the above tests at: https://treeherder.mozilla.org/perf.html#/alerts?id=22295

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/TestEngineering/Performance/Talos

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

Comment 1

[Mike Hommey [:glandium]]

It's very unlikely that this is the right regressor. Practically speaking, no change happened. Especially not within Firefox.

Comment 2

[Marian Raiciof [:marauder]]

Thank you for the infos Mike!

Comment 3

[Mike Hommey [:glandium]]

The regression does look real though. Only misattributed. Why close the bug?

Comment 4

[Marian Raiciof [:marauder]]

I have reopened the bug.

According to the pushlog : https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=37229cef2cc79d44470afc9e04016bac8ddd0ae8&tochange=9d66a096d9e2872c33798b7633448b9144298885

if bug 1570598 is not related to this alert, then we remain with two variants:
https://bugzilla.mozilla.org/show_bug.cgi?id=1570240
https://bugzilla.mozilla.org/show_bug.cgi?id=1570798

Can you give some insights about those two bugs ?

Comment 5

[Mike Hommey [:glandium]]

I don't see how any of those can be related.

Comment 6

[Mike Hommey [:glandium]]

From the graph, I'd say the regression comes from somewhere before that, on autoland. Hard to say when exactly.

Comment 7

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Bebe: Could you help Marian to identify the correct culprit here?

Comment 8

[Ionuț Goldan [:igoldan]]

(In reply to Mike Hommey [:glandium] from comment #6)

From the graph, I'd say the regression comes from somewhere before that, on autoland. Hard to say when exactly.

You are right about this. mozilla-inbound is just a downstream.

Comment 9

[Ionuț Goldan [:igoldan]]

The real regression seems to have appeared on autoland around August 2. Skimming over the talos perf regressions which appeared around that time, I've come across bug 1572682. That bug has 2 different regressions of the same magnitudes; it landed a fix around August 12, which canceled this regression also.

Comment 11

[Marco Castelluccio [:marco]]

Was bug 1570598 regressed by bug 1572682 then? Could you update the regressed by field of bug 1570598?