1151889 - 5-34% Windows cart regression on Mozilla-Inbound-Non-PGO (v. 40) on April 06, 2015 from push 4450f65d2ba9

Status: RESOLVED FIXED

(Testing :: Talos, defect)

Description

[Vaibhav (:vaibhav1994)]

Talos has detected a Firefox performance regression from your commit 4450f65d2ba9 in bug 1151306.  We need you to address this regression.

This is a list of all known regressions and improvements related to your bug:
http://alertmanager.allizom.org:8080/alerts.html?rev=4450f65d2ba9&showAll=1

On the page above you can see Talos 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, please see: https://wiki.mozilla.org/Buildbot/Talos/Tests#TART.2FCART

Reproducing and debugging the regression:
If you would like to re-run this Talos test on a potential fix, use try with the following syntax:
try: -b o -p win64,win32 -u none -t svgr  # add "mozharness: --spsProfile" to generate profile data

To run the test locally and do a more in-depth investigation, first set up a local Talos environment:
https://wiki.mozilla.org/Buildbot/Talos/Running#Running_locally_-_Source_Code

Then run the following command from the directory where you set up Talos:
talos --develop -e <path>/firefox -a cart

Making a decision:
As the patch author we need your feedback to help us handle this regression.

Our wiki page oulines the common responses and expectations:
https://wiki.mozilla.org/Buildbot/Talos/RegressionBugsHandling

Comment 1

[Vaibhav (:vaibhav1994)]

Markus Stange, could you look what might be causing this regression?

Comment 2

[Markus Stange [:mstange]]

I started three profiling try runs:
baseline: https://treeherder.mozilla.org/#/jobs?repo=try&revision=92e0bd8b1eb7
+ bug 1151306: https://treeherder.mozilla.org/#/jobs?repo=try&revision=e12732a982a3
+ bug 1151346: https://treeherder.mozilla.org/#/jobs?repo=try&revision=8859057cdeab

Comment 3

[Markus Stange [:mstange]]

This is a regression from bug 1151346. We must be layerizing too aggressively; there's a lot more time spent in nsNativeThemeWin::ClassicDrawWidgetBackground in the XP profile (882 samples vs 71 samples).

Looks like I need to run CART on my Windows machine.

Comment 4

[Markus Stange [:mstange]]

I think I'll just back out bug 1151346 again. I have no explanation for why it causes additional repaints during the customization animation (which animates margin-left and margin-right), and it's going to be a lot of work to make the animation run smoothly, so I'm going to defer this until I can spend more time on it.

Comment 5

[Vaibhav (:vaibhav1994)]

A couple of more regressions seen on cart test for windows on mozilla-inbound: http://alertmanager.allizom.org:8080/alerts.html?rev=4450f65d2ba9&showAll=1&testIndex=0&platIndex=0

Comment 6

[Joel Maher ( :jmaher ) (UTC -8)]

should we close this bug, or should we make it dependent on bug 1009693?

Comment 7

[Markus Stange [:mstange]]

Oops, I forgot to update this bug.
Bug 1151346 has been backed out. Is this regression fixed?

Comment 8

[Joel Maher ( :jmaher ) (UTC -8)]

yes, this is resolved!