1018561 - 2.2% linux32 dromaeo css regression on inbound (fx32) May 30

Status: RESOLVED WONTFIX

(Core :: DOM: Core & HTML, defect)

Description

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

We have a slight regression in dromaeo dom:
http://graphs.mozilla.org/graph.html#tests=[[72,131,33]]&sel=none&displayrange=7&datatype=running

I did some retriggers to help verify the exact changeset:
https://tbpl.mozilla.org/?tree=Mozilla-Inbound&fromchange=db25dbf0e8d0&tochange=e09be54f43d6&jobname=Ubuntu%20HW%2012.04%20mozilla-inbound%20talos%20dromaeojs

While this is a slight regression the patches from bug 957086 show a clear drop in the numbers (lower is worse) as prior to this the retriggers show a 4800-4900 range and for this changeset we dip into the 4700 range for most of the values.

Comment 1

[Boris Zbarsky [:bzbarsky]]

I really don't see why bug 957086 would have affected Dromaeo scores unless it's a PGO artifact.  But do we actually do PGO on this platform?

Comment 2

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

the test is regression on non pgo, we do pgo and it appears we have regressed there as well:
http://graphs.mozilla.org/graph.html#tests=[[72,131,33],[72,63,33]]&sel=1401107828173,1401712628173&displayrange=7&datatype=running

Comment 3

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

on aurora we seem to have an overall win (higher is better):
http://graphs.mozilla.org/graph.html#tests=[[72,63,33],[72,131,33],[72,52,33]]&sel=1397236726302,1405012726302&displayrange=90&datatype=running

so I will unblock this from firefox 32.