<Regression> Mozilla-Aurora - Tp5 No Network Row Major MozAfterPaint (Main Startup File IO Bytes) - WINNT 6.1 (ix) - 6.49%

[nob...@cruncher.build.mozilla.org]

Regression: Mozilla-Aurora - Tp5 No Network Row Major MozAfterPaint (Main Startup File IO Bytes) - WINNT 6.1 (ix) - 6.49% increase
----------------------------------------------------------------------------------------------------------------------------------
Previous: avg 45299525.000 stddev 631096.473 of 12 runs up to revision d480e05bd276
New : avg 48240683.333 stddev 1398247.120 of 12 runs since revision 68399cfe3799
Change : +2941158.333 (6.49% / z=4.660)
Graph : http://mzl.la/1Jzzy9o

Changeset range: http://hg.mozilla.org/releases/mozilla-aurora/pushloghtml?fromchange=d480e05bd276&tochange=68399cfe3799

Changesets:
* http://hg.mozilla.org/releases/mozilla-aurora/rev/68399cfe3799
: Kyle Zentner <kzen...@mozilla.com> - Bug 1178168 - Unremove box-sizing: padding-box in Gecko Tests. r=dholbert a=test-only
: http://bugzilla.mozilla.org/show_bug.cgi?id=1178168

Bugs:
* http://bugzilla.mozilla.org/show_bug.cgi?id=1178168 - Bring back 'box-sizing: padding-box' tests on Aurora, as long as we're still shipping support for 'box-sizing: padding-box'

[dhol...@mozilla.com]

This regression was "caused" by a reftest-only change (so, NPOTB) which couldn't possibly have impacted Tp results.

I'm guessing this was just caused by the uplift (so, it regressed [maybe gradually] while this code was on mozilla-central over the past 6 weeks).

The regression does seem pretty real -- looking at the graph, it's mostly stuck at a new, higher value. There is one point that dropped back down to the earlier baseline, though (which makes it believable that the regression really happened before Kyle's test-only patch, and the samples before his were just sporadically low in a post-regression world).

[nob...@cruncher.build.mozilla.org]

Regression: Mozilla-Aurora - SVG, Opacity Row Major - Ubuntu HW 12.04 - 7.71% increase
--------------------------------------------------------------------------------------
Previous: avg 281.669 stddev 4.652 of 12 runs up to revision d480e05bd276
New : avg 303.377 stddev 7.274 of 12 runs since revision 68399cfe3799
Change : +21.708 (7.71% / z=4.666)
Graph : http://mzl.la/1Nzwzwj