Closed Bug 1117106 Opened 9 years ago Closed 9 years ago

Sunspider 3d cube 16% regression

Categories

(Core :: JavaScript Engine, defect)

Product:
Core
Component:
JavaScript Engine
Version:
37 Branch
Platform:
x86_64
Windows 7
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED INVALID

People

(Reporter: alice0775, Unassigned)

References

Details

(Keywords: regression, Whiteboard: [Dupe me]) 

Steps to reproduce:
1 run Sunspider https://www.webkit.org/perf/sunspider/sunspider.html

Regression window:
Good: 26.1ms +/- 4.7%
https://hg.mozilla.org/integration/mozilla-inbound/rev/6f1696df360b
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:37.0) Gecko/20100101 Firefox/37.0 ID:20141212114057

Bad: 30.4ms +/- 2.5%
https://hg.mozilla.org/integration/mozilla-inbound/rev/7837628feeca
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:37.0) Gecko/20100101 Firefox/37.0 ID:20141212115254

Pushlog:
https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=6f1696df360b&tochange=7837628feeca

Triggered by: Bug 1090636
[Tracking Requested - why for this release]:
tracking-firefox37: --- → ?
In local build
Last Good: 2676562ab2b3
First Bad: 44c3c726c7ae

Triggered by:
	44c3c726c7ae	Jason Orendorff — Bug 1090636, part 3 - Shadow an existing property even if pobj == obj. r=efaust.
I see in 38.0a1 (2015-01-12) Win 7 x64:
cube: 20.1ms +/- 11.2%
(In reply to Paul Silaghi, QA [:pauly] from comment #3)
> I see in 38.0a1 (2015-01-12) Win 7 x64:
> cube: 20.1ms +/- 11.2%

The value is no meaning other than the fact that your PC is fast than mine.
You're right.
In Nightly 2014-12-14 - cube: 17.1ms +/- 9.9%
Status: UNCONFIRMED → NEW
Ever confirmed: true
Boy am I glad I split that patch into tiny chunks. The blamed changeset is a one-line change. Looking at it now.
Failed to reproduce this in the shell. Building a browser to try to repro there.
Can't reproduce in the browser either.

AWFY shows very little change in this test over time, and no difference around January 1 which is when my patches landed.

http://arewefastyet.com/#machine=28&view=single&suite=ss&subtest=cube
(In reply to Jason Orendorff [:jorendorff] from comment #8)
> Can't reproduce in the browser either.
> 
> AWFY shows very little change in this test over time, and no difference
> around January 1 which is when my patches landed.
> 
> http://arewefastyet.com/#machine=28&view=single&suite=ss&subtest=cube

okay, small enough
Status: NEW → RESOLVED
Closed: 9 years ago
status-firefox36: unaffected → ---
status-firefox37: affected → ---
tracking-firefox37: + → ---
Resolution: --- → INVALID
You need to log in before you can comment on or make changes to this bug.