776233 - [meta] sunspider regression in AWFY

Status: RESOLVED INVALID

(Core :: JavaScript Engine, defect)

Description

[Olli Pettay [:smaug][[e-mail]]]

there is some major regression in http://arewefastyet.com/

Comment 1

[Boris Zbarsky [:bzbarsky]]

Is this the string-fasta thing?

Comment 2

[Boris Zbarsky [:bzbarsky]]

http://arewefastyet.com/?a=b&view=breakdown says "dna", "format-tofte", "tagcloud", "unpack-code" all took a hit around the 20th.

Comment 3

[Jan de Mooij [:jandem]]

See bug 775915 and bug 776200.

Comment 4

[David Mandelin [:dmandelin]]

Tracking even though it's a meta bug because we don't necessarily know exactly the cause of all the regressions.

Comment 5

[David Mandelin [:dmandelin]]

OK, I tracked down all the regressions. All but 3ms are fixed. The last 3ms is 777174. Keeping this open until it's all better but we'll just track bug 777174.

Comment 6

[David Mandelin [:dmandelin]]

I think this has turned into a meta bug that doesn't need to be tracked in itself.

Comment 7

[Paul Wright]

So this bug can be closed now, right?

Comment 8

[Andrew McCreight [:mccr8]]

Depends, does AWFY show that the regression is fixed? ;)

Comment 9

[Paul Wright]

You're right.  On July 19, JM on 64-bit (which is the one I care about, fwiw) was at 319-320.  As of today, we are at 327.  It's roughly the same on 32-bit.  So, cannot consider this fixed.

Comment 10

[Guilherme Lima]

Due to Ion landing, JM removal and BC landing, I think it's invalid now.