361703 - Page pegs cpu for 30+ sec

Status: RESOLVED WORKSFORME

(Core :: CSS Parsing and Computation, defect)

Description

[Mike Pinkerton (not reading bugmail)]

Go to the above page, it pegs the CPU with the majority of the time (sample forthcoming) spent in CSS parsing. 

It pegs Safari as well, but only for about 10 seconds. Gecko definitely has more problems with it.

Comment 1

[Mike Pinkerton (not reading bugmail)]

Attachment: 10sec profile from sampler

Comment 2

[Mike Pinkerton (not reading bugmail)]

tested with camino 1.1a1 (2006101801) which is mozilla1.8.1. I'd imagine the trunk has the same problems, and I'd also imagine it's XP, but I haven't tried it.

Comment 3

[Mats Palmgren (inactive)]

WFM, Firefox 2.0.0.4 and trunk on MacOSX 10.4.9 and Linux (all 4).

Mike, do you still see this problem?

Comment 4

[Wayne Mery (:wsmwk)]

FWIW, 20+sec, but pegs CPU only very briefly Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9a8pre) Gecko/2007081905 Minefield/3.0a8pre  pentium 4 3.2G

Comment 5

[Wayne Mery (:wsmwk)]

ditto SM trunk, ~20 sec
Build identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a8pre) Gecko/2007081503 SeaMonkey/2.0a1pre

But, virtually at the same time ~8 seconds FF2 on an 8yr-old, small porker AMD, again, high cpu only a couple sec. 
Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8.1.3) Gecko/20070309 Firefox/2.0.0.3

So, there may be some variables in play?

Comment 6

[Wayne Mery (:wsmwk)]

(In reply to Wayne Mery (:wsmwk) from comment #4)
> FWIW, 20+sec, but pegs CPU only very briefly Mozilla/5.0 (Windows; U;
> Windows NT 6.0; en-US; rv:1.9a8pre) Gecko/2007081905 Minefield/3.0a8pre 
> pentium 4 3.2G

I think that was my old vista desktop.

current desktop is win7, i7 2.8Ghz, fast ethernet, nightly build
< 2 seconds, and CPU usage is a blip

HOwever my 6yr old vista laptop, on home wireless and cablemodem, using nightly is 12 seconds, but CPU is not bad

So I think we can call this WFM