146983 - Mozilla Consumes 100% CPU Viewing Some Pages

Status: VERIFIED DUPLICATE of bug 123273

(Core :: JavaScript Engine, defect)

Description

[nyb0dy]

From Bugzilla Helper:
User-Agent: Mozilla/5.0 Galeon/1.2.1 (X11; Linux i686; U;) Gecko/20020417
BuildID:    1.0rc3

I've tried on 3 different OSes (rh 7.2, win xp, win 98) and the results were the
same: mozilla (1.0rc3) eats up a lot of cpu power when requested to display that
page.

I think that the problem is caused by the javascript engine: by disabling
javascript the cpu time is perfectly normal.

Reproducible: Always
Steps to Reproduce:
1.Load the page
2.Wait for mozilla to load the uppermost banner


Actual Results:  Mozilla takes up (almost) all the cpu time

Expected Results:  It should have used less cpu power

First of all the page is marked "optimized for IE v5.5+", which is not a plus.
I think that the guy use javascript to change the banner, and (maybe for a flaw
in his code) the javascript engine takes all the cpu. 

Maybe it would be useful to put a limit (chosen by the final user or the
developer, I don't know) on the javascript cpu power (i.e. a limit of
instructions per second);

Comment 1

[Mathias Bogaert]

Confirming. Build 2002052408 Win2K. Works fine with IE 6.0.

Comment 2

[Moiz Dohadwala]

Reproducible on win2k sp2, m1rc3 ( 2002052306 ).
I think it the marquee code for the status line at the bottom, the one which
says "Bienvenido a la pagina de JL-Picard".

Comment 3

[Boris Zbarsky [:bzbarsky]]

This is almost certainly not a JS engine bug but rather a DOM or layout bug.

Mathias, are you using a branch or trunk build?

Could someone run a Venkman or jprof profile of the page in question?

Comment 4

[joakim_46]

WFM with build 2002052306 under Windows ME. I can't see any big CPU 
consumption. The page renders very quickly indeed.

Comment 5

[Jerry Baker]

The included URL pegs my CPU with RC3. I came to Bugzilla to report similar
trouble with a different page. Changing summary and marking as NEW.

Comment 6

[Jerry Baker]

It is the status bar JS. I will attach a reduced test case.

Comment 7

[Jerry Baker]

Attachment: HTML with offending JS

Comment 8

[Boris Zbarsky [:bzbarsky]]

Jerry, thank you for the testcase.  This is partly bug 123273 ("setTimeout
(something, 0) causes 100% CPU constant") and partly bug 135305 ("Should 
throttle rate of status string updates for performance").

The former is likely more relevant here, so marking duplicate of that.

*** This bug has been marked as a duplicate of 123273 ***

Comment 9

[Jerry Baker]

Mass removing self from CC list.

Comment 10

[Jerry Baker]

Now I feel sumb because I have to add back. Sorry for the spam.

Comment 11

[Phil Schwartau]

Verified Duplicate. Those interested in this issue should be cc themselves 
on bug 123273, "setTimeout(something, 0) causes 100% CPU constant"