490163 - A script on this page has been stopped due to a low memory condition.

Status: VERIFIED DUPLICATE of bug 472097

(Core :: DOM: Core & HTML, defect)

Description

[myashwani]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1b3) Gecko/20090305 Firefox/3.1b3
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.1b3) Gecko/20090305 Firefox/3.1b3

when a few tabs are opened, the mozilla firefox 3.1 beta 3 hangs and the msg pops up saying "A script on this page has been stopped due to a low memory condition."

Reproducible: Always

Steps to Reproduce:
1.A script on this page has been stopped due to a low memory condition.
2.
3.
Actual Results:  
A script on this page has been stopped due to a low memory condition.

Comment 1

[Tyler Downer [He/Him]]

Thank you for this report. However, in order to make it more useful, please follow the steps found at http://quality.mozilla.org/bug-writing-guidelines and report back with your results.

Comment 2

[Wayne Mery (:wsmwk)]

without more info, and no response, this is 
=> incomplete

if you can add information, please do so

Comment 3

[Cristian Gómez Alvarez]

Hi, I've experienced this problem too on Firefox 3.5 on Windows XP 32 bits. I have 700 MB of RAM and a Pentium IV 2.4 GHz Processor.

All I did was open a few tabs and I got this error always.
For futher information I have uploaded this screenshots:

[1] Shows firefox bug and system monitor with memory usage of firefox
http://img33.yfrog.com/i/errorfirefox.jpg/
[2] Shows firefox bug
http://img16.yfrog.com/i/firefoxbug.jpg/
[3] Shows addons installed on firefox
http://img16.yfrog.com/i/complementosc.jpg/

I hope this information can help to solve this issue and we can still enjoy firefox as always.

Thanks

Comment 4

[Masatoshi Kimura [:emk]]

I've also experienced the problem. Fx 3.0 had worked perfectly on the same hardware. This is a regression from bug 453480, bug 453836, and bug 452485. IsLowMemory seems to report too premature "low memory" condition on Windows.
It reports even if pagefile have a room for extending.
I think we should disable checking low memory condition on Windows unless we could detect it more reliably because OOM killer is not present on Windows.

Comment 5

[Masatoshi Kimura [:emk]]

Confirming.

Comment 6

[Jared]

Confirmed on Windows 7 32bit

Comment 7

[Jared]

I read that the problem MAY have had to do with Firebug, but I disabled it and the error kept happening.

Comment 8

[Maark]

There are additional effects - browsing buttons and buttons on toolbars dissapear as well as pictures on the newly opened tabs. Also a tab icons dissapear sometimes.
The condition is probably a certain minimum number of tabs opened - currently I have 11 tabs opened and last three lost their icons.
No javascript running. Closing a number of tabs helps.

Comment 9

[Maark]

(In reply to comment #8)
> There are additional effects - browsing buttons [...]
You can click the places where browsing buttons (arrows) were.

Comment 10

[Jared]

Confirmed again on Windows XP 32bit. Haven't yet seen it on Ubuntu 9.04 32bit.

Comment 11

[Masatoshi Kimura [:emk]]

Workaround:
1. Go System Properties > Advanced tab > Performance options > Advanced tab > Vitrual Memory. (It may be slightlly different on Vista/7)
2. Select Custom Size
3. Make Initial size value larger than system default. Don't forget to press Set button.
Maximum size will not take effect. IsLowMemory seems to report "low memory" condition even if there is room for paging file growth.

Comment 12

[Jared]

That workaround seemed to help me on my Windows 7 boot, I have 1GB of ram and I set 1GB minimum size of paging file, 2GB max. I got the problem still after running a few more graphically intense games, but it did not happen as quickly.

Changed the minimum to 1.6GB, I'll see how that works.

Comment 13

[Christian Axelsson]

Confirmed on Vista 64 bit with 4GB ram. Swap file size is set to 4GB and system reports everything from 50% to 70% of memory usage when this occurs. 
The browser also tend to crash about 1-2 times a day, often when left idle during night.

Comment 14

[Tubusy]

Major problem, Vista 32; occurs in safe mode with everything disabled.

Comment 15

[Jared]

I have not had this problem AS OFTEN after I increased the size of my paging file to about 2x the amount of my physical RAM. This is not an ideal solution but it will console me until a fix is released (possibly with the next Firefox update?).

I /still/ get the problem if I'm running multiple memory-intense web pages, but this fix is an improvement from when it would occur just from viewing text-only webpages.

Comment 16

[Christian Axelsson]

If this is a problem with Windows not reporting free memory back as it should, then I think that the most sane option for now is to have a user option to disable the OOM detector.

Comment 17

[Jim Jeffery not reading bug-mail 1/2/11]

Even though this needs a reduced test case, or better STR, I'm requesting blocking due to the seemingly large number of complaints comming into the support forums.  Thread here: 

http://forums.mozillazine.org/viewtopic.php?f=38&t=1350495

Comment 18

[Steve Magruder]

This is happening to me just as described above on Windows Vista SP2.  I haven't the foggiest of how to produce a test case.  It just happens with normal use, but the funny thing is that it seems to be happening faster and faster after each browser restart.  I have Firebug and YSlow disabled.  And I've even cleared my cache a couple times today.

Comment 19

[Tubusy]

I do think this should be elevated to a Critical priority to be fixed in 3.5.1. I'm not certain how it works around here, but should this be filed under Firefox rather than Core? Somewhat alarmed it's going to be overlooked for a while. :|

Comment 20

[Tyler Downer [He/Him]]

https://bugzilla.mozilla.org/page.cgi?id=fields.html#importance
Not exactly critical. And no, this is a core component of Firefox so it goes in the core product. A ? is on blocking 1.9.1.1, so it may make it into 3.5.1.

Comment 21

[Steve Magruder]

I don't think it gets much more critical than this, IMHO.  This is the kind of problem that could drive users away from Firefox, and that would be most regrettable.

Anyway, the paging file size increase (hopefully temporary) seems to be working out so far.

Comment 22

[Jared]

It is definitely critical for me. I've downgraded Firefox on both of my computers to 3.0.11 to what worked until there is a bug fix. This bug made 3.5 completely unusable, which I am disappointed about.

Comment 23

[Tubusy]

This is what we know so far:

It's emerged in 3.5 (not certain if it's in betas)
It's not a problem with 'genuine' system memory limitations.
It's happening in XP and Vista.
It happens in Safe Mode, with no add-ons, clean caches and fresh profiles.
It corrupts both rendering and interface.
It can snowball into a full crash (reports submitted); or it can recover to a degree.
It can occur with just a few tabs open. 
It's period-of-time related; rare in a few minutes, very common after an hour.
It claims to be script related; and I seem to see a correlation with ajax-heavy sites. Could it be TraceMonkey?

That's all I can think of right now, any other clues?

Comment 24

[Jared]

I have some more details:

- It was in at least one of the release candidates, I found the error while I was testing it. I really thought it would go away by the time the official release came out.
- It's also happening in Windows 7
- It always recovers if I close many tabs -- especially the ones taking much of my memory [games and a music streaming site]
- I don't think it's script related, I think the error says it's stopping scripts to conserve memory, but that does not mean a script caused the problem.

Comment 25

[Jared]

Oh, and I can confirm that this does not occur in Ubuntu Jaunty 9.04

Comment 26

[Calisun]

I can confim this is happening on both in Win XP and Vista. The vista machine has 2GIG of memory and it is happening on regular basis.
In my experience this is happening when I have 3 or 4 tabs open. 
This is only a suspicion, but I believe that it has something to do with flash player. This error seems to happen more often when I am on sites that have a lot of flash, like myspace.

Comment 27

[Jared]

I'm disappointed that this is not fixed in Firefox 3.5.1.

Comment 28

[Mike Beltzner [:beltzner, not reading bugmail]]

Comment 23 doesn't contain any usable steps to reproduce. The error message is being thrown here:

http://mxr.mozilla.org/mozilla-central/source/dom/base/nsJSEnvironment.cpp#953

so --> DOM

jst: any idea why we might be throwing that error? is it possible that IsLowMemory is lying to us?

Not blocking for now, but please renominate if you have more information!

Comment 29

[dsmozilla]

(In reply to comment #20)
> https://bugzilla.mozilla.org/page.cgi?id=fields.html#importance
> Not exactly critical. And no, this is a core component of Firefox so it goes in
> the core product. A ? is on blocking 1.9.1.1, so it may make it into 3.5.1.

On my computer it effectively meets the "critical" definition. When I get this error, it usually starts popping up relentlessly on tab after tab over and over again, and even if I wanted to try to remedy it by closing tabs (despite having plenty of free memory!), half the time I can't actually click my way out of the modal error dialog boxes fast enough to start closing tabs out. My usual "workaround" to this problem is to pull up Task Manager and end-task Firefox, using session manager to restore my tabs.

Also, even when it's not so bad that you are stuck in modal error boxes, you generally can't choose to close the tabs you want, because ever 2-3 seconds a new error pops up, which not only initiates the modal dialog box but also forces focus to the tab and window of the page with the script that's stopped. This makes it practically difficult to close the tabs you suspect are eating up lots of memory (for example, Hulu or Facebook). For this reason, I usually wind up end-tasking Firefox even when I'm not literally trapped in modal dialog after modal dialog.

Seems to me that may as well be called a "crash." If the error was done similar to in Google Chrome where it pops up non-intrusive, non-modal error in the information bar saying, e.g. "the Flash plugin has crashed," THEN I would think it fair to call it not critical. As it is now I have to end-task Firefox roughly once every 30 minutes.

My system specs:

Vista
3GB RAM, *with windows paging file completely disabled"

Plugins: download statusbar, downloadhelper, firebug, logmein, personas, ramback, and session manager

Comment 31

[Jared]

Ok I'm not a developer, but from what I can see, the problems seem to be different. Most of the people who posted on this bug have had low RAM, under 1GB, and that problem has been reported for people who have large amounts of RAM over 4GB.

I may be wrong, I'm just an outsider. :l

Comment 32

[j.]

Occured to me on Windows XP SP3 machine with 1GB of RAM.
Firefox 3.5 worked fine for a few days. Then the problem gradually (yes) got worse and i was forced to install preview version. 
Also had to send a more then few crash reports. 

Happened in safe mode with the first page opened - google.com/ig

Comment 33

[Tubusy]

Agree it seems to be a duplicate, and resolved. Fix working in Minefield, looking forward to main Firefox update.

Comment 34

[Tyler Downer [He/Him]]

Well, this looks like it will make it into 3.5.2.

Comment 35

[Jared]

Confirmed to have been fixed in 3.5.2. I got my memory usage up to 850kb/1gb, and no error. Had 850kb/2gb memory if you include the paging file.