Closed Bug 277547 Opened 20 years ago Closed 13 years ago

Firefox freezes or hangs frequently and temporarily at zero or low cpu - HT / hyperthread / multiple processor only - NOT resolved by BIOS update

Categories

(Firefox :: General, defect)

Product:
Firefox
Component:
General
Version:
3.0 Branch
Platform:
x86
Windows XP
Type:
defect
Priority:
Not set
Severity:
major

Tracking

()

Status:
RESOLVED INCOMPLETE

People

(Reporter: luto, Unassigned)

References

Details

(Keywords: perf, Whiteboard: workaround comment 31, do not dup to Bug 282392!)

Attachments

(1 file)

Snapshot of Resource Explorer showing FireFox near 100%
55.57 KB, application/x-zip-compressed
Details 
User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0

Firefox very frequently freezes (windows detects it as not responding, the mouse
shows an hourglass, and input has no effect).  It will usually unfreeze after a
few seconds or minutes, but it will eventually freeze again.  This sometimes
happens immediately after starting Firefox and sometimes just when interacting
with the browser.  CPU usage is 0 when this happens.

This happens on my computer, after replacing CPU and motherboard, removing a bad
hard drive, reformatting, then reinstalling just XP SP2 (slipstreamed), the
network driver, and Firefox.  I've also seen it on two network adapters.

Now for the strange part: if I run tcpview (www.sysinternals.com) and close the
TCP connection that Firefox opens to itself, firefox starts taking 100% CPU, but
it also functions flawlessly, will unhang itself, and will not hang again.  I
suspect some kind of deadlock or missed wakeup, and when I kill the connection,
all the select()s or whatever Firefox uses return immediately, and the problem
is hidden.

Is there anything I can do to provide more information?

Reproducible: Sometimes

Steps to Reproduce:
1. Run Firefox
2. Do anything.  www.slashdot.org usually does it, but so do most pages.

Actual Results:  
Firefox freezes for awhile.

Expected Results:  
It shouldn't freeze.

I have two monitors at different resolutions.  That's the only odd thing about
my setup I can think of.
(In reply to comment #0)
> User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5)
Gecko/20041107 Firefox/1.0
> Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5)
Gecko/20041107 Firefox/1.0
> 
> Firefox very frequently freezes (windows detects it as not responding, the mouse
> shows an hourglass, and input has no effect).  It will usually unfreeze after a
> few seconds or minutes, but it will eventually freeze again.  This sometimes
> happens immediately after starting Firefox and sometimes just when interacting
> with the browser.  CPU usage is 0 when this happens.

I have the same problem on Mac OS X with Firefox, Mozilla and Camino.
Reproducable with www.n-tv.de.

Bojan
I can confirm this for
"Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107
Firefox/1.0"
too, and for 
"Mozilla 1.7.5
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041217"
and Mozilla 1.8a4.

Freeze happens quite often and may occur with blank start page and moving mouse
or right click. No URL visit needed.

Bojans desc probably belongs to a different issue, since this is independent of
URLs and not reproducible with a specific one.

I deinstalled Mozilla 1.8a4, installed 1.7.5 instead, deinstalled Firefox 1.0,
moved away personal configs to have a clean start, reinstalled, no change.

More similarities to Andys desc.: The day before this started, I changed cpu and
dvd drive. Nothing else and no change in installed software.

The temporary freeze seems to occure less often if i'm running systernal procmon
along. This may hint to a thread deadlock.

Can by now confirm Andys workaround, after closing the local loopback connection
in tcpview, firefox immediately "un-freezes" and seems not to freeze again.

I'm joining Andys on his question: Is there anything I can do to provide more
information?
I can confirm this bug as well. 

I am running: 
- XP SP2 with a dual head setup 
- 3.0Ghz w/ HT
- 1GB Ram
- Asus P4S800 Mobo (Sis 648 chipset)
- Onboard LAN
- Matrox P650 (AGP 8x) 

The workaround with tcpview seems to work. This lockup seems to affect all parts
of the browser. It will hang in the preferences as well as just normal.

This also occurs with Mozilla (as stated previously). Therefore I'm thinking it
is a more general gecko issue than a Firefox issue.

If there is any more information I can provide, please let me know.

-Ted
Theodore:

That's interesting.  I also have a P4 3.0 HT and dual monitors.  I'll try it
with HT disabled tomorrow.
(In reply to comment #4)
> Theodore:
> That's interesting.  I also have a P4 3.0 HT and dual monitors.  I'll try it
> with HT disabled tomorrow.

I also have a P4 3.0 HT, a brand new system using an Intel D865PERL mb (865PE 
chipset).  My system was exhibiting the exact same symptoms, and I thought I 
had some kind of a driver issue because I've never seen this anywhere else with 
Firefox.  I tried updating the BIOS and all hardware/INF drivers to the latest 
version from Intel's website, and I also tried uninstalling/reinstalling 
Firefox, and using Netscape 7.2 instead of Firefox, with the same effects (or 
lack thereof).

More importantly, disabling hyperthreading in the BIOS as Theodore mentioned 
above DID solve the problem.  With HT disabled, I could not get the browser to 
hang anywhere.

Whether this bug has to do with hyperthreading, or just this particular 
processor (I have a P4 2.8GHz w/HT that I have never seen this issue with), 
this is most definitely reproduceable.  Disabling HT in the BIOS seems to be a 
good workaround.
This confirms that it is more than likely a threading issue within Firfox/gecko.
I'd like to actually clear up some information that it was Andy who originally
mentioned disabling hyperthreading. I'd also like to point out that while
disabling hyperthreading is a fine workaround, there are adverse speed
side-effects to doing such. The tcpview workaround is a valid one as well,
although it is a little more annoying if you close your browser a lot. It
doesn't lend itself to slowing down other programs along with it though.

I have additionally noticed that there is the 'Quality Feedback Agent' which has
no window but likes to hang around in the taskbar after manually disconnecting
the localhost connections. This leads to the conclusion that the localhost ports
are used to talk to the Quality Feedback Agent. This would make sense for
cross-platform IPC. This also gives a better look as to where exactly the
threading problem is occuring. (Where exactly is used very broadly in this
context ;) )

I hope this issue can be resolved quickly, although I know from experience that
threading deadlocks are horribly difficult to track dwon.

-Ted
First, the tcpview workaround sucks, as it causes Firefox to spin, using 100% of
one virtual CPU.

But here's a better one: Start Firefox, then, before it freezes, use Task
Manager to set its affinity to just one CPU.  Seems to work for me (it hasn't
frozen in 10 minutes -- I usually get about 10 seconds before a freeze).

Note that I see this problem in every Firefox, Mozilla, Thunderbird (rarely),
and K-Meleon.  I haven't seen it in Sunbird yet.

I'm going to guess that somehow the P4 3.0 HT can trigger a race almost every
time.  When affine to one processor, though, the race is hidden again.
-- workaround

Talking 'bout workarounds, here is mine, which eliminates the need for doing
something every time firefox is started:

I'm using a personal firewall and added a filter rule denying firefox access to
127.0.0.1 (localhost). This additional rule must have a higher precedence than
rules allowing firefox generel access or than rules allowing generel access to
127.0.0.1.

So far I have not observed any negative side effects. By the way, what is the
use of this localhost connection in firefox (gecko)?


-- question

Since this seems to happen on some machines only and maybe only if some hardware
was changed after firefox was installed for the first time, I still wonder if
some other configuration data could be deleted to avoid the bug. What is stored
in the registry? Is there any hardware dependent initialization?

Martin

PS: pls vote for this bug, to make developers aware of this issue
Same issue here, same CPU, P4 3.0 HT.  Been using Firefox on my old machine 
forever, upgraded machines, and I get this problem.  Regardless of what I'm 
doing, within about 30 seconds of starting Firefox for the first time, I start 
getting lockups that last 5-30 seconds, with no more than 5-10 seconds between 
lockups.  Preferences, extensions window, themes window, in-browser, doesn't 
matter.  The only time I don't get the lockup is if I open the preferences 
window without opening Firefox itself.  Haven't tried any of these workarounds 
but I'm sure they'll work the same way.
Did run into the same Problem with the new Computer of a friend:
Intel D865PERL mb
P4 3Ghz HT
2Gb RAM
On the completely fresh install of Windows XP Pro SP2 ("selfmade" Disk with SP2
included) <- could this be important?, Firefox 1.0 had severe lockups (which
Site seemed not to matter), there where absolutely no Software/Drivers installed
(even not Intel inf update etc.)
Setting the affinity to just one CPU solved the Problem.

    
Whenever i opened a new browser window the whole browser froze for about ten
seconds. After some time i was able to track the error down to my current
version of the add block extension. Upgrading the extension and restarting fixed
the bug.
Regarding <a href="#c11">#11</a>:

Mark, I think your description does not belong to this bug.

I experienced this bug with freshly installed versions of firefox and mozilla
without extensions. Others are talking about fresh installations, too.

Furthermore, you reported the hangs to be bound to a specific action (opening
new window) which is different from experiences reported above.

You could try tcpview or something else to monitor whether you have a hanging
connection to localhost to verify this.
Another workaround is to create a shortcut for Firefox in XP that forces the
program to be run in Windows 98 compatability mode. This disables hyperthreading
for Firefox. This effectively does the same thing as the 'affinity' workaround,
without the need to set it everytime you restart firefox.

Hope this helps,
Ted
I just tried latest-trunk (Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US;
rv:1.8b2) Gecko/20050318 Firefox/1.0+).  The issue seems a bit rarer but is
still there.
Installed (upgraded) to a P4 3Ghz & instantly FF, Thunderbird & Netscape 
started exhibited freezing issues. Performed Task Manager Affinity setting to 
1 processor & problem went away immediatly. Problem is repeatable, reenable 2 
Processor Affinity freezing starts again. 
Hi, I am new to bugzilla, so please forgive any silliness.

I can reproduce this on my P4 3.2 GHz with HT, XP Home SP2.

Note that during the hang there are TWO Firefox instances listed in Task 
Manager.
I too can confirm this with mozilla 1.7.6 on a brand new P4 3.2, 865PERLL mb,
with a fresh install of Windows XP Professional which I just purchased retail
and included SP2 (In other words, I did not install the service pack on top of
old XP version).
Reseting the 'Affinity' as described above by others proved to be the solution
for me.
Found another workaround that emulates the affinity setting, really obvious
actually, set the compatability mode of the .exe to windows 98/Me (this works
for Firefox & Thunderbird. This is reliable & has worked for 2 days now. But I
would still like to see it fixed.
Ooooooops. Please forgive my haste in confirming this bug. I had not updated the
BIOS on my motherboard. After I ran Intel's Express BIOS update, everything is
working fine for me now. Sorry.
<a href="#c19">comment #19</a> is very interessting. It made me update the bios
and clearing CMOS thereby. Now the bug seems to be gone for Firefox 1.0.3 and
Mozilla 1.7.7. The localhost connection is still there, but doesn't hang anymore.

If others can verify this behaviour, this should close the bug. Try clearing
CMOS checksum without updating the bios first, to check what part is pivotal.

It's still confusing that things deep in hardware configuration affected a
Mozilla/Firefox localhost connection and, apparantly, only those.

Just for the record, before that, the bug persisted thru new installs of
Firefox/1.0.1 (including a deinstall of 1.0, moving away personal configs,
attempt to clean registry, starting in Safe Mode) and Mozilla 1.8b, 1.7.6, 1.7.7.
I upgraded to ECS BIOS 050322S (on my 915P-A 1.1), which not only appears to
have fixed this issue but to have improved other things as well.  I haven't been
able to trigger the freeze (yet, but I doubt it'll trigger), browsing feels
snappier, and Acrobat has stopped crashing from within Firefox.

I'm OK with closing this bug, but, given that it seems to hit people pretty
frequently, it might be nice to document it somewhere or somehow detect these
buggy BIOSes.  No clue how, though...

Thanks!
I would prefer if this bug was NOT closed for the following reason:

It would be beneficial to find out exactly where and why the deadlock is
occurring in the code. This may help with other deadlocks and/or hyperthreading
issues.

Not all BIOSes affected have had upgrades yet. However much the bug may be the
fault of the BIOS, Mozilla Foundation software seems to be the only software
affected. To those end-users unfortunate enough to have these hyperthreading
issues and without the savvy to investigate, "it's not our fault" sounds rather
unprofessional.

Investigation can't hurt and may help a great deal.
WFM on Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050530
Firefox/1.0+
Any sure indication this is actually HT related?

I think this is a dupe of: 266627
(In reply to comment #24)
> Any sure indication this is actually HT related?
> 
> I think this is a dupe of: 266627

This is definitely HT related. Turning off HT in the BIOS makes it go away.
Setting affinity to one processor (or equivalently running in Win98/Me
compatibility mode) makes it go away.

The repeated freezing isn't related to any particular action on the part of the
user, unlike 266627.
I believe I am encountering this problem on an older dual 600Mhz with a P2B-DS 
motherboard... it has the latest BIOS (granted, it hasn't been updated in a 
while, but it's an obsolete board), and no hyperthreading. Perhaps this is an 
issue w/ dual CPUs of any type vs hyperthreading?

(In reply to comment #25)
> (In reply to comment #24)
> > Any sure indication this is actually HT related?
> > 
> > I think this is a dupe of: 266627
> This is definitely HT related. Turning off HT in the BIOS makes it go away.
> Setting affinity to one processor (or equivalently running in Win98/Me
> compatibility mode) makes it go away.
> The repeated freezing isn't related to any particular action on the part of 
the
> user, unlike 266627.


This appears to be happening to me while using Mozilla/5.0 (Windows; U; Windows
NT 5.1; en-US; rv:1.8b5) Gecko/20051006 Firefox/1.4.1 (The 1.5b2 Firefox build).
 Strangely, this never seemed to happen before the second beta release.
The shows the resource usage over time.  At the time the snapshot was
taken,Firfox was using 100% CPU.  The period is about 18 seconds.  No other
activity was taking place.

This did not happen with earlier Firefox.  (This is Firefox 1.0.7 on Win2k.)
Is anyone who experiences this able to make a debug build and provide a stack trace? Not being able to reproduce a problem makes it nearly impossible to fix.
Assignee: firefox → nobody
I can reproduce the problem on Firefox 1.5b1 and 1.5b2.
On todays trunk ( Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a1) Gecko/20051027 Firefox/1.6a1 ) I was unable to reproduce it, seems like the problem is fixed, will test more and report here if it will appear.

most if not all hyperthread issues have been duped to Bug 282392 which details updating the PC's bios should FIX the problem

Jeff comment 5, John comment 22 and Eric comment 26 may be exceptions - so more details are needed from you guys per Gavin's comment 29

workaround for HT machine from bug 282392:

Download imagecfg.exe from http://www.robpol86.com/misc_pgs/imagecfg.php
Execute "imagecfg.exe -u <program-name>.exe" (according to your paths). 
It will always start in a single cpu mode


[bill's comment 28 is prob a different bug]
Severity: critical → major
Keywords: perf
Summary: Firefox freezes/hangs frequently and temporarily; possible deadlock? → Firefox freezes or hangs frequently and temporarily at zero or low cpu - HT / hyperthread / multiple processor only - NOT resolved by BIOS update
Whiteboard: workaround comment 31, do not dup to Bug 282392!
(In reply to comment #0)
> User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5)
> Gecko/20041107 Firefox/1.0
> Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5)
> Gecko/20041107 Firefox/1.0
> 
> Firefox very frequently freezes (windows detects it as not responding, the
> mouse
> shows an hourglass, and input has no effect).  It will usually unfreeze after a
> few seconds or minutes, but it will eventually freeze again.  This sometimes
> happens immediately after starting Firefox and sometimes just when interacting
> with the browser.  CPU usage is 0 when this happens.
> 
> This happens on my computer, after replacing CPU and motherboard, removing a
> bad
> hard drive, reformatting, then reinstalling just XP SP2 (slipstreamed), the
> network driver, and Firefox.  I've also seen it on two network adapters.
> 
> Now for the strange part: if I run tcpview (www.sysinternals.com) and close the
> TCP connection that Firefox opens to itself, firefox starts taking 100% CPU,
> but
> it also functions flawlessly, will unhang itself, and will not hang again.  I
> suspect some kind of deadlock or missed wakeup, and when I kill the connection,
> all the select()s or whatever Firefox uses return immediately, and the problem
> is hidden.
> 
> Is there anything I can do to provide more information?
> 
> Reproducible: Sometimes
> 
> Steps to Reproduce:
> 1. Run Firefox
> 2. Do anything.  www.slashdot.org usually does it, but so do most pages.
> 
> Actual Results:  
> Firefox freezes for awhile.
> 
> Expected Results:  
> It shouldn't freeze.
> 
> I have two monitors at different resolutions.  That's the only odd thing about
> my setup I can think of.
> 

I have read in another forum, that this ,ay have to do with computers that large amounts of fonts installed on them i.e. in the 1000's. wikipedia.org freezes for me, i believe it is scanning for all the fonts and this creates a delay about 20 seconds in my case
Summary says "HT / hyperthread / multiple processor only" and OS: field is set to "Windows XP". 
Can /usepmtimer in boot.ini be a workaround when MS Win XP SP2?

See Bug 307527(Bug 307527 Comment #28 and following some responses to it.)
See also Bug 307527 Comment #101.
I think that I was having this problem.  My firefox would hang on page loads, but it would not unhang within 1 minute (I'm not waiting longer than that).  I have a P4 2.8 HT proc.

https://bugzilla.mozilla.org/show_bug.cgi?id=266627#c30 suggested that this nightly might help:
http://stage.mozilla.org/pub/mozilla.org/firefox/nightly/experimental/threadmanager/

running that version seems to get rid of the hanging problem I was having with ff 1.504
*** Bug 361319 has been marked as a duplicate of this bug. ***
I have the same problem but using Windows Task Manager, when my Firefox freezes, McSheild.exe is hogging all the CPU (88-96%). This only occurs when clicking on Firefox. It does not occur when clicking on Explorer.
Interestingly enough I reported this same problem in Bug 361319 and used the imagecfg method to work around it.  I was using Firefox 2.0 at the time.  Since then I've updated to Firefox 2.0.0.1, 2.0.0.2 and 2.0.0.3 and never had an issue.  I didn't realize until today that the imagecfg fix only fixes the executable so since Firefox 2.0.0.1 I've been running on both "processors" without issue.  I confirmed this today by checking the Affinity in Task Manager to see it was using both CPUS.

The machine I had the problem on was accidentally not downloading and installing Windows patches from Microsoft since October 2006 until a few weeks.  Since I reported that bug in November 2006 and Firefox 2.0.0.1 was released at the end of December 2006, that would mean that the problem would have had to have been corrected somehow in Firefox 2.0.0.1.
(In reply to comment #37)
> The machine I had the problem on was accidentally not downloading and
> installing Windows patches from Microsoft since October 2006 until a few weeks.
>  Since I reported that bug in November 2006 and Firefox 2.0.0.1 was released at
> the end of December 2006, that would mean that the problem would have had to
> have been corrected somehow in Firefox 2.0.0.1.

 Patch for Bug 307527 changed from use of QueryPerfomanceCounter() to use of timeGetTime() in NSPR, in order to bypass MS Win's problem when mutiple CPU & acpi(PowerNow! when AMD), and the patch is involved in Firefox 2.0(branch-1.8.1+ on 2006-02-06). Example of problems in MS Win is in Bug 307527 Comment #101. Following is list of MSKB in the comment.
> MS knowledgebase'es,
>  - http://support.microsoft.com/kb/Q821893 
>  - http://support.microsoft.com/kb/Q835730 
>  - http://support.microsoft.com/kb/Q896256
 Solution of MS Win's problem is already provided by Hotfix(inludes HAL update) for Win XP SP2(and porobably for Win-2K SP4), although it is not shipped as Service Pack yet.
And "driver update"(includes HAL update) by CPU/PC vendor, which is equivalent or similar to MS's Hotfix, is also already released by some vendors.
 So, if you don't experience problem with Hotfix for MS Win, or if you don't experience problem with Firefox 2.0 on MS Win without Hotfix or bypass(e.g. /usepmtimer, disabling of PowerrNow!, and so on), cause of your problem is possibly MS Win's problem fixed by the Hotfix.

To Michael Kraft:
Do you think your problem is resolved by the Hotfix?
Do you think that your problem is same(not only external view of problem but also cause of problem) as bug opener's one and all problem reporters one before you reported?

FYI.
Name of ACPI technology which relates to MS's Hotfix.
 - Intel : "Enhanced Intel SpeedStep technology"
 - AMD   : "PowerNow!"/"Cool'N'Quiet"
I am having the "temporary" screen freeze also. I have 2 monitors, and a Intel Core 2 Duo E6700, so the HyperThreading bug may have reared its ugly head again.

Here is what happens: 
Open FF, open multiple tabs and even multilple windows, and on occasion, the app will freeze. Mouse will work, but screen appears frozen. I can go to the other window and grab a different (non-FF) app window, and drag it 1/2 way across the 2 screens, and it will "unfreeze" the frozen screen. Just before it unfreezes, the whole screen goes black (both monitors) for about 1 second, and then everything is ok again (until it happens again, which can be from 1 second to 5 minutes later). I cannot narrow down if it is a certain type of site that is tripping it up. After it is unfrozen, FF is fine. 1 FF will freeze on 1 screen, and the other on the other screen is fine, almost like it is the entire screen that is frozen (further evidence is that my Taskbar does not "unhide" when the frozen FF is on the main screen). 

My extensions are minimal: del.icio.us 1.2.1, Download Statusbar 0.9.5, IE Tab 1.3.3.20070528, No Script 1.1.6.02, Talkback 2.0.0.4 

My system is as follows: 
Asus P5N32-E SLI nForce 680i SLI Socket 775 1333MHz DDR2-800 
CORSAIR XMS2 TWIN2X2048-8500C5, DDR2 1066 (PC2 8500), 240 PIN 
Intel® Core 2 Duo E6700 Conroe Processor 2.67GHz, 1066FSB 
2x Western Digital Raptor WD1500ADFD 150GB Serial ATA 10,000RPM Hard Drive w/16MB Buffer 
XFX|GF 7800GT 256MB 
CREATIVE LABS SB AUDIGY 2 ZS 
PLEXTOR PX-716A 4X DVD+R/RW 
MITSUMI Black 1.44MB 3.5" Internal USB 2.0 digital card reader with Floppy Drive 
ZBOARD KEYBOARD 
WIN XP HOME 
2X DELL 1905FP 19" MONITORS 
NZXT CASE 

(In reply to comment #40)
> I am having the "temporary" screen freeze also. I have 2 monitors, and a Intel
> Core 2 Duo E6700, so the HyperThreading bug may have reared its ugly head
> again.
> 
> Here is what happens: 
> Open FF, open multiple tabs and even multilple windows, and on occasion, the
> app will freeze. Mouse will work, but screen appears frozen. I can go to the
> other window and grab a different (non-FF) app window, and drag it 1/2 way
> across the 2 screens, and it will "unfreeze" the frozen screen. Just before it
> unfreezes, the whole screen goes black (both monitors) for about 1 second, and
> then everything is ok again

It sounds that problem of external software(such as Graphics driver/Mouse driver) when TSC is used by QueryPerfmanceCounter or TSC is used by the software on PC of MS Win/muti CPU with acpi. 
Read all MSKB articles pointed in this bug. See next article for what happens on QueryPerfomanceCounter.
  http://www.gamedev.net/community/forums/topic.asp?topic_id=370281

Similar phenomenon was reported to Bugzilla Japan, and reporter said that "/usepmtimer" in boot.ini resolved his problem.
Will "/usepmtimer" be a workaround? (MS Win XP SP2 is required)
Firefox locks up often (many times throughout the day).

Every time Firefox does this, the Firefox.exe process is at 50% (sometimes 49%, sometimes at 51%, but the vast majority of the time, it is 50%).

I started noticing this in the early Version 2.0.
The first couple were fine (2.0), but after the 2nd or 3rd version (like 2.0.0.4), Firefox hangs most of the time.

This can happen by just clicking on another tab, without reloading any web pages and/or going to a different page.

For me, it can last minutes before Firefox starts responding again.

And, when Firefox hangs, ALL Firefox browser windows hang.
No scrolling, can't change tabs, etc.

I have about 4 Firefox windows open.
One of them only had one web page open.
Another has about 30 tabs open.

OS: XPSP2
RAM: 4GB
Current Firefox version: 2.0.0.14

I'm thinking about uninstalling Firefox and going back to one of the earlier ones.
I am having this problem also.  I can't figure out any pattern to what is making it freeze.

OS: XPSP2
RAM: 2GB
Firefox version: 3.0.1
I'm seeing this on occasion as well.  I've seen two types of freezing:

1. Something in javascript causes a hangup and the page freezes for about 30 seconds and then pops up a message about a script taking a long time and asking me to stop it.  This type uses a lot of CPU.

2. The window just freezes.  Opening windows task manager shows that it says Firefox has stopped responding. What's bizarre is that on the Applications tab, it shows two entries for Firefox even though only one window was open.  Right clicking on an entry and choosing "Go to Process" jumps to the firefox.exe entry on the processes tab.  Doing the same on the other entry jumps to "explorer.exe" which is Windows Explorer.  I think this problem is triggered when Firefox tries to load Flash or PDF content, but I'm not positive.
Version: unspecified → 3.0 Branch
The last report here was three years ago, and there's nothing specifically actionable in this bug without further input from reporters, so I'm just going to close this.
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → INCOMPLETE
I've been experiencing this since I've upgraded to dual-core (AMD) a few weeks ago.
CPU uses drops down while Firefox hangs temporarily.
I'd be happy to provide more input if anyone picks this up, and/or I'll open a new bug if told so.

Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:20.0) Gecko/20100101 Firefox/20.0
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: