Why is Firefox shutting down my PC?

December 19, 2007 8:29 PM

Firefox is shutting down and restarting my PC, what gives?

Suddenly, Firefox randomly shuts down my PC and I am baffled as to why it is doing this. I've tried reinstalling it, even deleting old profiles. I run adware, errorkiller, and hitman pro every few days.

My Flock browser has even started doing the same shutdown and restart while IE and Opera run without shutting down and restarting my PC. All other programs run fine.

I have 1.0 gig of RAM on an XP run PC. Could this be a memory issue? Virus or worm?

posted by jasonspaceman to computers & internet (15 comments total)

Random shutdowns sound like a hardware problem. Is it a laptop? Could it be a heat issue? Perhaps if you watch youtube videos in firefox that is stressing the CPU and it is shutting down to avoid overheating. Also you could check if it's a memory issue by downloading and running memtest.
posted by PercussivePaul at 8:37 PM on December 19, 2007

I also forgot to mention ... I've checked all of my fans and they are running.

Paul, it is desktop.
posted by jasonspaceman at 8:38 PM on December 19, 2007

When Windows XP randomly shuts down and restarts, it's almost always because of a BSOD. The default settings hide the bluescreen by instantly rebooting.

Open up your event log. (Control Panel -> Administrative Tools -> Event Viewer.) Look at the system log for any entries of number 1000 or 1001. (Click the View menu, then Filter. For "Event ID" specify 1000 and then click OK, then do it again for 1001.) Pick the most recent one and copy/paste it here. It'll look something like this:

The computer has rebooted from a bugcheck. The bugcheck was: 0x000000ea (0x8213cc90, 0x838b4778, 0xf7b1acb4, 0x00000001). A full dump was not saved.

For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.

If you'd like to change it so it shows you the bluescreen instead of rebooting, go to the Control Panel, and open the System panel. Go to the Advanced tab, and under Startup and Recovery, click the "Settings" button. Uncheck the "automatically restart" checkbox.
posted by CrayDrygu at 9:03 PM on December 19, 2007 [3 favorites]

You say the fans are all running; are they running hard (especially the CPU fan)? Because if they are, you may just have CPU or video card heatsinks clogged with dust.
posted by flabdablet at 2:25 AM on December 20, 2007

CrayDrygu, there aren't any 1000 or 1001 entries in the log. The last entries that do show up are SidebySide errors.

Event Type: Error
Event Source: SideBySide
Event Category: None
Event ID: 32
Date: 12/19/2007
Time: 11:14:22 PM
User: N/A
Computer: xxxxxx
Description:
Dependent Assembly Microsoft.VC80.MFCLOC could not be found and Last Error was The referenced assembly is not installed on your system.


For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.


The others are Service Control Manager source type of errors. To me, all of these look like they are from when the computer restarted. I could be wrong though.

Flabdablet, the fans are not running hard and they have been recently cleaned out.
posted by jasonspaceman at 5:45 AM on December 20, 2007

You say "Firefox shuts down my PC". What do you mean by that? What does Firefox have to do with it?
posted by dmd at 6:07 AM on December 20, 2007

dmd, when I am using firefox, my pc shuts down. In frustration, I am saying it is shutting down my pc. In the rest of my text, you can see I am open to it being something other than firefox.
posted by jasonspaceman at 6:45 AM on December 20, 2007

Even if there are no 1000 or 1001 errors in the log, I suggest following the steps in the last paragraph of my post above. This way, the next time it reboots on you, one of two things will happen:

1) You'll get a bluescreen error. Copy down the name of the error from the top of the screen (it'll be all-caps and have underscores, like "DRIVER_IRQL_NOT_OR_EQUAL"), and the string of hex numbers from near the bottom (something like "0x000000ea (0x8213cc90, 0x838b4778, 0xf7b1acb4, 0x00000001)").

2) You'll get no bluescreen, it'll just reboot. If it happens this way, it's most likely a hardware problem.
posted by CrayDrygu at 12:49 PM on December 21, 2007

#2 happened. That's what I was afraid of. Not really sure what to do next other than buy a new PC.
posted by jasonspaceman at 7:40 PM on December 21, 2007

You can probably narrow down the problem, at least. The most likely culprit is a memory problem.

Download memtest86+ -- use the "Download - Pre-Compiled Bootable ISO (.zip)" link. If you need a program that can burn an ISO to disc, I like Active@ ISO Burner.

Boot your computer off of that CD and let it run overnight (or at least 2-3 hours). If you've got red text on your screen in the morning, that means your RAM is bad.

If that test doesn't indicate anything, it's probably your motherboard. More testing could narrow it down, but the tests would probably involve swapping components.
posted by CrayDrygu at 8:03 PM on December 21, 2007

The test never lasts more than four or five minutes, the pc shuts down. So, I am going to take out one of the memory sticks and see if that helps.
posted by jasonspaceman at 6:14 AM on December 22, 2007

jason, thats a bad sign. The problem is likely something other than the memory -- or else the memory is failing in some spectacularly bad way. In any case, you can be sure it's a hardware problem and not something do to with your windows installation. The next thing I would do is remove any extraneous parts. Take out all cards except your video card. Re-run the test. Then you can try removing RAM if it fails again.

Aside from memory, another failure that can cause spontaneous reboots is blown capacitors. You can spot these visually - just look for a bulge or a leaking electrolyte. They can sometimes be replaced, but you would probably need a professional to do it so it would likely cost you.
posted by PercussivePaul at 9:42 AM on December 22, 2007

I agree with PercussivePaul there; it's not a good sign that memtest won't keep running. I've honestly never seen that before.

At this point, I'd say it's most likely a motherboard problem. It can't hurt to swap the memory and other components and try memtest again, of course. But if that doesn't help, you'll probably need to replace the motherboard. I'd take it to a smaller local PC shop for that (not a big-box retail shop), and let them know what troubleshooting steps you've gone through. It'll probably cost you half as much as buying a whole new PC would. Assuming you don't have a Dell, anyway.
posted by CrayDrygu at 1:43 PM on December 22, 2007

You don't need a truly spectacular failure to crash Memtest with bad RAM. A single stuck address line will do it.

If the highest address line in the array is stuck, the RAM will effectively be operating as two clones of half the address space. If the line is stuck high, only the top half of the RAM array will actually be in use; if it's stuck low, only the bottom half will be in use - but the half that is actually in use will appear in both halves of the address space.

If the second-highest line is stuck, you get a similar failure pattern except that the first and second quarters of the address space become mirrored, as do the third and fourth quarters.

Memtest has to run in some part of RAM in order to function at all. So, if there's a stuck high-order address line, Memtest will end up inadvertently writing over some part of itself as it tests the address space where its mirror image is.

If that's what you're dealing with, you will find that Memtest dies after the exact same runtime on every try, and two things will fix it: (a) remove the faulty RAM card from the machine altogether, so that Memtest loads into and tests the other one - in this case it should complete without failure; (b) swap the RAM cards around, so that Memtest loads into the good one, in which case it should report the address line fault in the bad one.

Note that if your mobo is configured to turn on dual channel RAM when it sees two identical RAM cards, a stuck RAM address line will still crash Memtest in case (b). So unless you're looking specifically for a failure that only occurs in dual-channel mode, it's generally easiest to run Memtest with only one RAM card installed at once.
posted by flabdablet at 11:51 PM on December 22, 2007

If you find that Memtest is taking variable amounts of time to die, then a permanently stuck address line (either in the RAM card or on the mobo) is probably *not* the cause. The most likely cause is excessive bus noise, and this in turn is probably due to bad power. If you've got capacitor plague (look for bulging or leaky caps as described above) then the source of that bad power is the DC to DC converters built onto your mobo. If the mobo caps look OK, the most likely cause is the power supply itself.

A PSU is a cheaper and easier thing to replace than a mobo. If you whip the case off the PSU and see bulging or leaking capacitors inside it, or something smells bad and/or looks burnt in there (just don't *touch* anything electronic inside the PSU case, since there are capacitors in there that remain charged to full mains voltage even after the input power is disconnected) then a replacement power supply is definitely called for.
posted by flabdablet at 11:59 PM on December 22, 2007