Firefox 3.5.6 crashes Athlon X4 based system

  • Thread starter Thread starter tanix
  • Start date Start date
T

tanix

I have an Athlon 2, x4 620 quad core processor system on\
Asus M4A78 motherboard.

When I run firefox 3.5.5 or 3.5.6 and refresh several tabs in
rapid succession, it freezes the box unde xp sp3.
What happens is task manager -> performance tab shows one
of cores going upto 100% of CPU useage, mouse freezes, keyboard
is dead and disk stops flashing every couple of seconds like it
normally does under windows. System is totally dead and needs
to be rebooted. Anyboyd knows anything about it?
 
tanix said:
I have an Athlon 2, x4 620 quad core processor system on\
Asus M4A78 motherboard.

When I run firefox 3.5.5 or 3.5.6 and refresh several tabs in
rapid succession, it freezes the box unde xp sp3.
What happens is task manager -> performance tab shows one
of cores going upto 100% of CPU useage, mouse freezes, keyboard
is dead and disk stops flashing every couple of seconds like it
normally does under windows. System is totally dead and needs
to be rebooted. Anyboyd knows anything about it?

Have you asked Firefox support?

http://support.mozilla.com/en-US/kb/

Do you have access to any other computers where you can see if the issues
occur in other hardware configurations?

I installed Ff on a computer here w/Phenom x4 955 @ 3.6 GHz and 8 Gig,
running W7 Ultimate x64 (as close as I can come to your hardware) and tried
to replicate your scenario. No issues at all.
 
I have an Athlon 2, x4 620 quad core processor system on\ Asus M4A78
motherboard.

When I run firefox 3.5.5 or 3.5.6 and refresh several tabs in rapid
succession, it freezes the box unde xp sp3. What happens is task manager
-> performance tab shows one of cores going upto 100% of CPU useage,
mouse freezes, keyboard is dead and disk stops flashing every couple of
seconds like it normally does under windows. System is totally dead and
needs to be rebooted. Anyboyd knows anything about it?

Had something similar after upgrading ram from 2 to 4gb. Memtest ran
fine, but when my system ran mythcommflag it would crash.Turned out to be
the ram voltage, which was rated at 2.1v was only being set to 1.8
something. Which was not enough under a large system load. Set it
manually to 1.9 in bios iirc and never had another problem. ymmv.
 
Have you asked Firefox support?

Yep, I did. So far, no luck. Its been 2 days already.
But there is so much traffic on this forum, I am not sure anyone
would care about this issue since it is not win 7 related.
http://support.mozilla.com/en-US/kb/

Do you have access to any other computers where you can see if the issues
occur in other hardware configurations?

Works like a champ on a single core processor.
Never had a problem with it under ff 3.5.5.
I installed Ff on a computer here w/Phenom x4 955 @ 3.6 GHz and 8 Gig,
running W7 Ultimate x64 (as close as I can come to your hardware) and tried
to replicate your scenario. No issues at all.

Well, the thing is it DOES work find under win 7.
Someone said the xp scheduler is crappy and it is not a firefox problem.
Could be, who knows. May be ff just creates some nasty conditions and
scheduler can't handle it.

Someone even said there is a fix for xp scheduler.
But I was not able to find anything so far.

This is pretty bad situation for me. Cause I am doing a lot of important
work and may loose all sorts of stuff if I forget that ff may lock up
the box.
 
Had something similar after upgrading ram from 2 to 4gb. Memtest ran
fine, but when my system ran mythcommflag it would crash.Turned out to be
the ram voltage, which was rated at 2.1v was only being set to 1.8
something. Which was not enough under a large system load. Set it
manually to 1.9 in bios iirc and never had another problem. ymmv.

I ran memetest86 for at least 20 hrs. No problems.
I also ran this Prime95 stress test, which loads up all the cores
to the max doing some heavy duty math processing.
Also ran fine for several hours.
Except that one does not do much testing of other things.
Could be a network driver or just about anything you can imagine.
So, doing just math processing isn't good enough of a test.

I wonder if there is some other stress test, especially for XP
multicore issues.
 
What happens when you run Firefox in safe mode?

Well, I did not try it yet. Some people suggested doing it.
I am just not sure what does it achieve.
Right now I disabled all the addons except of imacros,
cause that is the one I need the worst and never had a problem
with it on a single core system.
 
Hi, Tanix.



Like Wes, I had problems after doubling RAM from 2 GB to 4 GB last year.
I've never run FireFox at all. When Vista went RTM 3 years ago, I rebuilt
my computer to celebrate. I got a new mobo/CPU/RAM all from NewEgg a kind
of a matched set: EPoX MF570sli mobo/AMD Athlon 65 X2 5000 CPU/2x1 GB OCZ
PC-6400 SDRAM. This ran Vista Ultimate x64 just fine for about a year.
Then I ordered another 2x1 GB OCZ RAM from NewEgg, identical to the first
pair. Everything was good for about 6 months.

Then random BSODs started happening. I tried MANY fixes, starting with a
through internal cleaning, new Antec PSU, even a new 1 TB SATA HDD.
Memtest86 said all the RAM was fine. Finally I gave up and took it to a
reliable local shop that I've dealt with before. Casey tested it thoroughly
and replaced 3 fans, but the problems quickly returned, so I took it back to
the shop. He ran some more stringent RAM stress tests and said that it
passed them all, but when he took out one of the two new sticks, the BSODs
stopped. I ran for the next few months without that one DIMM and had no
problems.

When I contacted OCZ for RMA, the tech insisted that I send BOTH new sticks
so that they could be tested by the factory as a matched pair. OCZ sent me
2 new DIMMs; again, apparently identical to the original pair. That was
last Spring and I've had no more BSODs. ;<)

My point is that some RAM problems are so subtle that we really can't
diagnose them in the field. And the problems they create can mimic almost
any kind of computer defect, sending us on many wild goose chases. The one
consistency is inconsistency; we seldom get the same Stop Code twice. We
can suspect the PSU, the hard drive, drivers, applications, heat buildup,
RFI...just about anything.

You might want to contact the maker of your RAM about a warranty replacement
of some or all of your memory. Many brands sell with a lifetime warranty.

RC

Yep, there are things like DMA transfers, burst mode and you name it
related to ram. I bet the memtest86 is way too primitive to exhaustively
test all the access modes. Just flipping bits around can only detect
the obvious cases, such as some data bits being shorted out and things
like that. But ram was the first thing i suspected. We'll see how it goes.
 
Back
Top