BSOD: irql_not_less_or_equal AND page_fault_in_nonpaged_area

koolkat77 · 21 Feb 2013

Can you upload the latest reports?

Make scans with the following:

-Kaspersky TDSSKiller

Anti-rootkit utility TDSSKiller

-ESET online scanner

Free Online Virus Scanner | ESET

Microsoft Security Essentials.

Recommended from a strict BSOD perspective, compatibility & stability compared to other antiviruses/internet security software. It is free and lightweight:-

Warning

Do not start the free trial of Malware Bytes; remember to deselect that option when prompted.

Run a full scan with both (separately) once downloaded, installed and updated.

DebugDan · 22 Feb 2013

I appreciate the quick response, but I don't think it can be a software problem.
I did some debugging with a friend of mine who is good with computer hardware.

We stripped the system down to just the power supply, motherboard, and CPU.
We switched out the RAM for some from a different working computer.
We switched the hard drive with a different one, with a different working copy of windows 7.
The result: the BSOD happened almost immediately after windows booted.

Since we ruled out the hard drive and RAM, we figure it must be the CPU causing the problem.
According to the specs on the error from the microsoft website, the "page fault in nonpaged area" error is being caused by a memory read that fails (it actually mentioned the L2 cache on a CPU as a potential problem)... I forget the link.

I was going to take the CPU out and have it tested somewhere.
Do you agree that the CPU is probably the problem?
It couldn't be the motherboard itself somehow, could it?

Again, Thank You.

essenbe · 22 Feb 2013

When you ran memtest86, did it run for 8 full passes. Each pass has 10 tests. It seems really quick to run that many passes that quick. It usually takes me 8 hours to run 8 passes. It takes 8 passes to fully check the ram. Each pass checks different things.

Also, can you look on the side of the PSU and tell me the make and model as well as the amps for the +12V? Also how old is the PSU? According to AMD, the PSU is right at the required voltage, meaning if the PSU has degraded or is not putting out the rated voltage, you can have all kinds of problems.

EDIT: also, please go to bios and go to the hardware monitor section and tell us the readings for the +12V, +5V and the +3.3V.

If you don't have it already please post screenshots of CPUz showing the Mainboard tab, the memory tab and the SPD tab (2 shots showing each ram module)
Screenshots and Files - Upload and Post in Seven Forums

essenbe · 22 Feb 2013

You may find this helpful RAM - Test with Memtest86+

You may also find it helpful to read the user feedback on your motherboard.

DebugDan · 22 Feb 2013

Here is all the information about my PSU.
Bought it new at the same time we bought the rest of the components for the system (early 2012).

I couldn't find the +12V, +5V, or +3.3V readings you mentioned in the BIOS...
(BIOS is up-to-date according to ECS website)
Is there another way to make sure the PSU is working properly?

According to AMD, the PSU is right at the required voltage, meaning if the PSU has degraded or is not putting out the rated voltage, you can have all kinds of problems.

Are you referring to the voltage specs for the CPU? Where did you find this information?

It has gotten to the point now that I can no longer use windows... it crashes before, during, or immediately after the boot every time.... so no CPUz, sorry.
The RAM can't be the problem, can it? When I was testing, I removed both sticks and put in a new stick from my working machine and still got the error.
The memtest looked like it completed the passes successfully... but if I need to I can run it again.

Wow... those are some terrible motherboard reviews.
Wish I would have known about them when we bought the thing (early 2012).
Given that, I'm now leaning more toward the motherboard being the culprit.
(That, and I was reading and apparently CPUs rarely die that quickly)

I feel like we're getting closer but how do we decide which is the problem between the PSU, Motherboard, and CPU?

Thanks,
Daniel

essenbe · 22 Feb 2013

If you have another PSU or have a friend with one, try another PSU before spending money on one in case that is not the problem. In your bios you should have a hardware monitor section that should tell you the values of the PSU as well as other things too. I was wrong about the minimum power requirements, I read the wrong system specs I guess.. Your PSU is plenty if it in fact puts out 700W. Many of the cheaper ones, claim to, but do not.

Bad ram can cause all kinds of serious problems. We can test the ram, CPU, GPU. You can only test the PSU with a multi meter, even then it only tells the values at idle. The problem with PSU's is there is little way to test them under stress. As best I can determine, your PSU has 2 rails with 23A on each rail. Make sure they are divided up correctly. From what I can find with a quick look is the max power draw for your card is 116W, so wattage should be OK.

When you said you put in a working ram stick from your machine and still got the error. Are you talking about an error on memtest? I really don't suspect the CPU unless it is overclocked. If you are overclocking anything, stop. Set your bios to optimized defaults. Also, set your ram by manufactures specs. Set it by hand, set the frequency, timings and voltage by hand. The info for that should be on the ram sticks.

If your machine will not run, it is difficult to test anything. We need to get it running first. Do you BSOD while in bios? You can test the hard drive with Seatools, Seagate's diagnostic program, Stress test with Prime 95 for the CPU, stress test the GPU with Furmark. Have you received any error messages other than the bsod messages?

DebugDan · 22 Feb 2013

Given some time, and only if necessary, I can gain access to a 400W and/or 550W PSU to test with.
Both of which would involve dismantling an existing system to remove the PSU.
I would like to save this as a last resort, but let me know if we must go down this path.

No overclocking here... I haven't changed any BIOS settings.
I reset the CMOS just because I hadn't tried that... didn't seem to help.
The BIOS that came installed from the factory is the most up-to-date version, so all settings are set to their defaults.
I looked in the BIOS and found some voltage information (looked everywhere, these were the only ones listed).
Current CPU Voltage = 1.3500 V
CPU vCore = 1.3500 V
VDIMM = 1.608 V
NB Voltage = 1.248 V
HT Voltage = 1.20 V

I am quite confident that the RAM sticks are just fine. (I suppose the RAM slots in the board could be bad though)
I swapped equivalent sticks from my working machine (the sticks from the bad machine are now in the good machine and vice-versa). No problems to report on my good machine (working fine for about 1.5-2 hours).
The bad machine worked for a minute or two longer than the norm, but then blue-screened all the same.
I never got an error in memtest with any memory stick.

Also, I don't think the graphics card is relevant because if I take it out and use the onboard graphics the BSOD happens just the same.

The BSOD never happens in BIOS (I can let it stay idle in BIOS for hours without a problem, if I wanted to).
I have not received any error messages other than the BSOD.

After about 10 attempts, I manged to get the data from all tabs in CPU-Z
Attached are the screenshots as well as the generated report.

What's the next step?
Thanks,
Daniel

essenbe · 22 Feb 2013

To the best of my knowledge, when in bios all of yor hardware is being used, but you get no bsod. Try to run the Seatools for DOS I linked to. Run it from disk and run the extended test. Also, if you haven't already run sfc /scannow. SFC /SCANNOW Command - System File Checker

DebugDan · 22 Feb 2013

The BSOD error occurs regardless of which hard drive I use.
I borrowed a different hard drive with a working copy of windows 7.
The errors happen just as frequently... I'd say that rules out the hard drive and OS...

Do you have a different suggestion for how to proceed?

Perhaps the BIOS uses all the hardware, but maybe not to the extent required to cause an error?
Is there a good usb-bootable utility to test the system? (power supply, board, cpu)

Thanks,
Daniel

essenbe · 22 Feb 2013

Wait just a minute. If you tried a hard drive from another machine, it will almost always BSOD. Different hardware, different drivers, chipsets. Why not try the bootable Seatools? You only seem to BSOD while in windows or using the drive, however you choose to look at it. You said you could stay in bios. In bios, the hard drive is not involved, only the hardware.

there is not test that I am aware of that will test the motherboard, and not reliably the PSU. The Seatools is a bootable hard drive test, much like Memtest. Try downloading Linux Mint or any of the live distros. Puppy Linux is probably the smallest download. Burn the ISO to a DVD or USB and boot from that. It does not run from the hard drive, it runs from ram. See if you BSOD there. But run the extended test from Seagate. You can't figure this stuff out taking shortcuts.