Random freezes with no BSOD

Thedoctor44 · 27 Apr 2014

0xA bugchecks are IRQL_NOT_LESS_OR_EQUAL
These are caused mainly by buggy device drivers, so it is important we get some new dump files.

I still think memtest86 needs a go for 8 passes.

Laurie · 27 Apr 2014

Yep, you got me there. Sorry about that! I really should have checked. I'm pretty sure the files are there this time.

OK I'll run Memtest86 overnight then so it gets to do a full eight runs.

Thedoctor44 · 27 Apr 2014

You should enable Special Pool in the Driver Verifier settings.

Code:

BugCheck A, {0, 2, 0, fffff8000329bbd4}

Probably caused by : Pool_Corruption ( nt!ExDeferredFreePool+1283 )

Followup: Pool_corruption

EDIT: I've found out your graphics driver is significantly outdated.

Code:

fffff880`0f26c000 fffff880`0fee0100   nvlddmkm   (deferred)             
    Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Timestamp:        Mon Jun 28 05:08:34 2010 (4C282042)
    CheckSum:         00C773D1
    ImageSize:        00C74100
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

Please update it, if you are unsure how here's some instructions.

NVIDIA Drivers - Avoid Problems

You don't have to pay attention to this but I just wanted to go into detail.

I'm fairly new but here's my analysis:

Code:

fffff880`031d8ca8 fffff800`03286169 : 00000000`0000000a 00000000`00000000 00000000`00000002 00000000`00000000 : nt!KeBugCheckEx
fffff880`031d8cb0 fffff800`03284de0 : 00000000`80000000 fffff880`031d8fe0 fffff880`031d8fc0 fffff980`16412ee0 : nt!KiBugCheckDispatch+0x69
fffff880`031d8df0 fffff800`0329bbd4 : fffff800`03401e80 fffffa80`08b51040 00000000`000018ca fffff800`032906f7 : nt!KiPageFault+0x260
fffff880`031d8f80 fffff800`032795f7 : 00000000`00000000 fffff880`031d5000 00001f80`00100000 0053002b`00000000 : nt!IopCompleteRequest+0xc64
fffff880`031d9050 fffff800`032798a7 : 00000000`00138743 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiDeliverApc+0x1c7
fffff880`031d90d0 fffff800`0325ed96 : fffffa80`07510198 fffff880`031d9410 fffff880`031d92d0 00000000`00000004 : nt!KiApcInterrupt+0xd7
fffff880`031d9260 fffff800`03348376 : 00000000`0002fd48 00000000`00000000 fffff880`031d93f0 00000000`00138743 : nt!RtlpLookupFunctionEntryForStackWalks+0xb7
fffff880`031d92d0 fffff800`03349743 : fffffa80`07510198 fffff880`00000011 00000000`00000000 fffff800`00000004 : nt!RtlpWalkFrameChain+0x2e6
fffff880`031d9970 fffff800`0334a5cb : 00000000`00000003 fffffa80`07510198 00000000`00000000 00000000`00000000 : nt!RtlWalkFrameChain+0x63
fffff880`031d99a0 fffff800`0372013c : fffffa80`07510180 00000000`00000040 00000000`00000003 fffff800`03376911 : nt!RtlCaptureStackBackTrace+0x4b
fffff880`031d99d0 fffff800`037221aa : fffff880`031d5000 fffff880`031db000 00000000`00000040 00000000`00000000 : nt!ViPoolLogStackCallout+0x1c
fffff880`031d9a00 fffff800`03724a4a : fffff8a0`00c87d10 00000000`00000004 00000000`00000001 00000000`00000003 : nt!ViPoolLogStackTrace+0x8a
fffff880`031d9a30 fffff800`033b9c57 : fffff8a0`00c87d00 00000000`00000040 00000000`00000003 00000000`00000001 : nt!VfFreePoolNotification+0x4a
fffff880`031d9a60 fffff800`03632476 : fffff8a0`007ea730 00000000`00000000 fffff880`031d9c10 fffff800`000004ef : nt!ExDeferredFreePool+0x1283
fffff880`031d9b10 fffff800`035bc3da : fffff880`0000001d fffff880`031d9c10 fffff8a0`00c87500 fffff880`0000001d : nt!CmpCallCallBacks+0x4f6
fffff880`031d9be0 fffff800`034eb48e : fffff880`031d9fb8 fffff880`031da1a0 fffff8a0`00023010 fffff8a0`000945d0 : nt! ?? ::NNGAKEGL::`string'+0x1212b
fffff880`031d9c60 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt! ?? ::NNGAKEGL::`string'+0x2c9a6

It looks like what happened is an APC Interrupt has occurred when the IRQL level is too high.

Code:

CURRENT_IRQL:  2

The IRQL for an APC Interrupt to complete is 1 and because Windows priorities the highest IRQL first the Interrupt cannot complete as a result a page fault has occurred so a bugcheck is then initiated.

Now I'm not too sure about the verify free pool functions that are happening before, I need to look into that.

Laurie · 28 Apr 2014

Thanks for taking a look Thedoctor.

OK so I've run Memtest86 for a full 8 passes, and it passed all of them, so I think that's OK. I've now followed your instructions and switched Driver Verifier back on with Special Pool checked. Just to clarify though, I only checked Special Pool. Is that what you meant, or did you mean me to tick some of the others as well? Anyhow for now, I'll just keep using the machine and see if I can force it into another BSOD.

Thedoctor44 · 28 Apr 2014

No what you've selected is fine.
Just let us know what happens.

Laurie · 29 Apr 2014

Since I switched Driver Verifier back on with the new settings, I've experienced a lot more freezes, particularly whilst browsing the web. This makes the computer quite hard to use, as I'm having to restart it quite often. However I've still had no more BSODs.

As an additional test, I ran the Furmark test, but I wasn't entirely clear from the tutorial how I should know if it had passed or not. I ran it for a couple of minutes and I got an alert message saying that the temperature had exceeded 100 degrees, and as the text was in red and block capitals, I figured it was probably time to stop the test in case it damaged anything.. I didn't notice any visual artifacts though, which was the other thing the tutorial said to look out for.

Thedoctor44 · 29 Apr 2014

Your GPU is overheating, is your case big enough?
Is their sufficient airflow?
You might need to get a new GPU.

Laurie · 29 Apr 2014

Well the computer is a laptop and I can't easily remove the graphics card, so I'm not sure there is much I can do on that front. Having said that, I've had the computer for over four years now and it's been working fine. Is an overheating graphics card the kind of thing that could cause these freezes? If so, what could have changed to make it start overheating? The fans are still working so is there some kind of thermal paste like on a CPU that could have dried out or something?

Thedoctor44 · 29 Apr 2014

Thermal paste has nothing to do with GPUs.
Well the problem with laptops being so compact is there isn't enough airflow so overheating can damage components.

You could change the graphics card but it depends whether it's soldered in or not.

Laurie · 29 Apr 2014

I'm pretty sure it's soldered in. The manual that came with it has guides for replacing most components, but it says nothing about the GPU. I can hardly even see it when I disassemble the machine, as it's hidden behind other things.

But I mean, how sure are you that the GPU overheating is causing the freezes? Is it just an off chance or is it definitely going to be the problem? I ask because it's not clear to me that the GPU is overheating when the freezes occur. With normal use, it seems to be humming along at 61-2 degrees; it was only when I ran the Furmark test that it started overheating.