Solved BSOD - Several Error Types pointing to ntoskrnl.exe

[Arc]

the BSODs in games I get screen flickering because Windows reports GPU timeout and trying to recover, and after 3-4 attempts it goes to BSOD. Once I lower the GPU clocks these problems disappear so guess it's either the GPU, PSU or the PCIE bus. But a stress test should give a hint if there's a problem, shouldn't it? :/

Something like this? A balloon from the systrey?

If so, it is the GPU. But furmark will not be able to catch it.

I am also a victim of this, with my earlier GPU

 

[v4l33]

Yes, exactly that. But like I said the problem goes away once I lower the default factory OC clocks, so guess it's related to the OC/PSU. The question remains, could it be the cause to the ntoskrnl.exe BSODs during idle/low load after I lower the GPU clocks? I mean can a combination of GPU core/memory frequencies lead to system instability, even tho it supports higher clocks?
E.g.
HD 7770 default AMD clocks - 1000/1125
HD 7770 default Sapphire OC clocks - 1100/1250 <- video driver timeout
HD 7770 OC lowered clocks - 1050/1200 <- no video driver timeout, possible ntoskrnl.exe BSOD cause?

 

[Arc]

Factory overclocks, as it is said by our overclocking team, are designed to bear that overclocking.

If the issue goes away with reducing the magnitude of the OC, thats a good news.

Though I doubt that the stability will be carried forward for long, but my best wishes will be with you.

I am not very experienced with overclocking, but our overclocking team is the authority. I am requesting them to suggest you here.

 

[v4l33]

That's what I thought also, and having the PC holding up during Prime95, Furmark and OCCT full load stress tests I suppose it rules out a bad PSU also since the power drainage is maxed during those tests.
I don't know what else to do, guess I'll have to bear with these crashes until I change my PC or the faulty hardware fails for good. :/
Anyway, thanks a lot for the support, learned alot during this debugging.

 

[v4l33]

Ok so got another BSOD, see attachment, and I think this one is a lucky one.
This is what WinDbg says:

STACK_COMMAND:  kb
SYMBOL_NAME:  ZEROED_PAGE_CORRUPTED
FOLLOWUP_NAME:  MachineOwner
MODULE_NAME: hardware
IMAGE_NAME:  hardware_ram
DEBUG_FLR_IMAGE_TIMESTAMP:  0
FAILURE_BUCKET_ID:  X64_0x1a_8887_ZEROED_PAGE_CORRUPTED
BUCKET_ID:  X64_0x1a_8887_ZEROED_PAGE_CORRUPTED
Followup: MachineOwner

So it's safe to assume that the memory is faulty afterall, even if memtest86 passed succesfully?

 

[Britton30]

The debug files really make no distinction between types of memory Although yours says hardware_ram, it could be memory in the GPU or CPU even, the different cache levels. It may possibly also be the hard drive cache, I'm not sure.

 

[v4l33]

Dang it. So back to square one I guess. I'll run memtest86 again for the 3rd time overnight and see what results I'll get.

 

[Britton30]

Swap your RAM modules to different slots to see if still no errors.

 

[v4l33]

Did that already a week ago or so, wanted to test maybe the motherboard slots are faulty, sadly still getting crashes.

 

[Britton30]

Ok, sorry I can't help you. Back to you Arc.

 

[v4l33]

Nothing to be sorry about, I'll take any help or advice I can get atm.

 

[ICIT2LOL]

V my 2 cents worth mate try this it may turn out to be a a non event but while reseating stuff is good I always clean as well see this

SLOTCLEANING

Now as for the slots to clean them use astrip of old credit card about 1/2 inch wide nip the corners off one end - preferably round them off hold apiece of thin lint free cloth over the end dab on some form of alcohol - isopropyl alcohol (what we swab the skin with before an injection) is good and GENTLY swipe along the slots.

This can also be done for the GPU slot/s.

 

[v4l33]

I'll try that as soon as I get my hands on some isopropyl, tho the PC is about 6 months old, everything still shines like new, barely any dust on the components. Tried moving the video card on another PCIE slot but couldn't fit it since the card is too tick and hits the bottom of the case.
Tried lowering the clocks a little again for the GPU and 5 mins later I got a IRQ_NOT_LESS_THAN_EQUAL BSOD, so now I lowered them down to AMD stock clocks and see how it goes. If I won't get any crash anymore means the card is faulty and will have to RMA it.

 

[Arc]

If I won't get any crash anymore means the card is faulty and will have to RMA it.

That's the decision.

 

[ICIT2LOL]

I'll try that as soon as I get my hands on some isopropyl, tho the PC is about 6 months old, everything still shines like new, barely any dust on the components. Tried moving the video card on another PCIE slot but couldn't fit it since the card is too tick and hits the bottom of the case.
Tried lowering the clocks a little again for the GPU and 5 mins later I got a IRQ_NOT_LESS_THAN_EQUAL BSOD, so now I lowered them down to AMD stock clocks and see how it goes. If I won't get any crash anymore means the card is faulty and will have to RMA it.

Ok mate no need to clean then any alcohol will do I often use the methylated sort as well but if the thing is bad then do as you are going to do.

 

[v4l33]

It's been a week now with stock AMD clocks on my GPU and no BSOD of any kind whatsoever. Left the PC idle everytime I got out and once overnight trying to trigger the BSODs that plagued me but couldn't get to crash it again.
So it seems the GPU factory OC was causing them, I'm gonna return it as soon as I get a replacement since you never know if it's gonna fail with standard clocks too.
This was quite a journey to pinpoint the culprit, maybe Microsoft needs to improve the exception handling in the core a lil bit more and try to get more details about what went wrong in the future.
A big thanks for everyone who helped me, much appreciated. Hopefully I won't see you again.

 

[Arc]

So it seems the GPU factory OC was causing them, I'm gonna return it as soon as I get a replacement since you never know if it's gonna fail with standard clocks too.

As soon as possible That will be the best. Because, it is supposed that the factory OC should be able to carry the OC.

 

[v4l33]

It's been a while but I thought this is important to mention in case someone else have similar issues.

So in the past I've been plagued with random BSODs, a small minority of them being due to a faulty GPU which I replaced in the meantime. The BSODs due to the faulty GPU were easily reproduced and had an exact error code but the rest of the BSODs were unexplained for.

After months of searching, debugging and trying literally anything I thought about, even started slowly replacing every part of the PC, I finally found the culprit.
Apparently some AsRock motherboards do not get well with Windows if you have enabled SATA Agressive Link Power Management in your BIOS. Just searching on Google for both, rends hundreds of results with people having the same issues with the same motherboard, and the same solution.

So if you are experiencing unexplained and random BSODs it's worth trying to disable the feature in your BIOS and see how it goes.