BSOD ntoskrnl.exe errors

Jonathan_King · 31 Jan 2011

Well good luck with that, I hope that is the end.

JMH · 31 Jan 2011

rlguildford said:

>
The good news is it's been 24 hours now and the PC is still running - which is a significant milestone! I won't stick my neck out totally, but I'm hopeful that the good advice offered by you guys has significantly resolved my problems.

I'm flabergasted that McAfee appears to be the root cause of so much grief over the past few months - but it's hard to draw any other conclusion.

Many thanks for your advice.
Rob

Rob

Just passing by Rob,
Fingers crossed that your problems have indeed been resolved.
Good luck.

rlguildford · 04 Mar 2011

Darn it. Things were looking stable for a while, but the old problems seem to be back. I've not done anything knowingly to cause this (no software installs etc). In fact it seemed to start when I was away on holiday!

I was unable to start the machine normally this morning, but I managed to start up using Safe Mode (with networking). Anyway, the usual files are attached so hopefully you guys will be able to point me in the right direction again.

Many thanks
Rob

Jonathan_King · 04 Mar 2011

Hello,

Start by installing SP1: Learn how to install Windows 7 Service Pack 1 (SP1)

There is little helpful info in those dumps; I suspect we are looking at a hardware problem. Based on a small piece of circumstantial evidence, I do suggest you update your Broadcom ethernet drivers though. See here: Broadcom.com - Ethernet NIC Driver Downloads

Code:

k57nd60a.sys    Sat Jun 20 07:34:49 2009 (4A3CC959)

I think it's time to check the hardware. Start by running Prime95, specifically the Blend and Small FFTs tests: https://www.sevenforums.com/tutorials...t-prime95.html

Also, run Memtest86: RAM - Test with Memtest86+

...Summary of the dumps:

Code:


Built by: 7600.16695.amd64fre.win7_gdr.101026-1503
Debug session time: Fri Mar  4 02:25:28.502 2011 (UTC - 5:00)
System Uptime: 0 days 0:11:52.002
Probably caused by : memory_corruption ( nt!MiReleaseConfirmedPageFileSpace+87 )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0xA
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_0xA_nt!MiReleaseConfirmedPageFileSpace+87
SystemProductName = Studio XPS 8000
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16695.amd64fre.win7_gdr.101026-1503
Debug session time: Thu Mar  3 12:47:10.790 2011 (UTC - 5:00)
System Uptime: 0 days 0:00:24.961
Probably caused by : ntkrnlmp.exe ( nt! ?? ::FNODOBFM::`string'+21b46 )
BUGCHECK_STR:  0x1a_5003
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
PROCESS_NAME:  MsMpEng.exe
FAILURE_BUCKET_ID:  X64_0x1a_5003_nt!_??_::FNODOBFM::_string_+21b46
SystemProductName = Studio XPS 8000
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16695.amd64fre.win7_gdr.101026-1503
Debug session time: Tue Mar  1 22:00:19.555 2011 (UTC - 5:00)
System Uptime: 0 days 7:34:30.197
Probably caused by : ntkrnlmp.exe ( nt! ?? ::FNODOBFM::`string'+42ba5 )
BUGCHECK_STR:  0x1a_41287
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
PROCESS_NAME:  svchost.exe
FAILURE_BUCKET_ID:  X64_0x1a_41287_nt!_??_::FNODOBFM::_string_+42ba5
SystemProductName = Studio XPS 8000
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16695.amd64fre.win7_gdr.101026-1503
Debug session time: Sun Feb 20 19:11:10.214 2011 (UTC - 5:00)
System Uptime: 1 days 2:51:25.045
Probably caused by : memory_corruption ( nt!MiProtectPrivateMemory+2e1 )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0x50
PROCESS_NAME:  WerFault.exe
FAILURE_BUCKET_ID:  X64_0x50_nt!MiProtectPrivateMemory+2e1
SystemProductName = Studio XPS 8000
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``

rlguildford · 07 Mar 2011

Thanks Jonathan

I fear you're right regarding hardware problem. Previously I thought this unlikely because booting into the diagnostics partition (and running memtest86) seem rock solid.

I'll follow the steps you suggest and update the forum on the results.

Cheers and thanks again
Rob

rlguildford · 11 Mar 2011

Update:

I installed Windows 7 SP1.
During this I had a crash or two, but wasn't put off and the install was eventually successful.
I also downloaded and installed the latest Broadcom drivers for my Ethernet adaptor.

Next I downloaded and ran Prime95 to stress the hardware, at the same time as running RealTemp 3.6.

Prime95 detected some hardware problems immediately (it gave a message about a rounding error on several of the threads). RealTemp showed the CPU temperatore at about 25 degrees C. This all happened so quickly that I wondered if I had done it right, so I tried running the tests again.

Now the CPU temperatrues are at about 70 degrees, and Prime95 hasn't detected any errors for over an hour.

Of course I'll keep Prime95 running for a while longer, and in due course it may detect some more problems. Is it possible that the CPU/Memory etc is less stable when the system is cooler? Assuming this is a hardware problem (which seems pretty likely), what would be the next steps? My PC is a standard build with no over-clocking on my part.

Many thanks for your advice
Rob

rlguildford · 11 Mar 2011

I left Prime95 running, but the PC eventually went to sleep and then wouldn't resume so needed a hard reboot. By this time the CPU was cool again so I reran Prime95 straight away and all threads failed with the following:

[Mar 11 15:31] Worker starting
[Mar 11 15:31] Setting affinity to run worker on logical CPU #1
[Mar 11 15:31] Beginning a continuous self-test to check your computer.
[Mar 11 15:31] Please read stress.txt. Choose Test/Stop to end this test.
[Mar 11 15:31] Test 1, 6500 Lucas-Lehmer iterations of M12451841 using Pentium4 type-2 FFT length 640K, Pass1=640, Pass2=1K.
[Mar 11 15:31] FATAL ERROR: Rounding was 0.5, expected less than 0.4
[Mar 11 15:31] Hardware failure detected, consult stress.txt file.
[Mar 11 15:31] Torture Test completed 0 tests in 0 minutes - 1 errors, 0 warnings.
[Mar 11 15:31] Worker stopped.

So there does appear to be something of a pattern emerging; if the CPU is cool it gives errors, when it's warm it seems more reliable. Does this perhaps indicate a problem with the CPU itself? Can you suggest any guidance for narrowing down what's wrong?

Cheers
Rob

Jonathan_King · 11 Mar 2011

I have found that these "cold" errors are usually caused by the RAM, rather than the CPU. Try leaving the computer off for several hours, and then immediately booting into Memtest86 and running it for maybe 10 or 15 minutes. Memtest86 does not test the CPU worth saying so, so any errors you see are almost certainly caused by RAM.

If errors are found, you may try increasing the RAM voltage slightly. I have seen that as the solution perhaps a couple of times, but out of the hundreds of such cases I've dealt with, that's a very low success rate. Still worth a shot IMO. Typically the problem is fixed by replacing the memory.

If Memtest86 checks out fine on a cold boot, you might try increasing the voltage to the CPU slightly.

rlguildford · 12 Mar 2011

Thanks Jonathan

I Googled around for some hints on how to tweak the voltages, but from what I've read the Dell BIOS doesn't give you any control over this. Do you know if there's a way to do this with a Dell Studio XPS 8000?

I thought I'd start by removing 2GB ofthe 4GB RAM. Not sure how conclusive this will be, but I'll run some tests in this configuration.

Cheers
Rob

Jonathan_King · 12 Mar 2011

No, I suspect you'll have to send the computer in for Dell to repair, or just purchase some new RAM. Many OEMs such as Dell really lock down their BIOS setup programs.

Removing the 2GB is a good thing to try; worth a shot even if it doesn't work.

BSOD ntoskrnl.exe errors

Hardware problems

Futher update