[Okta Webinar] Learn how to a build a cloud-first strategyRegister Now

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1223
  • Last Modified:

Desktop blue screen, event ID 41, Kernel-Power

I have a fairly beefy desktop - even though it's over four year old now.  AMD Phenom II X4 940 on a Gigabyte MA78GM-US2H motherboard with 8GB RAM. I upgraded the system drive to an Intel 520 SSD and now use a very recently purchased 2GB Seagate mechanical drive for storage. I'm running Win8 64bit.

I'm getting recurring crashes which manifest themselves in a variety of ways:

- Mostly, within a short time of boot the machine simply restarts itself (sometimes I get a minidump and sometimes I don't)
-  Occasionally it just goes blank or freezes in the app,  or freezes with a crazy raster pattern
- Even less frequently it does the polite Win8 blue screen with the 'Windows has encountered a problem' and there's a an error message there but it's blurry and goes by too fast (something about ACLs though...)

Once it gets past these problems it's very stable (i.e. runs for days if you don't reboot). The Event Viewer gives the same description for all the crashes i.e. the very unspecific Event ID 41 with the helpful:

 "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly."

Today I got three in a row and with the actual Win 8 blue screen (minidumps attached). I don't read minidumps (I'd love to spend time figuring it out, but I'm supposed to be a writer and should get on with that...)

Prior to October 26 I was running dual boot Win7 and the Win 8 preview (off two different system SSDs). Now it's just Win 8 RTM, BUT I did the upgrade from Win7 and didn't reformat (the upgrade is so smooth and keeps all the settings wonderfully).

Back at my dual boot stage I messed with BIOS and IDE versus AHCI, etc. in order to be able to reset SSDs and get them working. (I'm pretty sure it's now in AHCI mode). I also flashed the BIOS prior to the final Win 8 upgrade.

Any help interpreting the minidumps much appreciated. My sense is that it's more to do with BIOS and drivers than Win8 and was probably lurking in my Win7 config too. One final possible clue is that I sometimes get that 'AMD Data Change...Update New Data to DMI' message on boot and sometimes I don't...
012113-7254-01.dmp
012113-32229-01.dmp
012113-7222-01.dmp
0
chriswor
Asked:
chriswor
  • 4
  • 2
  • 2
2 Solutions
 
joelsplaceCommented:
You need this program:  BlueScreenView  http://www.nirsoft.net/utils/blue_screen_view.html
It will read it and give you the supposed cause of the problem.  It will even do a Google search for you based on the result if you want it to.
0
 
joelsplaceCommented:
Since all three are different I would guess that you have a hardware issue.  Memory wourl be the first suspect.  I like to use Prime95.  It seems to catch more problems than Memtest.
http://files.extremeoverclocking.com/file.php?f=205


012113-7222-01.dmp      1/20/2013 7:12:44 PM      IRQL_NOT_LESS_OR_EQUAL      0x0000000a      00000000`00002002      00000000`00000002      00000000`00000000      fffff801`9f31a8e0      hal.dll      hal.dll+4c500                              x64      ntoskrnl.exe+7b340                              012113-7222-01.dmp      4      15      9200      128,798
      
012113-7254-01.dmp      1/20/2013 7:12:32 PM      PFN_LIST_CORRUPT      0x0000004e      00000000`00000002      00000000`0018c992      00000000`0022ffff      00000000`00000010      parport.sys      parport.sys+5deecc0                              x64      ntoskrnl.exe+7b340                              012113-7254-01.dmp      4      15      9200      126,366      

012113-32229-01.dmp      1/20/2013 7:12:39 PM            0x00000139      00000000`00000003      fffff880`0c7106b0      fffff880`0c710608      00000000`00000000      tdx.sys      tdx.sys+4a1f6b0                              x64      ntoskrnl.exe+7b340                              012113-32229-01.dmp      4      15      9200      126,878
0
 
nobusCommented:
i suggest to download ubcd, and run memtest86+ for testing the ram
best let the system cool for at least an hour, before testing

http://www.ultimatebootcd.com/                              ultimate boot cd
http://ubcd.mirror.fusa.be/ubcd511.iso                        direct link UBCD

then run the disk diag for your disk brand - that way you're sure about the basics

note : it is always best to run with a minimum devices connected for testing - to avoid problems; so disconnect extra devices, and peripherals
0
Independent Software Vendors: We Want Your Opinion

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
chrisworAuthor Commented:
Thanks joelsplace and nobus.  I did try Prime95, though only over an hour or so, and passed all tests on all four workers up to about the 4th level. I also tried BlueScreenView over my collection of minidumps since October. Certainly variety there, but ntoskrnl.exe was always in the mix somewhere.

So I started researching ntoskrnl.exe which led to memtest and disk diagnostics (as per nobus's subsequent recommendation). The Intel SSD was fine, but two instances of memtest, each set on 2GB, threw up two lots of memory copy errors in just 10 minutes of running.

I pulled out one of the four memory modules, and made sure the remaining three were all firmly seated, and ran the test again for an hour with no issues. It seemed unlikely that I'd pulled exactly the right misbehaving module, so I'm thinking the problem was all down to a poorly seated memory module or two.

By this time it was late and the nobus advice had just come in. I replaced the module I'd removed and disconnected all non-essential peripherals. Didn't have time to let it cool for an hour and didn't use the bootable memtest, but did leave it running all night with three instances of memtest (76% CPU and 86% memory). There were no errors in the morning.

So far I've been running for an hour after reboot with no interruption. It does look like the memory modules just weren't connecting as well as they used to and a bit of pushing and prodding has them talking properly again. I'll reconnect my peripherals and give it a couple of days then close this question if the machine has behaved nicely.

Thanks again
0
 
joelsplaceCommented:
Since you got errors in memory I'll bet everything else is fine.
0
 
nobusCommented:
your conclusion seems correct -  a bad contact in the slots
i would take out all sticks, and blow out the slots + verify the sticks their contacts, before putting them in
0
 
chrisworAuthor Commented:
This desktop hasn't missed a beat in two days despite several reboots and heavy use of all my usual apps. It was just the memory module connections.

I haven't done the recommended blowing out and housekeeping yet. At four years old it's probably time for my customary new CPU and new motherboard rebuild (I never used to last more than 2.5 years or so). But at a Windows Experience Index of 6.7, this one is still ahead of many (most) PCs on sale at retail. Hard to justify buying new toys...

Thanks for the advice.
0
 
joelsplaceCommented:
Glad it's working.  I'm with you on the upgrade.  My main PC is about the same as yours.
0

Featured Post

VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

  • 4
  • 2
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now