BSOD w/ AMD 890FX Chipset

[mbc362]

I recently upgraded my computer and immediately started having stability problems (intermittent POST failures and BSODs in Windows 7 x64). Originally, I believed that they had the same root cause, however I have basically swapped out all the hardware and while I'm no longer getting the POST failures, I'm still getting the BSODs.

I have RMA'd the first CPU I received, ran Memtest86+ on the RAM, replaced the RAM with units specifically listed in my motherboard's QVL, replaced the graphics card (upgraded from Nvidia 8800 GT to ATI HD5870), replaced the motherboard (switched from Asus Crosshair IV to MSI 890FXA-GD70), swapped out the power supply (Corsair HX1000W for my backup BFG MX 550 W), and disabled RAID mode on the SB850 controller (switched to IDE). When I replaced the Asus board with the MSI one, I no longer got the POST failures, however I am still getting the BSODs at roughly the same frequency as before. There does not seem to be any consistency to the BSODs, as I've gotten them during high load, light load, and no load usage.

I'm pretty much at a loss as to explain what is causing the issue. I've attached the BSOD report and system health report; if anyone has any suggestions as to the culprit, I would greatly appreciate it. System specs are as follows:

Windows 7 Ultimate x64
AMD Phenom II 1090T
MSI 890FXA-GD70
2x4GB OCZ Reaper DDR3 1333 OCZ3RPR1333C9LV8GK
XFX Radeon 5870 1GB HD-587X-ZNFV
Corsair HX1000W
WD Velociraptor 150GB (AMD SB850 RAID mode)

Please note that the system health report was made with the BFG MX 550 W power supply in place of the Corsair unit, and I had changed the chipset to IDE mode.

 

[reventon]

Hi,

A few BSODs here - and there appear to be 2 main causes.

Most of the BSODs mention graphics related drivers - these BSODs appear to have occurred all with your ATI card.

The second main cause is to do with your storage - driver mentioned in the oldest BSOD is ahcix64s.sys

This is the AMD AHCI driver. This BSOD could have had something to do with the RAID mode - not entirely sure.

So, the possibility exists that your new Graphics Card is faulty.

Run the following stress test and see if it causes you a BSOD - http://www.sevenforums.com/tutorials/100356-video-card-stress-test-furmark.html

Deal with that first and we will see if there are any storage related BSODs after that is fixed.

Regards,
Reventon

BUGCHECK SUMMARY

[font=lucida console]
Built by: 7600.16617.amd64fre.win7_gdr.100618-1621
Debug session time: Fri Sep 17 13:42:31.353 2010 (GMT+12)
System Uptime: 0 days 14:06:58.696
Probably caused by : ntkrnlmp.exe ( nt!NtRemoveIoCompletion+17c )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0x50
PROCESS_NAME:  plugin-contain
Bugcheck code 00000050
Arguments fffff880`0972e13f 00000000`00000001 fffff800`02f919dc 00000000`00000000
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16617.amd64fre.win7_gdr.100618-1621
Debug session time: Thu Sep 16 12:25:32.918 2010 (GMT+12)
System Uptime: 1 days 0:09:04.260
BugCheck D1, {fffff280098de810, 7, 0, fffff8800422212a}
*** WARNING: Unable to verify timestamp for atikmpag.sys
*** ERROR: Module load completed but symbols could not be loaded for atikmpag.sys
Probably caused by : dxgkrnl.sys ( dxgkrnl!DxgNotifyInterruptCB+6e )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0xD1
PROCESS_NAME:  System
Bugcheck code 000000D1
Arguments fffff280`098de810 00000000`00000007 00000000`00000000 fffff880`0422212a
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16617.amd64fre.win7_gdr.100618-1621
Debug session time: Wed Sep 15 09:33:21.049 2010 (GMT+12)
System Uptime: 0 days 5:33:01.391
Probably caused by : dxgmms1.sys ( dxgmms1!VIDMM_GLOBAL::IsPagingOperationPending+4c )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0x50
PROCESS_NAME:  System
Bugcheck code 00000050
Arguments fffff0a0`01db7b20 00000000`00000000 fffff880`043b0d78 00000000`00000007
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16385.amd64fre.win7_rtm.090713-1255
Debug session time: Tue Sep 14 15:44:31.606 2010 (GMT+12)
System Uptime: 0 days 1:05:25.948
*** WARNING: Unable to verify timestamp for atikmdag.sys
*** ERROR: Module load completed but symbols could not be loaded for atikmdag.sys
Probably caused by : atikmdag.sys ( atikmdag+64208 )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0xA
PROCESS_NAME:  System
Bugcheck code 0000000A
Arguments fffff280`09814af0 00000000`00000002 00000000`00000001 fffff800`02ac7d4f
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.16385.amd64fre.win7_rtm.090713-1255
Debug session time: Tue Sep 14 05:07:54.752 2010 (GMT+12)
System Uptime: 0 days 0:00:10.094
BugCheck D1, {fffff28008397670, a, 0, fffff88001899840}
*** WARNING: Unable to verify timestamp for ahcix64s.sys
*** ERROR: Module load completed but symbols could not be loaded for ahcix64s.sys
Probably caused by : ahcix64s.sys ( ahcix64s+31840 )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
BUGCHECK_STR:  0xD1
PROCESS_NAME:  System
Bugcheck code 000000D1
Arguments fffff280`08397670 00000000`0000000a 00000000`00000000 fffff880`01899840
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``[/font]

 

[mbc362]

Thanks for the response. Unfortunately, I forgot to mention that the BSODs I attached only represent the most recent ones (I forgot to save a lot of the older ones when I reinstalled the operating system). I was also getting the dxgmms1.sys BSODs back when I was using the Nvidia 8800 GT graphics card (thats actually why I had upgraded to the ATI card). I ran the program you suggested just in case; is it supposed to produce an output file? After about 5min my graphics card temp had leveled out at ~72 C and I experienced no artifacts.

 

[reventon]

There are no results - but that is considered a pass.

Your current PSU is the 1000W Corsair right?

I am afraid to say that this does look like your second motherboard is faulty aswell - unlikely, but not impossible.

Have you got the latest BIOS? - if not it would be worth updating just in case.

Why did you switch from RAID mode to IDE? (instead of AHCI?)

 

[mbc362]

The Corsair is my primary PSU, but the last two BSODs occurred with the BFG 550W PSU. The system health report, and the graphics card stability test were run with the BFG PSU. I do have the latest BIOS. I switched from RAID to IDE because I saw that one of the BSODs (which occurred during the win7 install) occurred due to the ahci driver.

 

[reventon]

Ok, replace the Corsair PSU back as it is far superior - obviously the PSU is not the problem though.

For lack of anything better to do (especially seeing as you have already swapped out all components) run Driver Verifier to try and find a software cause (if there is one).

http://www.sevenforums.com/tutorials/101379-driver-verifier-enable-disable.html

Also, it would pay to run the following hardware tests at some point, if you haven't already

http://www.sevenforums.com/tutorials/100352-cpu-stress-test-prime95.html (run the Blend test as well as the Small FFT test)

http://www.sevenforums.com/tutorials/105647-ram-test-memtest86.html

 

[mbc362]

I have run Memtest86+ on the RAM with no errors; I haven't had a chance yet to run Prime95 yet. After I enabled driver verifier I could not get my system to boot. The BSOD didn't list a cause, but the error code I got was 0x000000A. I got three in a row (none of which were logged, apparently). I ended up having to use a restore point to get my system to boot.

 

[reventon]

Ah, 0xA Driver IRQL Not Less or Equal

In other words - a software problem.

Turn driver verifier back on in Safe Mode.

While in Safe Mode save this program to the desktop: http://live.sysinternals.com/autoruns.exe

Right-click on it -> Select run-as-admin.

In Autoruns, go to Options -> make sure Hide Microsoft and Windows Entries is ticked.

Then, press F5 to reload the page. Once reloaded, go to the Driver Tab and untick the boxes beside everything that is in that list.

Then re-boot into normal mode.

Some things will not work as they should, but if it does get into normal mode without BSODing then we have made some progress.

The next step is to use the same program to re-enable some of the entries (this can be done in normal mode) and then re-boot and see if you get the BSODs.

Repeat this procedure until you have isolated the cause.

 

[mbc362]

I followed your directions, re-enabling driver verifier and deselecting all the non-Microsoft/Windows drivers in safe mode and then rebooted into normal mode. The first time I tried booting, I again got a BSOD. The second time, I was able to boot and login, however soon after my computer froze. Given these events (along with the fact that the first BSOD occurred during the Win7 install process), I think its safe to say that my problems are either being caused by a Windows driver, or hardware failure. Any thoughts?

 

[reventon]

Yes - Most likely hardware failure (contrary to popular belief - OS components are pretty reliable)

Make sure you have run all the hardware tests mentioned here - http://www.sevenforums.com/crashes-debugging/111796-bsod-w-amd-890fx-chipset.html#post963401

Just in case - where did you get your OS from? What kind of licence?

 

[mbc362]

I did run the tests with Furmark and Memtest86+. I ran a brief test with Prime95, but I haven't had a chance yet to run for a full 12 hours. I got the OS directly from Microsoft; its a personal license.

 

[reventon]

As in, you got an ISO file from Technet/MSDN (or similar)? If so, did you use DVD or USB to install it?

 

[mbc362]

No, I got a DVD directly from Microsoft.

 

[reventon]

Ok. Let us know how the hardware tests go.

 

[mbc362]

I ran the FFT test in Prime95 for roughly 10 hours and got no failures. An interesting thing has happened though. I reinstalled the OS again, but have not run Windows update or installed any of the hardware drivers (save the chipset RAID driver and the graphics driver), and I'm not longer able to reproduce the BSOD. I can boot up fine with driver verifier enabled, and I have gone 24 hours without a BSOD during use (which is, sadly, unprecedented).

 

[reventon]

Ok. This time - install the critical Windows updates manually one-by-one (or in groups of about 3). Reboot after each and see if you get BSODs. If you do get BSODs - make sure you know exactly what update caused it.

Then - if you get it fully updated without problems install any other necessary drivers one by one too.

 

[mbc362]

I installed several of the chipset drivers, and got a BSOD during boot with driver verifier enabled. After uninstalling one of the drivers, I was able to boot up properly with driver verifier enabled. However, I still got a BSOD during normal use. When I tried reinstalling win 7, I got a BSOD during the install (twice, in fact). I reran Prime95, and this time I did get a failure during the Blend test (all the results text file stated was that it was a fatal error due to rounding inaccuracy). Is the CPU to blame for this? Or could the motherboard/RAM/powersupply also be the cause?

 

[CarlTR6]

Upload those Driver Verifier enabled BSOD's so we can see what they point to.

 

[mbc362]

I've attached the BSOD report. Usbfilter.sys is the driver I had to disable to get my system to boot with driver verifier enabled. The second BSOD listed occurred when I had driver verifier enabled and was able to boot properly.

 

[CarlTR6]

I installed several of the chipset drivers, and got a BSOD during boot with driver verifier enabled. After uninstalling one of the drivers, I was able to boot up properly with driver verifier enabled. However, I still got a BSOD during normal use. When I tried reinstalling win 7, I got a BSOD during the install (twice, in fact). I reran Prime95, and this time I did get a failure during the Blend test (all the results text file stated was that it was a fatal error due to rounding inaccuracy). Is the CPU to blame for this? Or could the motherboard/RAM/powersupply also be the cause?

The short answer to your question is yes, it could be any of the causes you mentioned. Remove all of your RAM but one stick and see how your system runs. The only way I know to test a power supply is to borrow a known good power supply and try it. That fact that you got error running Prime indicates a definitely hardware problem.

Your Driver Verifier enabled dump pints to a Direct X driver as likely not the real cause. I have found that when Direct X is blamed, it can be the video card, the network adapter, or even the sound card as well as hardware problems like RAM or CPU.

Windows 7 is finicky about RAID arrays. I have seen RAID cause all sorts of irritating problems.

Before you do anything drastic, let's look at out of date drivers and update them. In general, third party drivers should be dated July 13, 2009 or later to be up to date for Win 7.

amdsbs.sys Fri Mar 20 14:36:03 2009 - AMD Technology AHCI Compatible Controller Driver for Windows family is a driver file from AMD Technologies Inc. belonging to AMD Technology AHCI Compatible Controller.

ahcix64s.sys Wed Jul 01 03:05:27 2009 - AMD AHCI driver

Rt64win7.sys Thu Feb 26 04:04:13 2009 - Realtek 8101E/8168/8169 NDIS 6.20 64-bit Driver for Realtek 8101E/8168/8169 PCI/PCIe Adapters