[SOLVED] Various BSODs

[AceInfinity]

If it returned errors at any point, that's a good enough indication that corruption is going on.

 

[Cl0ttERS]

HOWEVER with that in mind, the BSOD frequency has dramatically decreased, so already you have helped :)

It looks like its RAM, but I don't understand how- The motherboard is new, the RAM was new with it. So why is it failing?

 

[Cl0ttERS]

Oh yeah that reminds me-

I get incredible lag (As in choppy sound/mouse lag/video lag) when there is some sort of network change - Disconnecting WiFi/Starting virtual network/anything like that

About 5 processes of svchost spike at CPU when it happens. Any explanation for this? I think my network card is dodgy - or perhaps Microsoft Security Essentials... That also spikes during network change

 

[AceInfinity]

Try running this command in an instance of cmd prompt and paste the content that you see in the notepad within your next reply:

sc query "WinHttpAutoProxySvc" > 0 & notepad 0 & del /f/q 0

I'll let Wrench97 focus on your BSOD's, but for the CPU spiking, I know some people have this issue with this particular service. You can stop it and set it to Manual to test if it's currently active.

 

[Cl0ttERS]

Try running this command in an instance of cmd prompt and paste the content that you see in the notepad within your next reply:

sc query "WinHttpAutoProxySvc" > 0 & notepad 0 & del /f/q 0

I'll let Wrench97 focus on your BSOD's, but for the CPU spiking, I know some people have this issue with this particular service. You can stop it and set it to Manual to test if it's currently active.

Using your command outputted nothing - I had to change it to

sc query "WinHttpAutoProxySvc" > 0 & notepad 0 & del /f/q C:/Users/SacredSkull/Desktop/output.txt

Even though I guessed the syntax incorrectly, it actually responded with something

SERVICE_NAME: WinHttpAutoProxySvc 
        TYPE               : 20  WIN32_SHARE_PROCESS  
        STATE              : 1  STOPPED 
        WIN32_EXIT_CODE    : 0  (0x0)
        SERVICE_EXIT_CODE  : 0  (0x0)
        CHECKPOINT         : 0x0
        WAIT_HINT          : 0x0

 

[Wrench97]

Any errors in memtest indicates there's a issue there, if you want try removing 1 stick and rerun to see if you get any errors run 6 passes or until you see an error then swap sticks, if they both test good in the first slot test them in the second slot to see if 1, both or neither test bad there.

 

[Laxer]

If you installed the OS with bad ram it can cause all sorts of weird issues... I fought with something similar when I first built my machine.

As everyone has suggested above, run each stick through a full test individually to make sure they are good...

From there you can test to make sure all the slots are good by trying different configurations for the ram.

 

[usasma]

Just my 2¢ (I'm late to the party!)

Fix bad RAM first - then see if issues remain.

Latest version of GEARAspiWDM.sys in iTunes is v2.02.00.01 and it dates from 5/18/2009 (from my system)
BUT, the driver can be the correct version and correct date - and still be corrupted.
Replacing it with a freshly downloaded copy is always the safest thing to do.

We frequently see disabled drivers in memory dumps. I suggest:
- the best thing is to remove the driver software and then physically remove the device from the system.
- if you can't do that, then remove the driver software and then disable it in the BIOS
- if you can't do that, then - enable the device, replace the drivers with a freshly downloaded copy, then re-disable the device if so desired.
- if you can't do that, then please see my sig for further suggestions :0) (just kidding!)

Just FYI -
Back last fall a poster in another forum stated that Asus tech support had advised them to set their memory timings/settings/voltages in accordance with the RAM manufacturer's recommendations - and to NOT use the Auto function in the BIOS. They stated that there was a memory problem with the M4 series of motherboards.

Keep an eye on the Comodo stuff. It's a very complicated application and can cause memory access issues if it gets corrupted.

You have the AMD OverDrive driver installed (AODDriver2.sys) - which is known to cause BSOD's some Win7 systems. Please start with this:

AMD OverDrive (AODDriver2.sys) is either a stand-alone application, or a component of the AMD VISION Engine Control Center. This driver is known to cause BSOD's on some Win7 systems.

Please uninstall all AMD/ATI video stuff from Control Panel...Programs...Uninstall a program
Then, download and install a fresh copy of the ATI drivers from http://www.ati.com (in the upper right corner of the page)
- Select the "Custom" installation option and don't let it install AMD OverDrive or the VISION Engine Control Center
Then use the ATI uninstall manager to uninstall the Catalyst Control Center (CCC).

If the device (AODDriver or AODDriver4.01) remains a problem in perfmon /report, open Device Manager, select the "View" item.
Then select "Show hidden devices" and scroll down to the Non-Plug and Play Drivers section.
Locate the AODDriver entry, right click on it and select "Uninstall". Reboot for changes to take affect.

Sometimes the driver remains and continues to cause BSOD's. If this is the case for you, post back and we'll give further instructions for safely removing it.

If overclocking, please stop. Remove the overclock and return the system to stock/standard values while we're troubleshooting. Once the system is stable again, feel free to resume the overclocking.

 

[Cl0ttERS]

Would incorrect timings cause memory errors in memtest86 ?

And secondly Comodo corrupted itself, and wouldn't uninstall properly, I have used the community made cleaning tool, but I'm not entirely sure if it is gone

 

[Wrench97]

Yes it's possible for timings and/or voltages to cause the errors.

 

[Cl0ttERS]

I'd be interested to know more about how RAM actually gets damaged and how incorrect values affect its performance/ability to store information correctly, any links? :)

 

[Wrench97]

Tom's Hardware has a good article, it's a couple years old but covers the basics> http://www.tomshardware.com/reviews/pc-memory,1698-2.html

Damaged is another story, I would not call it damaged necessarily more a matter of a failed component or one that was not 100% in the first place and deteriorated with use. Damaged ram generally comes from over volting or sometimes voltage spikes and usually effects the entire stick.

 

[Cl0ttERS]

Oh yeah - One thing I have never understood is how to achieve XMP timings on AMD systems (If that makes sense) - From what I understand XMP sets the max timings up automatically, and often RAM is displayed with their XMP timing, can you get to these values with AMD processors?

 

[Wrench97]

Simple answer you can't XMP(Intel® Extreme Memory Profile) is a Intel standard, AMD had BEMP(Black Edition Memory Profiles), although you still see some advertised AMD never added the automatic settings to the bios like Intel did. On a Intel chipset retail performance board you can select XMP settings and it automatically raises the voltage and sets the timings to the XMP standard.

 

[Cl0ttERS]

Ah ok, so its just a preset really, and you can achieve them manually?

 

[Wrench97]

Yes that's correct.

 

[Laxer]

If you go into BIOS you should be able to modify the voltage, frequency and timing of your ram...

I have not heard of issues from timing/voltages with memtest... Regardless I would stand behind wrench's comment...

If you know the ideal settings for your RAM you can change them... If not just try default: 1.5V, 9-9-9-24 timing

 

[jcgriff2]

About 5 processes of svchost spike at CPU when it happens. Any explanation for this?

For info, svchost = Service Host - each runs multiple system services

It's not uncommon to have 12+ svchost running at one time

 

[Cl0ttERS]

I know they are seperate - Thats not the question I'm asking... I want to know why it causes a massive CPU spike when theres any change in network connectivity

Please understand - I am no noob - I have built my computer, so I know how to use the BIOS, change timings, ect. So don't hold out on technical terms. It's interesting how I've seen so many BSODs and so few kernel panics in linux.

 

[Vir Gnarus]

Different OS structures and exception handling are responsible for the difference between amount of BSODS in Windows and that in Linux. If I understand correctly, because of the modular construction of Linux, it's not as sensitive to certain tampering within the kernel as Windows is. Windows also is a lot more "robust" and active compared to typical Linux kernels, which can also be a reason. There's nothing wrong with any of that, of course, it's just a different approach in OS design.

the Svchost spiking can be caused by anything, as svchost is - obviously - a host process for various services and whatnot. You can check which services run under a particular Svchost process by dbl-clicking the process in Process Explorer and then going to the Services tab. You can also use the Threads tab to see the spiking threads and their associated service (if available) and even dbl-click the thread to get its stack for more granular details on what's going on. As a tip, you can press Spacebar at any time to freeze/unfreeze all currently visible data (this does not affect new windows opened and will only show empty data on em).

Process Monitor is also a good application to diagnose strange activity. The log grows quick so it's best to take small snapshots just enough to witness the spike activity.