BSOD 0x00000124

[Strawhat12]

Hi,

For some reason I get the BSOD stop 0x00000124 error message randomly whether my PC is idle or its nearing 100% usage. From the BSOD this is the full error;

stop 0x00000124 (0x0000000000000000, 0xfffffa8002a8f038, 0x00000000b2000040, 0x0000000000000800

I've gathered that this means its a hardware related error. I did have my CPU overclocked recently and I think I must have first got the message then but now everything is back at default, stock, auto and yet I still get the error.

I ran memtest and memtest86+ for 7 passes each with no errors.

So what could it be?

[cluberti]

Without an actual kernel dump it'll be hard to say WHAT failed, but yes, this is a hardware error reported by the CPU (not necessarily the CPU failing, but it ultimately does the error reporting to the OS).

[Strawhat12]

Do you want me to upload the minidump?

[cluberti]

The minidump will not contain the memory information that has the WHEA data, it will only contain enough data to display the stack and registers (and the WHEA data would not be included). You *need* a kernel dump to analyze a 0x124 dump, so if all you're currently configured for is a minidump, please change it and upload the kernel dump when the problem re-occurs.

[Strawhat12]

So had the PC BSOD on me again but I managed to upload a Kernel Dump of it. I did have a look at the dump myself using Windbg but I'd like to know what it all means.

Edited November 19, 2009 by Strawhat12

[cluberti]

First, please note this is not a kernel dump, it's a minidump:

"Mini Kernel Dump File: Only registers and stack trace are available"

However, you're in luck, there might be enough here to make a determination of where to look. It looks like you do have a 3GHz Intel Core2, and it did spit out a WHEA record into the minidump:

1: kd> !cpuinfo
CP  F/M/S Manufacturer  MHz PRCB Signature	MSR 8B Signature Features
 1  6,23,6 GenuineIntel 3000 0000060c00000000				   211b3ffe
					  Cached Update Signature 0000060c00000000
					 Initial Update Signature 0000060c00000000

1: kd> !whea
Error Source Table @ fffff80003019b38
0 Error Sources

That means the high 32 bits is 00000000b2000040 and the low 32bits is 0000000000000800, which equates to the following binary:

1011 0010 0000 0000 0000 0000 0100 0000 | 0000 0000 0000 0000 0000 1000 0000 0000

Taking this and putting it against the Intel model-specific MCA bitmask, you get:

   6		 5		 4		 3		 2		 1
3210987654321098765432109876543210987654321098765432109876543210
----------------------------------------------------------------
1011001000000000000000000100000000000000000000000000100000000000

This means, according to the intel processor documentation:

VAL - MCi_STATUS register is valid
Indicates that the information contained within the IA32_MCi_STATUS register is valid. When this flag is set, the processor follows the rules given for the OVER flag in the IA32_MCi_STATUS register when overwriting previously valid entries. The processor sets the VAL flag and software is responsible for clearing it.

UC - Error Uncorrected
Indicates that the processor did not or was not able to correct the error condition. When clear, this flag indicates that the processor was able to correct the error condition.

EN - Error Enabled
Indicates that the error was enabled by the associated EEj bit of the IA32_MCi_CTL register.

PCC - Processor Context Corrupt
Indicates that the state of the processor might have been corrupted by the error condition detected and that reliable restarting of the processor may not be possible.

Ultimately, this means the likelihood of failure rests in the processor socket somewhere - it could be the core itself, cache memory that is bad, etc - but ultimately, you have a bad CPU and probably fried something during the overclock enough to make it unstable. I'd suggest contacting Intel with this information and getting it replaced under warranty, if you still have warranty on the part.

[Strawhat12]

Thanks for the response. I suspected the CPU would be the problem.

You said that the file I provided wasn't a Kernel dump but a minidump. Well is there a different way of selecting a Kernel dump that I'm not aware of? I right-clicked 'My Computer' then 'Properties' and then 'Advanced Settings'. From there I selected 'Startup and Recovery' and from the drop down box there are three options; 'Small memory dump 256kb', Kernel dump', 'complete memory dump'. I obviously selected 'Kernel dump'. Is this not the way I should be doing it?

[Strawhat12]

Also one more question;

I checked and my processor is within the 3 year warranty period that Intel offer. If I was to contact them would it be best not to tell them it was overclocked? Does that void warranty? Even if I don't mention about the OC can they still find out somehow? Also, if I mention I get the BSOD randomly telling me its a hardware error they may say it could be to do with other components when I actually know its the processor via the minidump. Do I mention this to them?

[cluberti]

Thanks for the response. I suspected the CPU would be the problem.

You said that the file I provided wasn't a Kernel dump but a minidump. Well is there a different way of selecting a Kernel dump that I'm not aware of? I right-clicked 'My Computer' then 'Properties' and then 'Advanced Settings'. From there I selected 'Startup and Recovery' and from the drop down box there are three options; 'Small memory dump 256kb', Kernel dump', 'complete memory dump'. I obviously selected 'Kernel dump'. Is this not the way I should be doing it?

The change only takes effect on reboot, so either you hadn't rebooted yet, or it didn't stick. That's the right choice, but that's not the dump type you ended up with after the crash. As to the warranty repair, you can always point to them your dump and state that the other components in the machine test fine, and the machine only crashes when the CPU runs at 100% for a period of time. I wouldn't necessarily offer any more information than you have to, as overclocking may have voided the warranty, I'm not sure. However, if you simply state the facts of the crashing (CPU at 100%, has stock heatsink/fan attached and working properly, all other components have tested out properly under load) without offering what you did before the crash, you'll probably be fine. I doubt they'll ask if you've overclocked it, and there's no "black box" in your CPU that would indicate you've overclocked it, either, so there's no way they'd know if you didn't tell them.

[puntoMX]

I wouldn't necessarily offer any more information than you have to, as overclocking may have voided the warranty, I'm not sure.

On CPUs it will void the warranty, but it will be hard time for Intel to find out if it was OCed or not, and for just a simple CPU they won't do much to test it and just swap the part for another.

To the OP:

Note that in some countries you have to return the CPU with the cooler that came with it. If it was a tray version (no retail, no cooler included from Intel) you only had 1 year of warranty in most countries.

I also like to say that it's not common that a CPU dies on you, most of the time it's the motherboard, if possible I would try another CPU or that "defective" CPU in another motherboard. Either way, Intel probably will replace it when you send it to them, but it will cost you some shipping and time and still have a bad motherboard at the end .