9C: Machine Check Exception

[JustinCase]

Hi everyone,

I hope, it is okay to annoy you with my problem. I have been looking around for a while and was convinced that you guys here are that competent that you will be able to solve my problem.

Yesterday, my PC went to bluescreen while using WinXP in a normal way. It said "Machine Check Exception (0x0000009C)" with the following arguments:

Arg1: 00000004

Arg2: 00000000

Arg3: B2000000

Arg4: 00070F0F

Directly after having had this bluescreen, I tried to reboot the PC. It went to the screen where I can choose to use Windows in "Safe Mode" or the normal way. I was not able to choose anything. After the 30 seconds ran down, the machine tried to launch WIndows but got stuck with a black screen.

I wanted to find out whether my keyboard will react to anything so I went to BIOS. All my keyboard input there worked for approximately 30 seconds. Then BIOS was freezing.

After two days with these errors, it was finally possible to access Windows in "Safe Mode" and in normal mode again. So I used Google to find out what my problem could be and received, that this must be a hardware problem. Maybe a problem with my memory. So I downloaded MemTest86 and tested the memory. There, no error was found.

So, this is what I know now: nothing. I would be very glad, if some of yours would have any idea what to do know.

In case it might help, you will find the MiniDump-File attached.

Best wishes and thank you in advance!

Markus aka JustinCase

Mini072709_01.rar

[ripken204]

what happens when you try to boot into normal mode now?

[JustinCase]

Meanwhile everything works fine...

I tested the system with SiSoft SandraLite and there were no errors...but for two days the machine was not accesible at all..

[puntoMX]

I would check connections inside the PC and looking for bad capacitors.

[ripken204]

hmm, odd. but hey, at least it works now.

there may be issues with some piece of hardware still.. but i have no idea what is wrong.

so just hope that it doesn't go down again.

[cluberti]

It appears this is an AMD box, and a 9C means the hardware reported an error, and Windows was obliged to bugcheck with said error (it's how Windows works when an unresolvable hardware error is reported by your computer's CPU). I've discussed 9C errors in depth before, so this time I'll just dump out the analyze analysis of the issue, and try to explain it:

   NOTE:  This is a hardware error.  This error was reported by the CPU
   via Interrupt 18.  This analysis will provide more information about
   the specific error.  Please contact the manufacturer for additional
   information about this error and troubleshooting assistance.

   This error is documented in the following publication:

	  - Bios and Kernel Developers Guid for AMD Athlon(r) 64 and AMD Opteron(r) Processors
   Bit Mask:

	   MA						   Model Specific	   MCA
	O  ID	  Other Information	  Error Code	 Error Code
   VV  SDP ___________|____________ _______|_______ _______|______
   AEUECRC|						|			   |			  |
   LRCNVVC|						|			   |			  |
   ^^^^^^^|						|			   |			  |
	  6		 5		 4		 3		 2		 1
   3210987654321098765432109876543210987654321098765432109876543210
   ----------------------------------------------------------------
   1011001000000000000000000000000000000000000001110000111100001111

Looking closely, it appears that this is what it's saying:

VAL (bit 63) is set to 1, meaning a VALID ERROR was detected

UC  (bit 61) is set to 1, meaning the error is NOT correctable by hardware (so the machine should STOP whatever it's doing)

EN  (bit 60) is set to 1, meaning that error reporting is enabled for this error in the actual MCA control register, so we should report this

PCC (bit 57) is set to 1, meaning the actual state of the CPU may actually be corrupted because of the error condition, and as such whatever was executing should stop to avoid potential data loss (hence why Windows actually bugchecks - this is serious)

Bits 18, 17, and 16 are set to 1, meaning the extended error code came from the northbridge on the motherboard, and it reads 0111 (19 -> 16), which means the error was a BUS error, indicating an error in the HyperTransport or actual DRAM errors

Bits 11, 10, and 09 are reserved, and are ignored

Bit 08 is set to 1, meaning that an error was found by the DRAM scrubber (indicating potential memory problems)

Bits 03 and 02 are reserved, and are ignored

Bit 01 indicates there was an error associated with the CPU 01 core

Bit 00 indicates there was an error associated with the CPU 00 core

This means the DRAM scrubber caught a system watchdog error on the northbridge, meaning bad DRAM (or bad DRAM communications) somewhere on the motherboard. This could be CPU cache, it could be the actual RAM in the system, or it could indeed even be bad capacitors on the motherboard causing enough of a problem to throw the error (although this is rare, it's been known to happen, so I've included it in the possibilities).

Honestly, at this point I'd provide this to the vendor that sold you the machine, because you do have a hardware problem.