Windows Environment: Windows XP SP2, single 32bit P4 Processor with
hyperthreading disabled
When this happens: intensive stress operation
The custom PCI device is plugged in one of PCI slots. The memory BAR
space for the device is mapped to the user application. My process has
more than a dozen threads. One of thread is used to poll the hardware
state using the memory mapped register space. The other threads may
access the hardware or may not. Sometimes, the process will occupy 100%
CPU usage as indicated in task manager (for example, do
QueryPerformanceCounter in a while loop). CPU fan will speed up. When
this happens, I will get BSOD. The exception state looks like this:
FOLLOWUP_NAME: MachineOwner
SYMBOL_NAME: nt!KdPitchDebugger+ef4
MODULE_NAME: nt
IMAGE_NAME: ntkrnlpa.exe
DEBUG_FLR_IMAGE_TIMESTAMP: 42250a1d
FAILURE_BUCKET_ID: 0x9C_IA32_GenuineIntel_nt!KdPitchDebugger+ef4
BUCKET_ID: 0x9C_IA32_GenuineIntel_nt!KdPitchDebugger+ef4
Followup: MachineOwner
kd> dds @$csp ; kb
e0c27fb0 0000009c
e0c27fb4 00000000
e0c27fb8 e0c27ff0 nt!KdPitchDebugger+0xef4
e0c27fbc a2000000
e0c27fc0 84010400
e0c27fc4 00000000
e0c27fc8 e0c280f4 nt!KdPitchDebugger+0xff8
e0c27fcc e0ba8faf hal!HalpMcaExceptionHandler+0xdd
e0c27fd0 0000009c
e0c27fd4 00000000
e0c27fd8 e0c27ff0 nt!KdPitchDebugger+0xef4
e0c27fdc a2000000
e0c27fe0 84010400
e0c27fe4 00000000
e0c27fe8 00000000
e0c27fec 00000000
e0c27ff0 00000002
e0c27ff4 00000001
e0c27ff8 9a8769b0
e0c27ffc 01c667c1
e0c28000 00000000
e0c28004 00000000
e0c28008 00000000
e0c2800c 00000000
e0c28010 84010400
e0c28014 a2000000
e0c28018 00000000
e0c2801c 00000000
e0c28020 00000000
e0c28024 00000000
e0c28028 00000018
e0c2802c 00000000
ChildEBP RetAddr Args to Child
e0c27fc8 e0ba8faf 0000009c 00000000 e0c27ff0 nt!KeBugCheckEx+0x1b
e0c280f4 e0ba5cae e0042000 00000000 00000000
hal!HalpMcaExceptionHandler+0xdd
e0c280f4 0040ea45 e0042000 00000000 00000000
hal!HalpMcaExceptionHandlerWrapper+0x46
The Error code looks like this (interpreted from Intel Architecture
Software Developer’s Manual V3):
MCA Error: Internal unclassified error
READ/WRITE: READ
Model Specific Error: BQ_ERR_AERR2_TYPE
What does this error mean? How can I deal with it? Should I tell my
manager to destroy the instrument?
Thanks,
joshie