Run Burnin Test 5.3 Pro (1014) 2D & 3D test cause Windows Blue screen error

[Harry Wang]

Hi,
I have a BSOD (Windows Blue Screen of Death) error when I run the burnin test 5.3 2D and 3D test.
The BSOD Code: 0000008E (80000004, 806EBAD1, A9F53708, 00000000)
But this problem only happen when I turn on the windows screen saver!!
If I did not turn on the screen saver (cause display always ON),
this error will not happen.
Is this a software conflict for screen saver and burnin test?

System Config:
OS: Windows XP SP2 English
CPU: Intel Core Duo L2400 1.66GHz
MCH: Intel 945GME (Internal Graphics)
ICH: Intel ICH7M
RAM: DDR2-667 SODIMM 1GB
Graphics Driver: Intel IGP 6.14.10.4906
Or Newest driver form Intel website: Intel IGP 6.14.10.4926
DirectX: 9.0C (4.09.0000.0904)
Burnin Test: 5.3 (1014)

Test Setup:
1. Burnin Test 2D Test 100%
2. Burnin Test 3D Test 100%
3. Test time: Infinity
4. Turn on Windows screen saver (Blank, active after 10 mins)

[Harry Wang]

Update the test temperature and test result:

Test temperature:
25C, with extra cooling fan

Test result:
In most condition, this system will BSOD error in 3 hours burnin test.

[David (PassMark)]

Error 0x0000008E means,
KERNEL_MODE_EXCEPTION_NOT_HANDLED
The KERNEL_MODE_EXCEPTION_NOT_HANDLED bug check has a value of 0x0000008E. This bug check indicates that a kernel-mode application generated an exception that the error handler did not catch.

The 4 parameters 0x80000004, 0x806EBAD1, 0xA9F53708, 0x00000000 mean the following.

0x80000004
The exception code that was not handled. In this case the except was, STATUS_SINGLE_STEP.

0x806EBAD1
The address where the exception occurred. The high address indicates a device driver

0xA9F53708
The trap frame

Single-stepping on an x86 processor is a matter of setting the trace flag (value 0x100) in the CPU's EFlags register. When the trace flag is set, the CPU executes just one instruction, then generates a STATUS_SINGLE_STEP exception. After receiving the STATUS_SINGLE_STEP exception, the trace flag can be turned off to resume normal execution. If the driver enabling single stepping, and then not handling the exception? Would seem to be a driver bug.

Funnily enough it turns out we already analysed this problem in the past and concluded it was a device driver bug in pci.sys, probably 38 bytes into a function called PciReadWriteConfigSpace.

Of course pci.sys is not our software, so we can't actually fix the problem.