[Bug 8427] CPUZ fails to load

[Wine Bugs]

http://bugs.winehq.org/show_bug.cgi?id=8427

------- Additional Comments From [e-mail]  2007-18-05 06:24 -------
Hello,

ok, seems another bug (leave the non working kernel mode driver stuff aside)...

The first exception seen in trace is actually a nifty trick to detect if running
inside a virtual machine.

--- snip trace ---
000f:trace:seh:raise_exception code=c000001d flags=0 addr=0x4731a9
000f:trace:seh:raise_exception  eax=00000001 ebx=00000000 ecx=1d1dce56
edx=0061bd6d esi=0061bd28 edi=00610e20
000f:trace:seh:raise_exception  ebp=0034f534 esp=0034f4fc cs=0073 ds=007b
es=007b fs=0033 gs=003b flags=00010202
000f:trace:seh:call_stack_handlers calling handler at 0x461400 code=c000001d flags=0
--- snip trace ---

--- snip disasm ---
0047319E   . 53             PUSH EBX
0047319F   . BB 00000000    MOV EBX,0
004731A4   . B8 01000000    MOV EAX,1
004731A9     0F             DB 0F         ; causes invalid opcode exception
004731AA     3F             DB 3F                                   
004731AB     07             DB 07
004731AC     0B             DB 0B
004731AD   . 85DB           TEST EBX,EBX
004731AF   . 0F9445 E7      SETE BYTE PTR SS:[EBP-19]
004731B3   . 5B             POP EBX
-- snip disasm ---

Basically said, the app sets up SEH (try/catch blocks) and special VM parameters
(ebx, eax values) and then executes invalid opcode.
When running inside an VM this opcode is gracefully handled, while not running
in VM SEH is executed.

The second exception (c0000005) is the real one.
Seems there is no API involed regarding this case.
The directx issue you are seeing is probably another one, not directly related
with this crash.

Looks like it tries to access some class/object member data (vtable -> c++)
which is not present (ecx/this).
Lots of vtable indirections, indicating some kind of class/object hierarchy.
The missing member data could be caused much earlier thats why this thing is
hard to track down (speak: blame some API call for it).

Regards