AIDA64 hangs on OpenGL extension list with R337.50

[Squall Leonhart]

I'm reporting this here because i know you have contacts in nvidia who can get a fix for this,

 

The same hang occurs in GPU Caps viewer when closing a OGL/OCL test window in thread

 

nvoglv32.DLL!DrvValidateVersion+0x7450

 

and is reported as also affecting Heaven Benchmark.

 

 

The hang is resolved by terminating the instance of  nvoglv32.DLL!DrvValidateVersion+0x7450 with no constant cpu usage.

[Fiery]

It's odd, we haven't found such issues with ForceWare 337.50.  Please let me know what Windows version (and 32-bit or 64-bit) and what video card were you using for the test. GTX680 with 64-bit Win7 SP1 maybe?

[Squall Leonhart]

Yep, system is in signature, Windows is 64bit.

 

it was originally reported as affecting GPU Caps Viewer's OGL and OCL tests.

 

 

@ManuelG
Furthering my post about the OpenGL hang, i have identified that the hang always occurs in a nvoglv32.DLL!DrvValidateVersion+0x7450 thread in Aida64 and GPU Caps Viewer


When GPU caps viewer starts it spawns (TID random obviously)

48436 nvoglv32.DLL!DrvValidateVersion+0x7450

and

185300 nvoglv32.DLL!DrvValidateVersion+0x2040


the 0x2040 instance frequently ends and starts again with a new TID (so nice and clean)


Upon running the CL or Opengl tests a new 0x7450 thread spawns at TID 31400
31400 nvoglv32.DLL!DrvValidateVersion+0x7450


when closing the test window, this thread remains open with no cpu usage instead of terminating properly and the window stops responding. Terminating this thread in Process Explorer returns the application to working order.

The stack contents of the hung thread are

ntoskrnl.exe!KiSwapContext+0x7a
ntoskrnl.exe!KiCommitThreadWait+0x1d2
ntoskrnl.exe!KeWaitForSingleObject+0x19f
ntoskrnl.exe!KiSuspendThread+0x54
ntoskrnl.exe!KiDeliverApc+0x201
ntoskrnl.exe!KiCommitThreadWait+0x3dd
ntoskrnl.exe!KeWaitForSingleObject+0x19f
ntoskrnl.exe!NtWaitForSingleObject+0xde
ntoskrnl.exe!KiSystemServiceCopyEnd+0x13
wow64cpu.dll!CpupSyscallStub+0x9
wow64cpu.dll!Thunk0ArgReloadState+0x23
wow64.dll!RunCpuSimulation+0xa
wow64.dll!Wow64KiUserCallbackDispatcher+0x204
wow64win.dll!whcbfnHkINDWORD+0xb9
ntdll.dll!KiUserCallbackDispatcherContinue
wow64cpu.dll!CpupSyscallStub+0x9
wow64cpu.dll!Thunk0Arg+0x5
wow64.dll!RunCpuSimulation+0xa
wow64.dll!Wow64LdrpInitialize+0x42b
ntdll.dll!??_C@_0BN@KLOBBEB@Enabling?5heap?5debug?5options?6?$AA@FNODOBFM@+0x28204
ntdll.dll!LdrInitializeThunk+0xe
ntdll.dll!_ZwWaitForSingleObject@12+0x15
ntdll.dll!_RtlpWaitOnCriticalSection@8+0x13e
ntdll.dll!_RtlEnterCriticalSection@4+0x150
ntdll.dll!_LdrGetDllHandleEx@20+0x2f7
ntdll.dll!_LdrGetDllHandle@16+0x18
KERNELBASE.dll!_GetModuleHandleForUnicodeString@4+0x22
KERNELBASE.dll!_BasepGetModuleHandleExW@16+0x181
KERNELBASE.dll!_GetModuleHandleW@4+0x29
KERNELBASE.dll!_GetModuleHandleA@4+0x34
D3DOverriderHooks.dll+0x120b
ntdll.dll!_KiUserCallbackDispatcher@12+0x2e
nvoglv32.DLL!DllMain+0x692f
nvoglv32.DLL+0x71015a
nvoglv32.DLL!DrvValidateVersion+0x745c
ntdll.dll!___RtlUserThreadStart@8+0x70
ntdll.dll!__RtlUserThreadStart@8+0x1b


Once response is returned to the process, subsequent runs of the OpenGL or OpenCL tests do not spawn this hanging thread again.

[Squall Leonhart]

I've got a cross system reproduction of the issue now,

 

the missing variable is having Unwinders D3DOverride or RivaTuner Statistics server open.

[Fiery]

Thanks for the info. So the issue only comes up when RTSS is active?

[Squall Leonhart]

Yes, this (reportedly) has only recently been introduced with R337.

 

I can honestly see Alexi(Unwinder) refusing to fix this....

 

and thered be no fix for d3doverrider anyway >.>

 

Adding aida64 and gpu caps viewer to the list of profiles and turning detection/hooking off works around the issue.