VMware Communities
dopefish
Contributor
Contributor

Workstation 8 causing BSOD on multiple VMs using Direct3D

Hey,

I recently purchased vmware workstation 8 and now i got a pretty nasty problem.

Im using Win7 x64 as my host as well as Win7 x64 systems as my guests. All guests are running medium hardware consuming D3D applications.

So now i can run 2 or 3 guests from my SSD without any problems, but when i go up to 4 or 5 the host keeps on crashing  with a BSOD.

If i let the applications run on my host i can easily start 10 applications at the same time without a crash, also i already stress tested my graphics card to death no crashes and when crashes in vmware appear, the vga card is at like 20% load and 50°C.

i have all latest drivers and vmware tools installed.

I have tryed several fixes suggested by internet users.

- Installing about 10 different ati drivers

- Installing ati drivers without CCC

- Using MSI Afterburner to reduce the clocks of my gfx card - resulted in instant crash

- Reinstalling win7 from scrath, just adding drivers and starting vmware

- Reinstalling my guests from scratch

I think this error must have to do something with how vmware is passing over the direct3d devices but im not experienced in vmware.

So heres my host setup:

- Intel® Core™ i7-3770K CPU @ 3.50GHz
- Ram: 32 GB Kingston ValueRam 1333Mhz
- Asus AMD Radeon HD 6950 Direct CU II 2048MB
- Crucial M4 SSD 128 GB

- Asrock Z77 Extreme4 Board

For the guests i uses the following setup (tryed different setups with same result anyway):

- Ram: 2GB

- HDD: 40GB

- Lan: Bridge

- CPU-Cores: 2

I also disabled all non used stuff on the guests, disabled: USB, Floppy, CDROM, Printer, etc. basicly it just has land, gfx, hdd and cpu

When running multiple VMs cpu is not stressed at all it goes to 30% or so.

Any help would be highly appreciated,

Thanks in advance,

Lukas

p.s. here are some crashdumps:

Probably caused by : atikmpag.sys ( atikmpag+9250 )

Followup: MachineOwner
---------

6: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

VIDEO_TDR_FAILURE (116)
Attempt to reset the display driver and recover from timeout failed.
Arguments:
Arg1: fffffa801e2874e0, Optional pointer to internal TDR recovery context (TDR_RECOVERY_CONTEXT).
Arg2: fffff88007809250, The pointer into responsible device driver module (e.g. owner tag).
Arg3: 0000000000000000, Optional error code (NTSTATUS) of the last failed operation.
Arg4: 000000000000000c, Optional internal context dependent data.

Debugging Details:
------------------

ADDITIONAL_DEBUG_TEXT: 
You can run '.symfix; .reload' to try to fix the symbol path and load symbols.

FAULTING_MODULE: fffff80002c00000 nt

DEBUG_FLR_IMAGE_TIMESTAMP:  4fd61c26

FAULTING_IP:
atikmpag+9250
fffff880`07809250 4055            push    rbp

DEFAULT_BUCKET_ID:  GRAPHICS_DRIVER_TDR_FAULT

CUSTOMER_CRASH_COUNT:  1

BUGCHECK_STR:  0x116

CURRENT_IRQL:  0

STACK_TEXT: 
fffff880`0270fff8 fffff880`0ba85000 : 00000000`00000116 fffffa80`1e2874e0 fffff880`07809250 00000000`00000000 : nt+0x7f1c0
fffff880`02710000 00000000`00000116 : fffffa80`1e2874e0 fffff880`07809250 00000000`00000000 00000000`0000000c : dxgkrnl+0x5d000
fffff880`02710008 fffffa80`1e2874e0 : fffff880`07809250 00000000`00000000 00000000`0000000c 00000000`00000000 : 0x116
fffff880`02710010 fffff880`07809250 : 00000000`00000000 00000000`0000000c 00000000`00000000 00000000`00000000 : 0xfffffa80`1e2874e0
fffff880`02710018 00000000`00000000 : 00000000`0000000c 00000000`00000000 00000000`00000000 fffff880`0ba84c7c : atikmpag+0x9250


STACK_COMMAND:  kb

FOLLOWUP_IP:
atikmpag+9250
fffff880`07809250 4055            push    rbp

SYMBOL_STACK_INDEX:  4

SYMBOL_NAME:  atikmpag+9250

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: atikmpag

IMAGE_NAME:  atikmpag.sys

BUCKET_ID:  WRONG_SYMBOLS

Followup: MachineOwner
---------

Probably caused by : atikmpag.sys ( atikmpag+9250 )

Followup: MachineOwner
---------

7: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

VIDEO_TDR_FAILURE (116)
Attempt to reset the display driver and recover from timeout failed.
Arguments:
Arg1: fffffa801e06b4e0, Optional pointer to internal TDR recovery context (TDR_RECOVERY_CONTEXT).
Arg2: fffff8800795e250, The pointer into responsible device driver module (e.g. owner tag).
Arg3: 0000000000000000, Optional error code (NTSTATUS) of the last failed operation.
Arg4: 000000000000000c, Optional internal context dependent data.

Debugging Details:
------------------

Followup: MachineOwner

---------

ADDITIONAL_DEBUG_TEXT: 
You can run '.symfix; .reload' to try to fix the symbol path and load symbols.

FAULTING_MODULE: fffff80002c1a000 nt

DEBUG_FLR_IMAGE_TIMESTAMP:  4fd61c26

FAULTING_IP:
atikmpag+9250
fffff880`0795e250 4055            push    rbp

DEFAULT_BUCKET_ID:  GRAPHICS_DRIVER_TDR_FAULT

CUSTOMER_CRASH_COUNT:  1

BUGCHECK_STR:  0x116

CURRENT_IRQL:  0

STACK_TEXT: 
fffff880`0270cff8 fffff880`09137000 : 00000000`00000116 fffffa80`1e06b4e0 fffff880`0795e250 00000000`00000000 : nt+0x7f1c0
fffff880`0270d000 00000000`00000116 : fffffa80`1e06b4e0 fffff880`0795e250 00000000`00000000 00000000`0000000c : dxgkrnl+0x5d000
fffff880`0270d008 fffffa80`1e06b4e0 : fffff880`0795e250 00000000`00000000 00000000`0000000c 00000000`00000000 : 0x116
fffff880`0270d010 fffff880`0795e250 : 00000000`00000000 00000000`0000000c 00000000`00000000 00000000`00000000 : 0xfffffa80`1e06b4e0
fffff880`0270d018 00000000`00000000 : 00000000`0000000c 00000000`00000000 00000000`00000000 fffff880`09136c7c : atikmpag+0x9250


STACK_COMMAND:  kb

FOLLOWUP_IP:
atikmpag+9250
fffff880`0795e250 4055            push    rbp

SYMBOL_STACK_INDEX:  4

SYMBOL_NAME:  atikmpag+9250

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: atikmpag

IMAGE_NAME:  atikmpag.sys

BUCKET_ID:  WRONG_SYMBOLS

Followup: MachineOwner
---------

0 Kudos
4 Replies
jschellhaass
Enthusiast
Enthusiast

I would consider checking the system RAM.  I would start by pulling the second bank or RAM and seeing if it fails.  If it stiil fails swap the banks and try again.  You could also try running memtest86 and see if it finds any errors.

jeff

0 Kudos
dopefish
Contributor
Contributor

was just running memtest86 for 6 iterations, 0 errors.

0 Kudos
continuum
Immortal
Immortal

can you run several VMs without 3d support ?


________________________________________________
Do you need support with a VMFS recovery problem ? - send a message via skype "sanbarrow"
I do not support Workstation 16 at this time ...

0 Kudos
dopefish
Contributor
Contributor

i had 8 machines running at the same time without using d3d. all smooth, no lag at all.

but i did not stress those machines they were just idle on desktop and i was swapping around between them, could test multiple machines with high load if you think it might help.

i also temporarily ordered a gf 560gtx which arrives tomorrow just to make a check if its the ati card/driver.

just curious, how much use does the vms actually make of the hosts 3d card?

0 Kudos