LabView crashes from driver nimsdrk.dll

[U.ofT.]

Labview 7.1 crashes randomly while datalogging. The memory dump error message appears as:

DRIVER_IRQL_NOT_LESS_OR_EQUAL

Technical Information:
***STOP: 0x000000D1 (0x00000000, 0x00000002, 0x00000000, 0xF3EDAF29)

*** nimsdrk.dll - Address F3EDAF29 base at F3ED7000, Datestamp 40687658

I have updated the drivers to the lastest version, and this did not solve the problem. The problem can take anywhere from one minute to 3 days to happen. The NI server is not allowing me to attach files, else I would include my VIs. Here are the hardware components that are in use.

SCXI-1000
SCXI-1180 & SCXI-1302
SCXI-1102 * SCXI-1300
SCXI-1121 & SCXI-1327
Connected via PCI-6024e

Dell Optiplex GX270
-Windows XP Proffesional
SP2
-2.8 GHz with HT
-512 MB RAM
-NVIDIA GeForce FX5200

I can post any other specifics that you may need.
Thank you.

[George M.]

Hi,

What version of the driver did you instal? Have you had this setup ever working? If so, what has changed since it was working. Can you try the setup on another computer, and see if it does the same thing? Let me know. Thanks, and have a Great Day!

George

[Bob_Y.]

I would lkie to add a big ol' ME TOO to this.

I have a customer that has been seeing a system reboot about 1 to 3 times a week. After several weeks of telling him that it must be a microsoft problem, we figured out a way to get it to log more data during the crash. It turns out that it is reporting that NIMSDRK.DLL is the problem "driver". The blue screen that we finally got XP to provide shows the problem formatted as in the previous post with the following data:

DRIVER_IRQL_NOT_LESS_OR_EQUAL
STOP 0x000000D1 (0x00000000, 0x00000002, 0x00000000, 0xF2E89F7A)
NIMSDRK.DLL Address F2E89F7A Base F2E86000 Datestamp 40EDE044

Unfortunately, I did not get all of the message recorded, but merely took down the highlights. After the reboot, I looked at the dll's properties. Version 1.2.1.49152 or Version 1.2.1f0 depending on where you look. The name for this DLL is National Instruments Measurement Streaming DMA Runtime Co (which I assume is controller).

The system is a Pentium 4 (with HyperThreading) running XP-pro SP1; connected through a MXI-4 bus to a PXI-8331 in a PXI/SCXI chassis.

Let me know if you figure out what the problem is or better yet a fix for it.

Thanks
Bob Young

[Quantumbyte]

We're looking into this problem, and should have some more information soon. I'd recommend you disable the Hyperthreading in the BIOS as a workaround. It might prevent it from happening. Could both of you tell me what subsystems you were running (AI,AO,etc...) and whether all of them were using DMA as the data transfer mechanism?

[Bob_Y.]

It has crashed several times. These are not always in the same place or while doing the same things. This made it very difficult to track down. At first I believed that it was a Microsoft problem. The I was convinced that it was a BIOS problem. Then I spoke to our IT person and he gave me some help on getting more information about the crash. It turns out that it is a problem with the NIMSDRK.DLL.

That being said, usually just before the crash we were using LabVIEW to control a PXI/SCXI chassis with the following cards:
PXI-8331 to communicate with the PC
PXI-6509 Digital I/O to switch some relays
PXI-2530 with TB-2632 configured as an 8X16 matrix to simulate an encoder (via Switch Executive)
PXI-6071E MIO used mainly to communicate with the SCXI portion of the chassis
SCXI-1104C with SCXI-1300 AI is used to scan some digital (non-TTL level) lines
SCXI-1124 with SCXI-1325 AO is used to simulate some sensors
SCXI-1127 with SCXI-1332 configured as an 8x8 matrix to simulate some outputs

The customer has been unable to figure out a commonality among the crashes. He has not been able to figure out any particular setting or change or transition that is causing the crash. It is a lot of HW through MAX from LabVIEW (and even that sometimes being called from TestStand) and there is not an easy way to make it repeat. It does not always crash. You can do the same thing several times in a row and then suddenly, it will crash on exactly the same thing. Or on the first time through something.

I am not sure if it is more AI, AO, DI, DO, switch control or what. I know that this not much to go on, but it really is very random. One thing to note is that it seems to happen more often if there are a lot of things (like sub-vis) open at once. This may be a false clue however, since it's usually much more annoying when you are in the middle of debugging something so it may just _seem_ to happen more often then.

Thanks,
Bob Young

[sergesl]

Hello Bob Y.

The same issue is being debated on the post at http://forums.ni.com/ni/board/message?board.id=250&message.id=8928.
Please use that threads for future posts.

Serges L.