Intel® Hyper-Threading Technology conflicts with LabVIEW utilities (VISA, Scope GUI, IO Trace...)

[ejkaiser]

Hi altenbach,

before I go into details just one question: Even if the code is suboptimal and there may be race conditions, why would a simple mouseclick into the front panel (doesn't matter where) make a program continue that had stopped before.

Strange, isn't it?

Cheers

Edgar

[johnsold]

Do you have an event structure which is waiting for a front panel event?

Lynn

[ejkaiser]

No events. Front panel only has indicators and controls and the program is not waiting for any input. It runs for a while and then stalls without an obvious reason.

It is admittedly an ugly program with many locals, maybe race conditions ( I don't care, because they only apply to writing stuff into indicators).

What really puzzles me is this strange behavior: flawless operation, then unpredictable stall and going ahead from the last operation whith a mouse click into the front panel. Please note: I don't have to engage a control or something just a click on the grey non-functional surface inside the window.

I can't believe that this is a matter of my code.

[altenbach]

• Is the program actually running in the background during the stalls. Are only the front panel updates missing, or is the code itself stalling? (for example, if you place an indicator at the iteration counter, does it jump a large amount when activity returns?)
• What is the CPU use during stalling? What is it during typical use?
• How long is the time until the stall occurs (random, fixed amount, etc). How reproducible is it?
• You still have not told us the power settings of the PC. If you set it to "high performance", deos the problem still occur.
• Have you tried updating the video driver?
• What about antivirus and other utility software?
• Have you done a virus scan.
• What does the program actually do. Does it interact with instruments or DAQ hardware? If so, how?

[ejkaiser]

Here is a brief update:

Of course I checked all the bullet point in altenbach's message. I was taking care to avoid any interference from other software or background processes as far as you can do that on a Windows system. I had been doing that long before. It was all pointless. It just remained to be unpredictable.

It is funny what cured the error for the moment: I installed CorelGraphics X6 on the computer and the issue is gone. I can now do what I want, run many programs, video, etc. My Labview program runs flawlessly stable. I can even run number crunching FEA software in parallel, that drives the CPU close to 100%. Corel doesn't need to run, but in can run in parallel.

Looks like this installation has repaired something. Weird! Best thing is: it was the Corel test version, so the cure was for free.

Cheers

Edgar

[altenbach]

So, in summary, it wasn't hyperthreading, but probably some rogue shared library or other corrupt shared component. 'Glad you found a fix.

[ejkaiser]

Haha, it is more like the fix found me ...