I got a Custom Device which performs some CAN-Communications. The Custom Device runs for hours on end in Pharlap. I made a distribution for LinuxRT, at first the custom device seems to run correcty but after a while on of the Controller's (PXIe8840eQC) cores jump to 100% and the PXI responds no more.
In Phalap the load is kinda stable, in Linux the load jumps all over the place.
does anyone know how can I avoid this behaviour?
I know the post is a bit general but i can not post the CD. Maybe there are some considerations for the linuxRT Programming that differ from Pharlap?
Is this in LabVIEW? I'd try and run in the development environment to see when the CPU goes to 100%. Yes, there are some considerations. For example, if you were opening a file that's at a different path in Linux, then that error could make a function that normally waits patiently (eg dequeue) immediately return an error which might use 100% of the CPU
Yes the implementation is in LabVIEW.
I discoverd, that the problem was, that too many of my VIs had high priority. I set most of the VIs to normal priority and the problem is now gone.