Debugging on the cRIO

[linuxboy]

Hi everyone,
I had a question about running code on the cRIO right from the windriver interface. Say I run some code on the cRIO using right click run on target (that is probably not verbatim since I don't have windriver on here but it should be pretty correct) can I kill some tasks and be able to do it again without having to reset the cRIO?

Thanks,
Oliver

[kwojcik]

I figured out how to do this 2 years ago but I highly recommend you do not try to do this. Reason being: one time we deleted the tasks and the robot started driving full speed... very dangerous.

[whytheheckme]

Originally Posted by kwojcik

I figured out how to do this 2 years ago but I highly recommend you do not try to do this. Reason being: one time we deleted the tasks and the robot started driving full speed... very dangerous.

Really? I don't think this can happen due to the interface between the PPC and the FPGA - if the kernel crashes, the FPGA should disable the outputs.

[Greg McKaskle]

If the DS is still sending packets to the robot, then the system watchdog is happy and fed and will keep the FPGA alive. You can use the DS to disable or estop.

The user watchdog needs to be enabled in order for a breakpoint to stop the robot. The safety features of 2011 are equivalent to having the watchdog enabled, but will only stop a specific output.

Greg McKaskle