Its typical on bigger programs, that several processes must be watched with different timeouts.

So every process have its own variable, which was set to the desired timeout value whenever the process was executed right.

And in a timer interrupt all these variables are counted down. And whenever a variable reach zero, the interrupt enters an endless loop. Thus the hardware watchdog was no longer triggered and generate a reset.

This timer interupt must also have the highest priority, so no other interrupt can leave the endless loop. Furthermore this timerinterrupt can read the stack value, if it grows to big and then the endless loop can also be entered.


Peter