Generate a TRACEBACK list, but keep running

I wonder if there is a way to generate a traceback list as if a fatal error has occurred, but to keep running.

I'd like to be able to tell where in a long chain of routines an error has occurred and be able to track the calling chain to the source of the internal problem.

Obviously, I could cause a fault to occur, but many times an issue can occur that doesn't rise to a level of forcing the code to crash. It would be helpful in the debugging process to have the execution continue after noting (and logging) the issue.

Bill

FTN77 offered some routines that might be (have been?) helpful in what you are asking for. They are detailed in the FTN77 library manual pages 17-26, available online on this site.

There's a slight chance that one or more of the routines still work - perhaps Paul could advise.

Thanks for the pointer to the FTN77 routines.

After reading through them, there's nothing there that satisfies the desire to cause an error to be signaled, yet continue execution.

Still looking...

I will give this some thought but here are two quick solutions...

1. Use /DEBUG and set a break point in the debugger.

2. Use 'INT 3' in CODE/EDOC....

   program main code int 3 edoc print*, 'OK' end

Item 2 above may not be much use on its own.

A more direct approach is to use Plato where you have the option to 'Run to cursor'. Compile with /DEBUG, select the point in the code then use a right click and find 'Run to cursor'.

When in the debugger, the Stack/Status window gives you the trace-back.

Bill, apologies that the FTN77 library wasn't useful in this case, but there are some routines there that may prove useful in the future.

Eddie

Eddie, no problem. Yes, it will be useful in the future, I am sure.

Paul, many thanks for the options available to me. I'll see which one can work best for my situation.