Re: [Chicken-users] strange behavior - any reason?

[Felix Winkelmann]

Joerg F. Wittenberger wrote:
> > From the chicken source I'd expect this loop to terminate soon.  But
> this loops endlessly.
>
>
>         (else (values "uncaught exception" (format #f "N ~a: >>~s<<" n reason) 
> '() '())))))
>     ((join-timeout-exception? ex)
>      (values "time out" "join-timeout-exception") '() '())

                                                  ^ (intended?)

>     (else (values "unknown exception" (format #f "~s" ex) '() '()))))
>
> So how do I find out where this uncaught exception comes from?
>
> As to the threading question: when the code runs into the above
> endless loop, the system load is almost zero on my single processor
> machine.  This loop eats about 2.5% processor time and that's it.
> (Only if I print the value of 'n' in every loop, I get a reasonable
> load.)  So what does the executable do instead of processing my stupid
> endless loop?  Does chicken have some smart optimization to conserve
> cpu load for usless operations?  ;-)

Not that I'm aware of (:-).

When you call `thread-join!' to wait for a thread, *every* exception
(unless otherwise handled) will result in an uncaught-exception.
So it is not unlikely, that a simple bug can trigger endless looping.
On the other hand it's not unlikely that there is a bug in the
threading/exception system, since I assume you are heavily stressing
it.

Can you supply some sample code that triggers this behaviour?


cheers,
felix