timeout of Hserin, goto, gosub or both?

[aratti]

Difficult to give suggestions without knowing what you have in mind to achieve. The below is a simple way to get out of the While/Wend loop.

StrLIst    var byte[10] ' for example not knowing the true dimension
gotchar   var byte


BUfferChar:
gotchar=0
return_code=0
low HIGH RED_LED
While return_code<>13 or gotchar >9
  Hserin 50, NoData, [return_code]
  StrLIst[gotchar]=return_code
  gotchar=gotchar+1   
  WEND
RETURN


Nodata:
HIGH RED_LED   ; Show Error led (use the high state of this pin to workout   the timeout in the main routine)
Return

Al.

[flipper_md]

well, I believe as you suggest that it's ok to break a While loop...

but what's tickle me is the GOTO timeout of Hserin, he would break out of a GoSub.

If it's really a GOTO, there must not be any trace of where it was initiated in the stack, so NoData will not know where to return...
And even more buggy, the BufferChar Gosub will not be completed.
hence creating a bug.

So in the end, I would believe it's not safe to use either a GOTO or a HSerin inside a Gosub.

[aratti]

If it's really a GOTO, there must not be any trace of where it was initiated in the stack, so NoData will not know where to return...
And even more buggy, the BufferChar Gosub will not be completed.
hence creating a bug.

When your program jumps to NoData label, in the stacker is still pending the last GOSUB instruction (the one that directed the program flow to the routine under BUfferChar label), so no matter how many GOTO you will use, the first RETURN will clear the last pending gosub and pointer will return the program flow to the main program routine (if the gosub had been executed there). And this is what will happen with the RETURN under NoData label.

Al.

[flipper_md]

very interres-thing!

thanks much for this Al

So it means goto's are safe inside a gosub, as long as there is a return in the end. I like that!