[Icc-avr] Calculations

[John Baraclough]

My guess would be that the compiler has done the intermediate 
calculation as 'int' which will give some strange results. When using 
'long' or 'unsigned long' in calculations you should ensure that you 
stay in that domain until the calculation is complete and then translate 
to 'unsigned int'.

Try:

temp2 = (unsigned int)(0L/((16L*(long)baudrate)-1L));

That should work.

John



Bengt Ragnemalm wrote:
>
> Hi!
>
>  
>
> I have spent some of my spare time during Christmas solving some 
> strange problems I have had with a design. My problems was something 
> completely different but I ended up with this:
>
>  
>
> - I have found a calculation that may be a compiler error.
>
>  
>
> This is my test code, baudrate is 38400:
>
>  
>
> #define F_CPU 7372800
>
>
> temp2 = F_CPU/(16L*baudrate)-1;    // Normal calculation, result OK
> temp2 = 0/(16L*baudrate)-1;    // Results into 65535
>
> I guess that the last calculation error is because of some rounding 
> problem but obviously the correct answer is 0.
>
>  
>
> /Bengt
>
> ________________________________________________________________
>
> Bengt Ragnemalm                        Tel +46 13 22 24 97
>
> Research engineer                                       FAX: +46 13 10 
> 19 02
>
> Linköpings Universitet                                
> mailto:bengt.ragnemalm at imt.liu.se
>
> Inst. för Medicinsk Teknik Internet:        http://www.imt.liu.se 
> <http://www.imt.liu.se/>
>
> S-581 85 Linköping SWEDEN
>
>  
>
>  
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Icc-avr mailing list
> Icc-avr at imagecraft.com
> http://dragonsgate.net/mailman/listinfo/icc-avr
>