Re: [avr-gcc-list] (no subject)

[Ruud Vlaming]

Hi

If i try (changed your code a bit to make it compile)
  #define mbitPatternPowerMode (7<<2)
  #define mADCNoiseReduction 0x20 
  #define mbitSleepMask 0x04

  void Test_MCUCR(void)
  {  MCUCR=(MCUCR&(~mbitPatternPowerMode))|(mADCNoiseReduction|mbitSleepMask); }

I get the same result. If you however try the same with
  #define mADCNoiseReduction   0x10

You get 0xE3 as expected. The point is, defining mADCNoiseReduction 
with bit 5 set, it does not make a difference to the result if you
use 0xC3 or 0xE5. So the compiler "optimized" it for you.

greetings
Ruud.


 


On Saturday 28 June 2008 12:28, Arkadiusz Krysiak wrote:
> can anyone explain that behavior?
> 
> i have:
> #define mbitPatternPowerMode=(7<<2);
> 
> MCUCR=(MCUCR&(~mbitPatternPowerMode))|(mADCNoiseReduction|mbitSleepMask);
> 
> and i get before compilation:
> 
> +00000088:   B785        IN      R24,0x35         In from I/O location
> +00000089:   7C83        ANDI    R24,0xC3         Logical AND with 
> immediate
> ;                                    ^^^^ bad value i expect 0xE3 
> instead of 0xC3   :O
> 
> +0000008A:   6284        ORI     R24,0x24         Logical OR with 
> immediate
> +0000008B:   BF85        OUT     0x35,R24         Out to I/O location
> 
> ----------------------------------------------------
> Wystartuj z siłą bolidu w wyścigu o kombinezon i czapkę 
> ROBERTA KUBICY!  Rusza wielka aukcja charytatywna, 
> sprawdź więcej: 
> http://klik.wp.pl/?adr=http%3A%2F%2Fcorto.www.wp.pl%2Fas%2Faukcje7368578.html&sid=387
> 
> 
> 
> 
> _______________________________________________
> AVR-GCC-list mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/avr-gcc-list
>