Re: [avr-gcc-list] assembler expressions

[Jesper Hansen]

I know, and you also get the wrong result !
That's why the typecasts are needed. (Or a different way to obtain the
correct result).
I think it MUST be possible to do this in the assembler (or preprocessor or
whatever,
it's SO easy in C.

/Jesper

----- Original Message ----- 
From: "Ben Mann" <address@hidden>
To: "'Jesper Hansen'" <address@hidden>; <address@hidden>
Sent: Sunday, March 06, 2005 1:24 AM
Subject: RE: [avr-gcc-list] assembler expressions


> Hi Jesper,
>
> "Ah"
>
> ok, just remove the types.
>
> I compiled this, and got no errors:
>
> -------------
> #define F_CPU 18432000
> #define F1 20000
> #define F2 20001
> #define LO_F ( F1 *  F_CPU / (15*16777216) )
> #define HI_F ( F2 *  F_CPU / (15*16777216) )
>
> ldi  r21,lo8(LO_F)
> ldi  r22,hi8(LO_F)
> ldi  r23,hi8(LO_F)
> -------------
>
> Ben Mann
>
>
> -----Original Message-----
> From: Jesper Hansen [mailto:address@hidden
> Sent: Saturday, 5 March 2005 10:17 PM
> To: address@hidden; address@hidden
> Subject: Re: [avr-gcc-list] assembler expressions
>
>
>
> No, the c-code like stuff just results in a lot of errors like these :
>
> file1.s:139: Error: missing ')'
> file1.s:139: Error: `)' required
> file1.s:139: Error: garbage at end of line
> file1.s:140: Error: missing ')'
> file1.s:140: Error: `)' required
> file1.s:140: Error: garbage at end of line
> file1.s:141: Error: missing ')'
>
> /Jesper
>
> ----- Original Message ----- 
> From: "Ben Mann" <address@hidden>
> To: "'Jesper Hansen'" <address@hidden>; <address@hidden>
> Sent: Saturday, March 05, 2005 2:32 PM
> Subject: RE: [avr-gcc-list] assembler expressions
>
>
> > Why doesn't this work?
> >
> > #define F_CPU 18432000
> > #define F1 20000
> > #define F2 20001
> > #define LO_F (unsigned long) ( (float) F1 *  F_CPU / (15*16777216) )
> > #define HI_F (unsigned long) ( (float) F2 *  F_CPU / (15*16777216) )
> > ...  ldi  r21,lo8(LO_F)
> >  ldi  r22,hi8(LO_F)
> >  ldi  r23,hi8(LO_F)
> > ??
> >
> > I have used code like this in the past with no problems.
> > I don't know what the implications of (float) are though.
> >
> > Regards
> >
> > Ben Mann
> >
> > -----Original Message-----
> > .equ LO_F,  F1 * F_CPU/(15*16777216)
> > .equ HI_F,  F2 * F_CPU/(15*16777216)