Le tridi 3 floréal, an CCXXIV, Nicolas George a écrit :
> > > +typedef union {
> > > + float f;
> > > + struct {
> > > + unsigned int mantissa : 23;
> > > + unsigned int exponent : 8;
> > > + unsigned int sign : 1;
> > > + } parts;
> > > +} CFloat;
> >
> > Will this work on all systems where FFmpeg can currently be
> > compiled? Or do you have to use the softfloat api from
> > libavutil or reuse lagarith code?
>
> Definitely not. This is the kind of code that should get someone who teaches
> it fired.This was a big ambiguous, thanks to Carl Eugen for pointing it: I meant that this code is ugly completely non-portable, and thus unacceptable in FFmpeg's code base. Regards, -- Nicolas George
signature.asc
Description: Digital signature
_______________________________________________ ffmpeg-devel mailing list [email protected] http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
