Re: [help-GIFT] Re: Patches AMD 64 next try...

[Wolfgang Mueller]

Wolfgang:
> 
> i'm still reviewing the last patch to make it work on my machine..
> 
> basically, if in gabor.c, if you change either the x or the y variable from a 
> uint32_t to a uint64_t, it starts magically working.
> 
> no, i've no idea. i'm still fighting this one. but after that, it magically 
> starts working, on my machine.
I am a fan of printouts. And my printouts tell me that some index is
getting negative, and the reason why makes sense. So my hunch is, that
if it works magically is that some overflow plays into our hands. But
this does not make the multiplication result correct.

Cheers,
Wolfgang
--
Dr. Wolfgang Mueller
LS Medieninformatik
Universitaet Bamberg