[racket] size_t and FFI?

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Sun Feb 10 09:00:01 EST 2013

I think that's likely to be right for all platforms where Racket runs,
but we should define it in one place.

Should we add `_size' to `ffi/unsafe', risking collisions with existing
code? Or add `ffi/size' (or `ffi/size_t')?

At Fri, 8 Feb 2013 20:45:07 +0100, Tobias Hammer wrote:
> This should work:
> 
> _uintptr = size_t
> _intptr  = ssize_t
> 
> Tobias
> 
> 
> On Fri, 08 Feb 2013 20:40:43 +0100, Tony Garnock-Jones <tonyg at ccs.neu.edu>  
> wrote:
> 
> > Hi all,
> >
> > When using Racket's FFI, is there a standard C numeric type
> > corresponding to size_t? If not, should there be? Many C APIs expose
> > functions along the lines of
> >
> >   ... myfunc(..., void *buf, size_t buflen, ...);
> >
> > and since size_t varies between 32- and 64-bit platforms, I'm unsure
> > what to supply there.
> >
> > I note that in collects we have
> >
> > ./db/private/generic/socket.rkt:   (define _size_t _int)
> > ./mred/private/wx/cocoa/image.rkt: (define _size_t _long)
> > ./racket/draw/unsafe/jpeg.rkt:     (define _size_t _intptr)
> >
> > ... of which the last one looks most likely to work properly.
> >
> > Cheers,
> >   Tony
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> 
> 
> -- 
> ---------------------------------------------------------
> Tobias Hammer
> DLR / Institute of Robotics and Mechatronics
> Muenchner Str. 20, D-82234 Wessling
> Tel.: 08153/28-1487
> Mail: tobias.hammer at dlr.de
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

Posted on the users mailing list.