Re: htonl, htons, ntohl and ntohs types

[Samuel Thibault <samuel dot thibault at ens-lyon dot org>]

Lars Munch, le Fri 21 Apr 2006 16:48:31 +0200, a écrit :
> On Fri, Apr 21, 2006 at 04:32:37PM +0200, Corinna Vinschen wrote:
> > On Apr 21 14:11, Lars Munch wrote:
> > > My code still gives me warnings due to a problem with stdint.h.
> > > 
> > > The Xint32_t typedef's uses long instead of int:
> > > 
> > > >From stdint.h:
> > > 
> > > typedef long int32_t;
> > > typedef unsigned long uint32_t;
> > > 
> > > I think they should be:
> > > 
> > > typedef int int32_t;
> > > typedef unsigned int uint32_t;
> > 
> > That's a problem with the code you're working on.  If it relies on the
> > fact that int32_t resp. uint32_t are typedef'ed as "int", then it
> > contradicts the purpose of stdint.h, which is, not relying on the
> > definition of underlying datatypes.
> 
> Thanks for the explanation and you are absolutely right, but isn't using
> long instead of int a potential problem, since long is usually 64bit on
> a 64bit system whereas int is usually 32 bit and both 32bit and 64bit
> systems? 

When 64bits will be supported, there will be an ifdef here.

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/