Re: ivtools build failures with gcc40 in 64 bit arches

[Agustin M. <agu...@hi...>]

On Fri, Oct 23, 2009 at 10:15:43AM -0700, Scott Johnston wrote:
> Agustin,
> 
> Does the contents of this patch resolve a compiler error or a
> warning?  Information is still being lost, but if it this fools the
> compiler, I'm fine with that.

Yes, otherwise gcc(>40) will fail in 64 bit arches because of the different
pointer length. I add a modified patch since gcc-4.4.2 complained about
uintptr_t not being available in context. But last Debian gcc-4.2.2 seems to
have some problems, so this may be a compiler problem.

> Casting a pointer to boolean, int, or unsigned int is kind of
> meaningless (other than comparison to NULL), and was only supported
> for debug purposes.

If this is the only purpose, using plain (long) and (unsigned long) 
may even be enough.

Cheers,

-- 
Agustin