home *** CD-ROM | disk | FTP | other *** search
- Path: sdrc.com!thor!scjones
- From: larry.jones@sdrc.com (Larry Jones)
- Newsgroups: comp.std.c
- Subject: Re: size_t
- Date: 2 Feb 1996 15:44:08 GMT
- Organization: SDRC Engineering Services
- Distribution: world
- Message-ID: <4etbg8$sva@info1.sdrc.com>
- References: <4eo3sc$l14@info1.sdrc.com> <4es27i$hd5@alterdial.UU.NET>
- NNTP-Posting-Host: thor.sdrc.com
- Originator: scjones@thor
-
- In article <4es27i$hd5@alterdial.UU.NET>, rex@aussie.com writes:
- > > larry.jones@sdrc.com (Larry Jones) writes:
- > > size_t must indeed be one of the standard types.
- >
- > Larry, I recall a discussion in committee re mapping pointers to some
- > integral type. Somewhere we require that I believe. IBM's AS400 came up as it
- > has (I think) 128-bit addresses but no integer type that big. I thought we
- > allowed magic integer types for some things but I may be misremembering.
-
- The requirement is in (ANSI) 3.3.4 Cast Operators:
-
- A pointer may be converted to an integral type. The size of
- integer required and the result are implementation-defined. If
- the space provided is not long enough, the behavior is
- undefined.
-
- My recollection of the discussion was that the committee agreed that
- the result need not be sensible (e.g., a conforming implementation
- could convert all pointers to 42) and that the implementation was free
- to document that *no* integral type is long enough.
- ----
- Larry Jones, SDRC, 2000 Eastman Dr., Milford, OH 45150-2789 513-576-2070
- larry.jones@sdrc.com
- You should see me when I lose in real life! -- Calvin
-