home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!olivea!spool.mu.edu!howland.reston.ans.net!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!cis.ohio-state.edu!caip.rutgers.edu!ghazi
- From: ghazi@caip.rutgers.edu (Kaveh R. Ghazi)
- Newsgroups: gnu.utils.bug
- Subject: autoconf-1.3, USG, BSD-string suggestions
- Message-ID: <Jan.24.13.06.45.1993.9444@caip.rutgers.edu>
- Date: 26 Jan 93 01:38:12 GMT
- Sender: daemon@cis.ohio-state.edu
- Distribution: gnu
- Organization: Rutgers Univ., New Brunswick, N.J.
- Lines: 25
- Approved: bug-gnu-utils@prep.ai.mit.edu
-
-
- I have two suggestions for autoconf.
-
- I was thinking that perhaps autoconf should not decide to define/undef
- USG depending on whether or not it finds <strings.h>, bcopy() and index(), but
- should instead use another symbol.
- I've found that USG is generally taken to mean SYSV (r3 or less).
- But some SYSV's do include the BSD string functions so assuming that their
- presence ==> ^USG may be inappropriate. It makes more sense to relate BSD
- string stuff to something like HAVE_BSTRING when these functions are available,
- similar to what emacs does. USG, if its defined by autoconf at all, would be
- more useful if it related to something at the kernel level like interruptable
- system calls. (But because of its ambiguity, I'd rather it not be used at all.)
-
- A second suggestion regarding BSD string functions, is that autoconf,
- in addition to its C file compile test, egrep for their presence in strings.h
- similar to what you do for memchr() in SunOS. I have a machine (Tek XD88
- running UTekV) which has bcopy() and index() in libc, and it has <strings.h>,
- but strings.h doesn't declare the functions. So I end up getting gillions of
- integer - pointer warnings.
-
- --Kaveh
- --
- Kaveh R. Ghazi CAIP Center, Rutgers University.
- ghazi@caip.rutgers.edu rutgers!caip.rutgers.edu!ghazi
-