[986] in Kerberos-V5-bugs
here's a weird one..
daemon@ATHENA.MIT.EDU (*Hobbit*)
Sat Dec 10 17:56:26 1994
Date: Sat, 10 Dec 1994 17:56:18 -0500
From: *Hobbit* <hobbit@asylum.sf.ca.us>
To: krb5-bugs@MIT.EDU
On oldish ultrix 4.2 on a microvax: /usr/include/stdarg.h exists, but
the compiler there is definitely *not* __STDC__. This breaks a lot of
things, because the top level aclocal.m4 defines CHECK_STDARG and
several "configure"s further down in the swamp call it, notably in
util/et, util/ss, include/krb5, and clients/ksu.
So the existence of stdarg.h is NOT a valid indicator that a compiler can
handle prototyping.
There are several places I can "fix" this, but someone else probably knows
where the "right" place to fix it is -- could they clue me in? I can
sleaze around it for now, I suppose...
The code is full of things like
#if __STDC__ || defined(STDARG_PROTOTYPES)
... what is STDARG_PROTOTYPES *really* for? Whether or not to include
stdarg.h or varargs.h doesn't seem to fall into the same pigeonhole as
whether or not to prototype! It probably just happens to work out that
way on a lot of systems.
_H*
