🚀 go-pugleaf

In building GNU cpio 1.5 on an ISC 386/ix v2.0.2 system, I discovered a
bug in the configure script.  I'm using GCC 2.1, and don't have bcopy.
However, when the script executes the test for bcopy, it compiles:

func=bcopy
main() { exit(0); ${func}(); }

Unfortunately, since the call to bcopy can't be reached (and GCC 2.1
knows that), the call is eliminated, and so there is no reference to the
symbol.  So I changed the compiled code to:

main(argc, argv) char**argv;{ if(argc==1)exit(0); ${func}(); }

And that correctly identifies that bcopy isn't found in any standard
library.

There is also another misfeature.  From Makefile.in:

OBJS = copyin.o copyout.o copypass.o dstring.o fnmatch.o global.o \
main.o util.o error.o getopt.o getopt1.o filemode.o version.o \
$(RTAPELIB) dirname.o idcache.o makepath.o xmalloc.o xstrdup.o \
@LIBOBJS@ @ALLOCA@ 

This works fine only if either @LIBOBJS@ or @ALLOCA@ is non-empty.
Otherwise, the backslash at the end of the previous line combined with
the space between @LIBOBJS@ and @ALLOCA@ confuses the system-supplied
make (I don't have GNU make yet).

Please acknowledge receipt of this message, as I'm not entirely sure
you'll receive it....

	Marty Shannon
--
--------------+ <Delapsus Resurgam: When I fall I shall rise. --(10cc)>
Marty Shannon | My opinions are just that.  You may share them.  No one
mjs@s4mjs.uucp| speaks for me, and I speak only for myself -- no matter
--------------+ where I post from.  Get it?		Post no flames.