This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Bootstrapping sparc-sun-solaris2.5.1 - missing rules for some .o files in libio


On Tue, 17 March 1998, 14:53:57, wilson@cygnus.com wrote:

 > 	OK, I just checked, and it's actually Solaris' make program to blame!
 > 
 > This is curious, but it does make sense.
 > 
 > The reason I say it is curious is because some of the people reporting the
 > problem did report that there were using GNU make.  However, now that I
 > think of it, I think those people were invoking it as `gmake'.  So perhaps
 > the problem is that somewhere we were accidentally using `make' for a
 > recursive invocation when we should have used ${MAKE}.  Or perhaps somewhere
 > a configure script invoked make and ${MAKE} hadn't been set yet.

I just grep'ed through the sources, and found only one reference to a 
plain `make' instead of ${MAKE}; it's my recent patch to config-ml.in
and shouldn't be related to this particular problem:

      (cd ${ml_dir}/${ml_libdir};
       ../${dotdot}${ml_unsubdir}symlink-tree ../${dotdot}${ml_unsubdir}${ml_libdir} "")
      if [ -f ${ml_dir}/${ml_libdir}/Makefile ]; then
	if [ x"${MAKE}" = x ]; then
	  (cd ${ml_dir}/${ml_libdir}; make distclean)
	else
	  (cd ${ml_dir}/${ml_libdir}; ${MAKE} distclean)
	fi
      fi

I really guess, people were accidentially reporting they've used GNU
make. At least, GNU make version 3.75 and 3.76.1 don't have this
problem.

I'm open to any proof of the opposite, though.

manfred


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]