This is the mail archive of the
libstdc++@sources.redhat.com
mailing list for the libstdc++ project.
Re: mkcheck.in partial patch for AIX
- To: Alexandre Oliva <oliva at lsd dot ic dot unicamp dot br>
- Subject: Re: mkcheck.in partial patch for AIX
- From: Phil Edwards <pedwards at disaster dot jaj dot com>
- Date: Mon, 6 Nov 2000 00:01:14 -0500
- Cc: David Edelsohn <dje at watson dot ibm dot com>, libstdc++ at sources dot redhat dot com, Mark Mitchell <mark at codesourcery dot com>
- References: <200011040524.AAA25622@mal-ach.watson.ibm.com> <or8zr08e4j.fsf@guarana.lsd.ic.unicamp.br>
On Sat, Nov 04, 2000 at 04:06:52AM -0200, Alexandre Oliva wrote:
>
> > Any suggestion on how one would prefer making -rpath optional?
> > Substitute @rpath@ when building mkcheck? Test @host_os@ at runtime
> > to define $RPATH or choose $CXX template?
>
> If we really must keep it, using libtool and -rpath (without -Wl) is
> the way to go.
I am in the process of libtoolizing mkcheck while waiting on wwwdocs
feedback. It's going well so far, but can't be tested. BTW, I have
incorporated the AIX size(1) stuff that David has posted.
Something that's killing my progress now is -- and it's probably a newbie
question -- how do you pass mode-specific arguments to this thing?
For mode=link, both the online manual and the output of mode-specific
--help list -rpath as such an option (along with -no-install, which makes
sense for the testsuite executables). But neither of those ever make it past
-*)
$echo "$modename: unrecognized option \`$arg'" 1>&2
$echo "$help" 1>&2
exit 1
;;
which is part of the libtool script generated in the library builddir.
Phil
--
pedwards at disaster dot jaj dot com | pme at sources dot redhat dot com
devphil at several other less interesting addresses in various dot domains
The gods do not protect fools. Fools are protected by more capable fools.