[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
libltdl's (1.5.24) Makefile.in configure ... autotool version mixing?
From: |
David Fang |
Subject: |
libltdl's (1.5.24) Makefile.in configure ... autotool version mixing? |
Date: |
Sat, 28 Jul 2007 18:59:20 -0400 (EDT) |
Hi all,
Concerning libltdl's build files: is it "safe" to forcibly
regenerate (local copies of) libltdl's {Makefile.in,configure,aclocal.m4}
from its {Makefile.am,configure.ac} using a slightly older version of
auto{make,conf} than they were originally created with?
Context: libtool-1.5.24 seems to have used automake-1.10 and
autoconf-2.61, however those are not the versions I'm autogen-ing with on
a particular machine, so when I try to regenerate, I get version mismatch
warnings -- and the targets don't get re-created as a result. Since I'm
bundling a copy of libltdl with the host package, I'd like maintainers to
be able to autogen using whatever single version of autotools they have
installed (provided that they meet minimum versions: 1.9, 2.59).
*Normally*, one should not need to regenerate libltdl's files, as
they have been pre-generated, however I found the following problem:
Makefile.in (by am-1.10) references 'mkinstalldirs' which is some script
that doesn't seem to be bundled with am-1.9, and thus, doesn't exist after
autogen. However, 'mkinstalldirs' is a dependency of 'distdir' in
libltdl, so a parent-level "make dist" fails to find this file, when it
recurses into libltdl/.
So I stripped out libltdl's Makefile.in, configure, aclocal.m4 and
force thosed to be regenerated with am-1.9, ac-2.59. Makefile.in no
longer references the nonexistent 'mkinstalldirs' and the top-level
dist/distcheck seems happy.
Is this a sin I'm going to pay for?
Thanks in advance.
David Fang
Computer Systems Laboratory
Electrical & Computer Engineering
Cornell University
http://www.csl.cornell.edu/~fang/
-- (2400 baud? Netscape 3.0?? lynx??? No problem!)
- libltdl's (1.5.24) Makefile.in configure ... autotool version mixing?,
David Fang <=