[Automake-commit] [SCM] GNU Automake branch, master, updated. v1.13.1b-193-gd90001e

[Stefano Lattarini]

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU Automake".

http://git.sv.gnu.org/gitweb/?p=automake.git;a=commitdiff;h=d90001e3a90ea37fe6866120eb96be7bed8ba66c

The branch, master has been updated
       via  d90001e3a90ea37fe6866120eb96be7bed8ba66c (commit)
      from  dc4eaab21b45b6b1fe4348db603380dd9681d5c3 (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
commit d90001e3a90ea37fe6866120eb96be7bed8ba66c
Author: Stefano Lattarini <address@hidden>
Date:   Tue Apr 30 22:26:46 2013 +0200

    compat: substitute '@mkdir_p@' as '$(MKDIR_P)' unconditionally
    
    We had started to do so only when the obsolete AM_PROG_MKDIR_P variable
    was used, but that has recently turned out not to be not good enough,
    because Gettext 0.18.2 (the latest one at the moment of writing) is using
    the modern macro AC_PROG_MKDIR_P in its '*.m4' files, but at the same time
    is still using the obsolete make variable '$(mkdir_p)' in its Makefile
    fragments.  This has been causing failures in real-world packages
    bootstrapped with latest Gettext (0.18.2) and with bleeding-edge automake.
    
    See for example:
    
      <http://lists.gnu.org/archive/html/bug-tar/2013-04/msg00009.html>
    
    And the worst things is that those failures have been basically silent
    and unnoticeable up until the final "make install" --- that is, after
    the affected package had been bootstrapped, configured and built.  Yikes!
    
    This issue has been reported to the Gettext developers:
    
      <http://lists.gnu.org/archive/html/bug-gettext/2013-04/msg00044.html>
    
    But even if they will quickly fix it, Gettext 0.18.2 will remain
    affected.
    
    The fix we can implement in Automake is small and unobtrusive enough,
    definitely worth having in order to save the users and developers from
    having to deal with this hard-to-spot issue.
    
    * m4/init.m4 (AM_INIT_AUTOMAKE): AC_SUBST '@mkdir_p@' to '$(MKDIR_P)'.
    * t/am-prog-mkdir-p.sh: Enhance.
    
    Signed-off-by: Stefano Lattarini <address@hidden>

-----------------------------------------------------------------------

Summary of changes:
 m4/init.m4           |    2 ++
 t/am-prog-mkdir-p.sh |   20 ++++++++++++++++++++
 2 files changed, 22 insertions(+), 0 deletions(-)

diff --git a/m4/init.m4 b/m4/init.m4
index f9b0965..9c7ab2f 100644
--- a/m4/init.m4
+++ b/m4/init.m4
@@ -22,6 +22,8 @@
 # release and drop the old call support.
 AC_DEFUN([AM_INIT_AUTOMAKE],
 [AC_PREREQ([2.65])dnl
+dnl Without this, packages using Gettext 0.18.2 might break horribly.
+AC_SUBST([mkdir_p], ['$(MKDIR_P)'])dnl
 dnl Autoconf wants to disallow AM_ names.  We explicitly allow
 dnl the ones we care about.
 m4_pattern_allow([^AM_[A-Z]+FLAGS$])dnl
diff --git a/t/am-prog-mkdir-p.sh b/t/am-prog-mkdir-p.sh
index ad5a810..b561bfe 100755
--- a/t/am-prog-mkdir-p.sh
+++ b/t/am-prog-mkdir-p.sh
@@ -67,4 +67,24 @@ $AUTOMAKE
 $MAKE check-local
 $MAKE distcheck
 
+# Now try using AC_PROG_MKDIR_P, but keeping the occurrences of
+# $(mkdir_p) and @address@hidden  This is to check against a regression
+# that hit us with Gettext 0.18.2.
+$MAKE maintainer-clean
+rm -rf autom4te*.cache
+
+sed 's/AM_PROG_MKDIR/AC_PROG_MKDIR/' configure.ac > t
+diff configure.ac t && fatal_ "failed to edit configure.ac"
+mv -f t configure.ac
+
+$ACLOCAL 2>stderr \
+  && $AUTOCONF -Wall -Werror 2>>stderr \
+  && test ! -s stderr \
+  || { cat stderr >&2; exit 1; }
+
+$AUTOMAKE
+./configure
+$MAKE check-local
+$MAKE distcheck
+
 :


hooks/post-receive
-- 
GNU Automake