Hi,
I decided to try fixing this bug, since it just looked like trilinos was
missing a Build-Depend on automake1.10 instead of automake. However,
there are two more problems:
1. trilinos still build-depends on python-numeric. This was removed from
Debian last month. [1] It also build-depends on python-numpy, which is
the suggested alternative. However, I'm not sure if just removing the
build-depends is enough, because trilinos won't build from source.
2. When the following patch is applied:
diff -u trilinos-9.0.3.dfsg/debian/control
trilinos-9.0.3.dfsg/debian/control
--- trilinos-9.0.3.dfsg/debian/control
+++ trilinos-9.0.3.dfsg/debian/control
@@ -3,7 +3,7 @@
Priority: extra
Maintainer: Debian Scientific Computing Team
pkg-scicomp-de...@lists.alioth.debian.org
Uploaders: Christophe Prud'homme prudh...@debian.org
-Build-Depends: cdbs, quilt, debhelper (= 7), automake, autoconf,
gfortran, libopenmpi-dev, libsuperlu3-dev, libsuitesparse-dev,
python-dev, swig1.3, python-numpy, python-numeric, libexpat1-dev,
libxml2-dev
+Build-Depends: cdbs, quilt, debhelper (= 7), automake1.10, autoconf,
gfortran, libopenmpi-dev, libsuperlu3-dev, libsuitesparse-dev,
python-dev, swig1.3, python-numpy, libexpat1-dev, libxml2-dev
Standards-Version: 3.8.1
Homepage: http://trilinos.sandia.gov/
Vcs-Svn: svn://svn.debian.org/svn/pkg-scicomp/trilinos/trunk/
The following error is produced:
make[4]: Entering directory
`/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg/packages/teuchos'
cd . /bin/bash
/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg/packages/teuchos/config/missing
--run aclocal-1.10 -I config
/usr/bin/m4:configure.ac:562: bad expression in eval: m4_Cursor + 0 + 1
/usr/bin/m4:configure.ac:562: bad expression in eval: ()
/usr/bin/m4:configure.ac:570: bad expression in eval: 1 ()
/usr/bin/m4:configure.ac:577: bad expression in eval: 2 ()
autom4te: /usr/bin/m4 failed with exit status: 1
aclocal-1.10: autom4te failed with exit status: 1
WARNING: `aclocal-1.10' is needed, and you do not seem to have it handy
on your
system. You might have modified some files without having the
proper tools for further handling them. Check the `README'
file,
it often tells you about the needed prerequirements for
installing
this package. You may also peek at any GNU archive site, in
case
some other package would contain this missing `aclocal-1.10'
program.
make[4]: *** [aclocal.m4] Error 1
make[3]: *** [all-recursive] Error 1
make[2]: *** [all-recursive] Error 1
make[4]: Leaving directory
`/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg/packages/teuchos'
make[3]: Leaving directory
`/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg/packages'
make[1]: *** [all] Error 2
make: *** [debian/stamp-makefile-build] Error 2
make[2]: Leaving directory
`/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg'
make[1]: Leaving directory
`/build/dmr-trilinos_9.0.3.dfsg-1.1-amd64-CFM9Lj/trilinos-9.0.3.dfsg'
dpkg-buildpackage: error: debian/rules build gave error exit status 2
I see no obvious source for this error. Line 562, 570, and 570 of
configure.ac are just macros testing for options on the configure call;
line 555 has the exact same macro but doesn't error out. The macro
doesn't even involve any comparison, to my eye.
Here are the relevant lines in the file:
551 #
552 # --enable-teuchos-boost
553 #
554
555 TAC_ARG_ENABLE_FEATURE_SUB( teuchos, boost, [Enable support for some
of boost], TEUCHOS_BOOST, no)
556 AM_CONDITIONAL(HAVE_TEUCHOS_BOOST, [test X$ac_cv_use_teuchos_boost
!= Xno])
557
558 #
559 # --enable-teuchos-comm_timers
560 #
561
562 TAC_ARG_ENABLE_FEATURE_SUB( teuchos, comm_timers, [Enable timers for
Teuchos::Comm related software],
563 TEUCHOS_COMM_TIMERS, yes )
564 #AM_CONDITIONAL(HAVE_TEUCHOS_COMM_TIMERS, [test
X$ac_cv_use_teuchos_comm_timers != Xno])
And the macro itself:
AC_DEFUN([TAC_ARG_ENABLE_FEATURE_SUB],
[
AC_ARG_ENABLE([$1-$2],
AC_HELP_STRING([--enable-$1-$2],[$3 (default is [$5])]),
ac_cv_use_$1_$2=$enableval, ac_cv_use_$1_$2=$5)
AC_MSG_CHECKING(whether to use [$1-$2])
if test X$ac_cv_use_$1_$2 != Xno; then
AC_MSG_RESULT(yes)
AC_DEFINE([HAVE_$4],,[Define if want to build $1-$2])
else
AC_MSG_RESULT(no)
fi
])
That's as far as I was able to get on this bug.
Regards,
Daniel
[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=546570
signature.asc
Description: OpenPGP digital signature
