>>>>> "IS" == Ivan Shmakov <i...@main.uusia.org> writes:
[...]
IS> BTW, I wasn't able to build i386 binaries: the # pbuilder build
IS> process has failed with:
> --cut--
> rm -f Swath_2.hdf Swathc_Test.hdf Swathf_Test.hdf testswathf.txt
> ./testswath_f90
> *** glibc detected ***
> /tmp/buildd/hdf-eos4-2.16v1.00.dfsg.1/testdrivers/swath/.libs/lt-testswath_f90:
> free(): invalid next size (fast): 0x087adbe0 ***
I was able to overcome this problem by disabling the Fortran
code below testdrivers/ entirely, like:
--cut--
diff -u hdf-eos4-2.16v1.00.dfsg.1/debian/rules
hdf-eos4-2.16v1.00.dfsg.1/debian/rules
--- hdf-eos4-2.16v1.00.dfsg.1/debian/rules
+++ hdf-eos4-2.16v1.00.dfsg.1/debian/rules
@@ -6,7 +6,8 @@
DEB_MAKE_CHECK_TARGET := check
DEB_MAKE_CLEAN_TARGET := clean distclean maintainer-clean
-DEB_CONFIGURE_EXTRA_FLAGS := --enable-shared --with-pic
+DEB_CONFIGURE_EXTRA_FLAGS := --enable-shared --with-pic \
+ he2_cv_fortran90_compiler=no
DEB_CONFIGURE_SCRIPT_ENV += \
CFLAGS="" \
CPPFLAGS="-Df2cFortran -I/usr/include/hdf"
diff -u hdf-eos4-2.16v1.00.dfsg.1/debian/changelog
hdf-eos4-2.16v1.00.dfsg.1/debian/changelog
--- hdf-eos4-2.16v1.00.dfsg.1/debian/changelog
+++ hdf-eos4-2.16v1.00.dfsg.1/debian/changelog
@@ -1,3 +1,10 @@
+hdf-eos4 (2.16v1.00.dfsg.1-2.is+0.4) 1gray-misc; urgency=low
+
+ * Pass he2_cv_fortran90_compiler=no to configure, effectively disabling
+ Fortran90 tests; hack configure.ac to allow it.
+
+ -- Ivan Shmakov <i...@main.uusia.org> Tue, 10 Nov 2009 22:44:33 +0600
+
hdf-eos4 (2.16v1.00.dfsg.1-2.is+0.3) 1gray-misc; urgency=low
* Hack debian/rules so that make check will find all the necessary
only in patch2:
unchanged:
--- hdf-eos4-2.16v1.00.dfsg.1.orig/configure.ac
+++ hdf-eos4-2.16v1.00.dfsg.1/configure.ac
@@ -204,11 +204,13 @@
if test -n "$FC"; then
AC_MSG_CHECKING([for fortran compiler $FC])
AC_LANG_PUSH([Fortran])
+ AC_CACHE_VAL([he2_cv_fortran90_compiler], [
AC_TRY_LINK([], [
INTEGER :: I
REAL :: R
I=TRANSFER(R,I)
], [he2_cv_fortran90_compiler=yes], [he2_cv_fortran90_compiler=no])
+ ])
AC_LANG_POP([Fortran])
if test ${he2_cv_fortran90_compiler} = "yes"; then
AC_MSG_RESULT([Fortran 90 code was successfully compiled; assume f90 or
later])
--cut--
[...]
I've put the resulting the source and binary (amd64, i386)
packages at:
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/source/hdf-eos4_2.16v1.00.dfsg.1-2.is+0.4.diff.gz
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/source/hdf-eos4_2.16v1.00.dfsg.1-2.is+0.4.dsc
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/amd64/libhdfeos-dev_2.16v1.00.dfsg.1-2.is+0.4_amd64.deb
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/amd64/libhdfeos0_2.16v1.00.dfsg.1-2.is+0.4_amd64.deb
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/i386/libhdfeos-dev_2.16v1.00.dfsg.1-2.is+0.4_i386.deb
http://waterlily.ipv6.uusia.org/~ivan/mini-dinstall/1gray-misc/i386/libhdfeos0_2.16v1.00.dfsg.1-2.is+0.4_i386.deb
--cut: /etc/apt/sources.list --
deb http://waterlily.ip.uusia.org/~ivan/mini-dinstall/ 1gray-misc/$(ARCH)/
deb http://waterlily.ip.uusia.org/~ivan/mini-dinstall/ 1gray-misc/all/
deb-src http://waterlily.ip.uusia.org/~ivan/mini-dinstall/ 1gray-misc/source/
--cut: /etc/apt/sources.list --
--
FSF associate member #7257
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
