I am having problems configuring and installing lam-7.1.1 on OS X 10.4.1
that follows on from the posting of Christophe Peyret on 5/10/05.
The bottom line is that all goes well until the "make install" command,
when midway through, that process aborts with the triple error messages:
Making install in etc
test -z "/usr/local/lam-7.1.1/etc" ||
/usr/local/lam-7.1.1/config/install-sh -d "/usr/local/lam-7.1.1/etc"
/usr/bin/install -c -m 644 'lam-conf.lamd'
'/usr/local/lam-7.1.1/etc/lam-conf.lamd'
install: lam-conf.lamd and /usr/local/lam-7.1.1/etc/lam-conf.lamd are
the same file /usr/bin/install -c -m 644 'lam-helpfile'
'/usr/local/lam-7.1.1/etc/lam-helpfile'
install: lam-helpfile and /usr/local/lam-7.1.1/etc/lam-helpfile are
the same file
make[2]: *** [install-sysconfDATA] Error 64
make[1]: *** [install-am] Error 2
make: *** [install-recursive] Error 1
/usr/local/lam-7.1.1 root#
I have set my various flags and library paths similar to that of C.
Peyret using XL Fortran, but specific to my installation of Apple
developer tools 2.1:
export CC=/usr/bin/gcc-3.3
export CFLAGS="-O3 -mpowerpc -mpowerpc64"
export CPP=/usr/bin/cpp
export CPPFLAGS=" "
export CXX=/usr/bin/g++
export FC=/opt/ibmcmp/xlf/8.1/bin/xlf90
export FFLAGS="-O3"
I run the configure command as:
./configure --without-dash-pthread --with-rpi=usysv
-prefix=/usr/local/lam-7.1.1
and then:
make all
and:
make install
which gives the error above. Also note that my bash shell from which I
run all the above is set for library paths, and compiler flags as follows:
PATH=$PATH:$HOME/bin:/usr/local/bin/:/opt/ibmcmp/xlf/8.1/bin:/usr/local/lam-7.1.1/bin:/usr/local/lam-7.1.1/share:
/usr/local/lam-7.1.1/include:/usr/local/lam-7.1.1/lib:/usr/local/lam-7.1.1/lib/lam
export PATH
FLIBS='-L/opt/ibmcmp/xlf/8.1/lib -lxl -lxlfpmt4 -lxlfpmt4_r -lxlfpmt8_r
-lxlfpmt8 -lxlopt -lxlf90 -lxlf90_r -lxlfmath -lxlfpad -lxlfpad_r'
export FLIBS
CLIBS='-L/usr/lib/gcc/darwin/3.3 -lgcc -lcc_kext -lstdc++ -lstdc++_ZeroLink -lsupc++'
export CLIBS
Any help would be greatly appreciated getting this problem solved!
THANKS!
Greg Burton
|
