On Monday, March 31, 2014 5:40:56 PM UTC+2, Rob McMahon wrote: > > On 30/03/2014 13:56, Jean-Pierre Flori wrote: > > > > It would be nice to confirm a 64 bit build is working. > > Just set SAGE64=yes and rebuild everything. > > As far as I remember, the only package not taking this env var into > > account was the GCC spkg. > > And ATLAS did not manage to use a priori good settings. > > See > > > http://trac.sagemath.org/wiki/ExoticPorts#SunUltrasparcT2runningSolaris1064bitkernel32bituserland64bitSagebuild > > > > I just went through the same steps as with the 32-bit version, only with > SAGE64=yes set. I.e. I ended up with > > env SAGE64=yes SAGE_PORT='x' CXX='g++ -D_GLIBCXX_USE_C99_MATH' > MAKE='gmake -kj32' gmake -k > env SAGE64=yes SAGE_PORT='x' CXX='g++ -D_GLIBCXX_USE_C99_MATH' > MAKE='gmake -j32' gmake > > Then > > env SAGE64=yes SAGE_PORT='x' CXX='g++ -D_GLIBCXX_USE_C99_MATH' > MAKE='gmake' gmake > > to untangle the error messages. It failed in mpir with > > libtool: compile: g++ -D_GLIBCXX_USE_C99_MATH -DHAVE_CONFIG_H -I. > -D__GMP_WITHIN_GMP -c cxx/dummy.cc -fPIC -DPIC -o .libs/dummy.o > libtool: compile: g++ -D_GLIBCXX_USE_C99_MATH -DHAVE_CONFIG_H -I. > -D__GMP_WITHIN_GMP -c cxx/dummy.cc -o dummy.o >/dev/null 2>&1 > /bin/sh ./libtool --tag=CXX --mode=link g++ -D_GLIBCXX_USE_C99_MATH > -Wc,-m64 -version-info 4:21 :3 -o libmpirxx.la -rpath > /home/cudcv/sage-6.1.1/local/lib dummy.lo cxx/isfuns.lo cxx/ismpf.lo > cxx/ismpq.lo cxx/ismpz.lo cxx/ismpznw.lo cxx/osdoprnti.lo cxx/osfuns.lo > cxx/osmpf.lo cxx/osmpq.lo cxx/osmpz.lo libmpir.la > libtool: link: g++ -D_GLIBCXX_USE_C99_MATH -shared -fPIC -DPIC > -nostdlib /usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2/crti.o > /usr/lib/values-Xa.o > /usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2/crtbegin.o > .libs/dummy.o cxx/.libs/isfuns.o cxx/.libs/ismpf.o cxx/.libs/ismpq.o > cxx/.libs/ismpz.o cxx/.libs/ismpznw.o cxx/.libs/osdoprnti.o > cxx/.libs/osfuns.o cxx/.libs/osmpf.o cxx/.libs/osmpq.o > cxx/.libs/osmpz.o -Wl,-R > -Wl,/home/cudcv/sage-6.1.1/local/var/tmp/sage/build/mpir-2.6.0.p4/src/.libs > > -Wl,-R -Wl,/home/cudcv/sage-6.1.1/local/lib ./.libs/libmpir.so > -L/home/cudcv/sage-6.1.1/local/lib > -L/usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2 -L/usr/ccs/lib > -L/usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2/../../.. -lstdc++ -lm > -lgcc_s /usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2/crtend.o > /usr/gcc/4.5/lib/gcc/sparc-sun-solaris2.11/4.5.2/crtn.o -m64 -Wl,-h > -Wl,libmpirxx.so.4 -o .libs/libmpirxx.so.4.0.21 > ld: fatal: file ./.libs/libmpir.so: wrong ELF class: ELFCLASS64 > ld: fatal: file /lib/sparcv9/libm.so: wrong ELF class: ELFCLASS64 > ld: fatal: file processing errors. No output written to > .libs/libmpirxx.so.4.0.21 > collect2: ld returned 1 exit status > gmake[6]: *** [libmpirxx.la] Error 1 > gmake[6]: Leaving directory > `/home/cudcv/sage-6.1.1/local/var/tmp/sage/build/mpir-2.6.0.p4/src' > > Everything in .libs in that directory is 64-bit except dummy.o. Oops. > In src/cxx/.libs, everything is 32-bit, so I guess the basic problem is > that the -m64 just isn't making it into CXXFLAGS. > > I would say this is an MPIR bug, as in Sage we just pass ABI=64 to mpir build system. What's strange is that I did not have similar issues during my attempt, but maybe that was because I was using a 64 bit GCC defaulting to 64 bit binaries, though I'm not sure anymore. Could you post your failed build log?
-- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
