Minh Nguyen wrote: > Hi folks, > > When building Sage 4.3.alpha1 on t2 (SPARC Solaris 10) with GCC 4.4.1 > and the Sun linker/assembler, the build failed. This time, it failed > when building c_lib. Here's a relevant error message snippet: > > CC: Warning: Option -fPIC passed to ld, if ld is invoked, ignored otherwise > "src/ZZ_pylong.cpp", line 47: Error: "ZZ_set_pylong(NTL::ZZ&, _object*)" is > expe > cted to return a value. > 1 Error(s) detected. > scons: *** [src/so_ZZ_pylong.o] Error 1 > *** TOUCHING ALL CYTHON (.pyx) FILES *** > /opt/SUNWspro/bin/CC -o src/so_ZZ_pylong.o -c -KPIC -fPIC > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include/python2.6 > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include/NTL -Iinclude > src/ZZ_pylong.cpp > CC: Warning: Option -fPIC passed to ld, if ld is invoked, ignored otherwise > "src/ZZ_pylong.cpp", line 47: Error: "ZZ_set_pylong(NTL::ZZ&, > _object*)" is expected to return a value. > 1 Error(s) detected. > scons: *** [src/so_ZZ_pylong.o] Error 1 > > ---------------------------------------------------------- > sage: Building and installing modified Sage library files. > > > Installing c_lib > /opt/SUNWspro/bin/CC -o src/so_ZZ_pylong.o -c -KPIC -fPIC > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include/python2.6 > -I/scratch/mvngu/sandbox/sage-4.3.alpha1/local/include/NTL -Iinclude > src/ZZ_pylong.cpp > CC: Warning: Option -fPIC passed to ld, if ld is invoked, ignored otherwise > "src/ZZ_pylong.cpp", line 47: Error: "ZZ_set_pylong(NTL::ZZ&, > _object*)" is expected to return a value. > 1 Error(s) detected. > scons: *** [src/so_ZZ_pylong.o] Error 1 > ERROR: There was an error building c_lib. > >>From the look of it, there's a mixture of GCC and Sun C compilers when > one should choose either and stick with it for the rest of the build > process. Also, the build now fails earlier than last I remember. > Previously, the build would fail when building the package > sage-x.y.z.spkg. The full install log is up on sage.math [1]. If it > helps, here's my PATH: > > [mv...@t2 ~]$ echo $PATH > /usr/local/gcc-4.4.1-sun-linker/bin:/usr/local/bin2:/usr/bin:/usr/ccs/bin:/usr/local/bin:/usr/sfw/bin:/bin:/usr/sbin > > [1] > http://sage.math.washington.edu/home/mvngu/doc/sage/install/sage-4.3/install-sage-4.3.alpha1-t2-sparc-solaris-10.log >
This is a bug I have known about for some time. It is trac http://trac.sagemath.org/sage_trac/ticket/6595 It needs either 1) Someone with knowledge of SCons to fix it. 2) Get rid of SCons for the library code and replace with a Makefile. I looked at that briefly, but it was not obvious to me exactly what the Makefile should consist of. 3) Not have Sun Studio installed. 4) Not have have Sun Studio installed at standard locations of /opt/SUNWWspro or /opt/sunstudio12.1, but some other non-standard location. If you look at the odd ticket I have created, where I've used SunStudio, you will see I give the path as /opt/xxxsunstudio12.1, which avoids SCons picking up the compiler, as it does not know where to look for it. One such example is http://trac.sagemath.org/sage_trac/ticket/7034 (this is on my own personal machine, not t2). I could easily stop this happening on 't2', but several people are using the Sun Studio compiler on t2, so I'm reluctant to disable/remove/hide it on 't2'. On my own machines, I just hide Sun Studio in a non-standard location. It would be really good if someone can look at this. It is the only thing that stops Sage building on any SPARC. There is only sage-bdist which stops us creating a binary. I know how to fix the sage-bdist issue, but not this one. Dave -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
