2009/7/28 Minh Nguyen <nguyenmi...@gmail.com>: > > Hi folks, > > So far in the 4.1.1 release cycle, David Kirkby has added Solaris (on > the machine t2) build support for about 7 packages. That is very > impressive work for an individual porter.
Thank you. > Out of curiosity, I compiled > Sage 4.1.1.alpha1 from source on t2. Now the compilation dies at the > point where the SPKG sage-4.1.1.alpha1 was being compiled. I may be > wrong here, but the context is that that SPKG tries to link to the NTL > library and the specific error is in building "c_lib". But don't take > my words for it. The full log is up at > > http://sage.math.washington.edu/home/mvngu/patch/install-solaris.log.bz2 > > and here's a relevant 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 expected 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 This is an issue we had in PolyBoRi and is related to the fact that SCons looks for a C++ compiler called 'CC' in /opt/SUNWspro/bin/ That path is hard coded into SCons. Again, like in PolyBoRi, a mix of Sun flags (-KPIC) and GNU flags (-fPIC) is being used. I dont know if this is a bug in SCons or a bug elsewhere. Alexander Dreyer fixed the problem in PolyBoRi though I'm personally not sure if the problem really lied in the PolyBoRi's code, or really the problems is a bug in SCons, that Alexander has managed to circumvent. it. I've created a quick *HACK* to SCons that makes it unaware of the /opt/SUNWspro/bin directory. You could try this hacked version of SCons if you wish. http://sage.math.washington.edu/home/kirkby/Solaris-fixes/scons-hack/scons-1.2.0.p0.spkg Hopefullly, it will get you a bit further. You should then reach a point where there is a compilain about paripriv.h Commenting out the offending lines will get the build going a bit further. There's one package while fails to build as an attempt to remove config.h files as the file does not exist. I touched that file, and then the build can progress a bit more. With a few hacks, it should be possible to get to the point where ecl fails. I dont have a clue how to fix that. Dave --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
