Hi, On Thursday, March 27, 2014 5:51:29 PM UTC+1, Rob McMahon wrote: > > I've worked around the problems, but I thought I should report them > here. I just tried building sage 6.1.1 on a new (to me ...) SPARC (T4-2) > box > > $ cat /etc/release > Oracle Solaris 11.1 SPARC > Copyright (c) 1983, 2013, Oracle and/or its affiliates. All rights > reserved. > Assembled 06 November 2013 > > The first make > > $ env MAKE='gmake -j32' gmake > > checking whether to enable maintainer-specific portions of Makefiles... > yes > configure: WARNING: you should use --build, --host, --target > configure: WARNING: you should use --build, --host, --target > configure: error: unrecognized option: `-j32' > Try `./configure --help' for more information > If you would like to try to build Sage anyway (to help porting), > export the variable 'SAGE_PORT' to something non-empty. > > Worked around with > > $ env SAGE_PORT='x' MAKE='gmake -j32' gmake > > The build failed at libfplll-4.0.4, which failed with complaints > > libtool: compile: g++ -DHAVE_CONFIG_H -I. -I.. > -I/home/cudcv/sage-6.1.1/local/include/ -fPIC > -I/home/cudcv/sage-6.1.1/local/include/ > -L/home/cudcv/sage-6.1.1/local/lib -MT fplll.lo > -MD -MP -MF .deps/fplll.Tpo -c fplll.cpp -fPIC -DPIC -o .libs/fplll.o > In file included from nr.h:430:0, > from numvect.h:19, > from matrix.h:21, > from wrapper.h:21, > from fplll.cpp:18: > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = double]’: > nr.cpp:868:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = long double]’: > nr.cpp:1078:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = dpe_struct [1]]’: > nr.cpp:1278:33: error: ‘isfinite’ was not declared in this scope > ... > libtool: compile: g++ -DHAVE_CONFIG_H -I. -I.. > -I/home/cudcv/sage-6.1.1/local/include/ -fPIC > -I/home/cudcv/sage-6.1.1/local/include/ > -L/home/cudcv/sage-6.1.1/local/lib -MT topenum.l > o -MD -MP -MF .deps/topenum.Tpo -c topenum.cpp -fPIC -DPIC -o > .libs/topenum.o > In file included from nr.h:430:0, > from numvect.h:19, > from matrix.h:21, > from util.h:23, > from util.cpp:16: > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = double]’: > nr.cpp:868:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = long double]’: > nr.cpp:1078:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = dpe_struct [1]]’: > nr.cpp:1278:33: error: ‘isfinite’ was not declared in this scope > gmake[5]: *** [fplll.lo] Error 1 > gmake[5]: *** Waiting for unfinished jobs.... > gmake[5]: *** [util.lo] Error 1 > In file included from nr.h:430:0, > from numvect.h:19, > from matrix.h:21, > from util.h:23, > from topenum.h:19, > from topenum.cpp:16: > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = double]’: > nr.cpp:868:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = long double]’: > nr.cpp:1078:23: error: ‘isfinite’ was not declared in this scope > nr.cpp: In member function ‘int fplll::FP_NR<F>::is_finite() const [with > F = dpe_struct [1]]’: > nr.cpp:1278:33: error: ‘isfinite’ was not declared in this scope > gmake[5]: *** [topenum.lo] Error 1 > gmake[5]: Leaving directory > `/home/cudcv/sage-6.1.1/local/var/tmp/sage/build/libfplll-4.0.4/src/src' > gmake[4]: *** [all-recursive] Error 1 > gmake[4]: Leaving directory > `/home/cudcv/sage-6.1.1/local/var/tmp/sage/build/libfplll-4.0.4/src' > gmake[3]: *** [all] Error 2 > gmake[3]: Leaving directory > `/home/cudcv/sage-6.1.1/local/var/tmp/sage/build/libfplll-4.0.4/src' > Error building libfplll > > A quick google search suggested compiling with > > $ env SAGE_PORT='x' CXX='g++ -D_GLIBCXX_USE_C99_MATH' MAKE='gmake -j32' > gmake > > which got past that problem. The build then aborted, the install log
Great! Would you mind oening a ticket on Sage's trac? Or even forward this upstream (possibly with a clena autotools patch)? Upstream (Damien Stehlé) has been more than reactive in the past to include fixes we needed for not so mainstream archs. > > ending in: > > Finished installing atlas-3.10.1.p7.spkg > gmake[2]: Leaving directory `/home/cudcv/sage-6.1.1/build' > gmake[1]: *** [all] Error 2 > gmake[1]: Leaving directory `/home/cudcv/sage-6.1.1/build' > > real 33m35.011s > user 361m10.877s > sys 23m39.620s > *************************************************************** > Error building Sage. > > The following package(s) may have failed to build: > > The build directory may contain configuration files and other potentially > helpful information. WARNING: if you now run 'make' again, the build > directory will, by default, be deleted. Set the environment variable > SAGE_KEEP_BUILT_SPKGS to 'yes' to prevent this. > > gmake: *** [build] Error 1 > > So I'm not sure what failed there. Re-ran the same make command, and it > aborted again with the same message this time after > > Successfully installed r-3.0.2.p0 > Deleting temporary build directory > /home/cudcv/sage-6.1.1/local/var/tmp/sage/build/r-3.0.2.p0 > Finished installing r-3.0.2.p0.spkg > > Then > > Successfully installed scipy-0.12.0.p1 > > and > > Successfully installed linbox-1.3.2.p0 > > Until > > gmake[2]: *** > It seems the error is at the following line: > [/home/cudcv/sage-6.1.1/local/var/lib/sage/installed/csage] Error 2 > So it should be when the libcsage.so is built. You can try to confirm that by forcing a rebuild of the file (with similar C*FLAGS). > gmake[2]: Leaving directory `/home/cudcv/sage-6.1.1/build' > gmake[1]: *** [all] Error 2 > gmake[1]: Leaving directory `/home/cudcv/sage-6.1.1/build' > > After that I had to re-run the make without the CXX=... setting. > > BTW, total time to build 2 hours, including typing this email. If > there's anything else I can do to help debug the problems, please let me > know. This is a test machine at the moment, so I pretty much have free > reign. I'm just running a ptestlong. > > 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 Cheers, JP -- 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.
