Ok for darwin8, I got rid of: -Wl,@LOAD@,$(ARPACKLIB)
in favor of the simple: $(ARPACKLIB) And the build succeeded, but strangely, validator now reports a different compatibility version (even though %v is the same). Validating .deb dir /sw/src/fink.build/root-arpack-atlas-shlibs-96patched-6... Error: Shlibs field says compatibility version for /sw/lib/libarpack.0.dylib is 1.0.0, but it is actually 1.1.0. Fang > On 19 Sep 2011, at 22:02, David Fang wrote: > >> Alex, >> I'm trying your arpack, and for both (. -atlas), I get link failures on >> powerpc-darwin8: >> >> ( mkdir tmp ) >> ( cd tmp; ar -x /sw/src/fink.build/arpack-96patched-6/ARPACK/libarpack.a ) >> ( /sw/bin/gfortran-fsf-4.6 -O -fPIC -single_module -ff2c -dynamiclib >> -install_name /sw/lib/libarpack.0.dylib -compatibility_version 1.1.0 >> -current_version 1.1.0 -o >> /sw/src/fink.build/arpack-96patched-6/ARPACK/libarpack.dylib >> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libLAPACK.dylib >> >> /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib >> >> -Wl,-all_load,/sw/src/fink.build/arpack-96patched-6/ARPACK/libarpack.a ) >> /usr/bin/ld: multiple definitions of symbol ___muldi3 >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_muldi3.o) >> >> private external definition of ___muldi3 in section (__TEXT,__text) >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_muldi3.o) >> >> private external definition of ___muldi3 in section (__TEXT,__text) >> /usr/bin/ld: multiple definitions of symbol ___negdi2 >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_negdi2.o) >> >> private external definition of ___negdi2 in section (__TEXT,__text) >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_negdi2.o) >> >> private external definition of ___negdi2 in section (__TEXT,__text) >> /usr/bin/ld: multiple definitions of symbol ___lshrdi3 >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_lshrdi3.o) >> >> private external definition of ___lshrdi3 in section (__TEXT,__text) >> /Volumes/Mercedes2/sw/lib/gcc4.6/bin/../lib/gcc/powerpc-apple-darwin8.11.0/4.6.1/libgcc.a(_lshrdi3.o) >> >> private external definition of ___lshrdi3 in section (__TEXT,__text) >> >> Haven't looked into this very deeply yet. Any ideas? > > Would try w/o the -all_load flag _ does apply here to libgcc.a also, and > seems to cause the above trouble. > (This might be why pogma discouraged the use of -all_load in favour of > -force_all_load (on >= 10.6 IIRC), > or the more robust "tmp/*.o" instead of > "-Wl,-all_load,/sw/src/fink.build/arpack-96patched-6/ARPACK/libarpack.a" > ) > > JF > -- David Fang http://www.csl.cornell.edu/~fang/ http://www.achronix.com/ ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2dcopy1 _______________________________________________ Fink-users mailing list Fink-users@lists.sourceforge.net List archive: http://news.gmane.org/gmane.os.macosx.fink.user Subscription management: https://lists.sourceforge.net/lists/listinfo/fink-users
