On Thu, Apr 11, 2013 at 09:47:39PM -0700, Alexander Hansen wrote: > I had a problem at one point on my G4 machine when we were testing the > prior atlas update, and that was why the powerpc version got held back > at 3.9.13 at that time. I don't have access to that machine right now > (turned off and 300 miles away), so I can't confirm or test any issues. > > Maybe we should do the following if a fix can't be found in a reasonable > time interval: > > 1) Copy the existing atlas.info in 10.4/ to atlas-powerpc.info. > 2) Tag that with Distribution: powerpc and Depends: 64bit-cpu (so that > people don't think they can build it on a G5 and install it on a G4). > 3) Tag 10.4/*/atlas.info with Distribution: i386, x86_64. > (Here I'm assuming that folks running 32-bit-only intel machines can > actually build atlas-3.10.1 .) > 4) Resurrect the old atlas-powerpc.info (3.9.13) as atlas-ppc32.info or > something, under the new package name name atlas-ppc32. > 5) Tweak its library build to avoid having %p/lib/libblas.dylib and > %p/lib/liblapack.dylib in atlas-ppc32-shlibs. > 7) Have all of the atlas-using packages in the 10.4/ tree carry > dependencies on atlas (>= 3.10-1) | atlas-ppc32 (and similarly for the > -shlibs). > > It's a bit ugly, but given that we don't have a way to differentiate > between G4s and G5s otherwise I can't really think of another option > that can be rolled out quickly. Plus, if we do find a fix, we can just > scrub atlas-ppc32 without changing the reverse-depends--they'll just > complain that atlas-ppc32 doesn't exist any more and install atlas. > > Abother alternative would be to have atlas-powerpc.info pegged at 3.9.13 > and also with the newer library structure. To make that easier to > handle we might want to cheat on the version: call it whatever the > current version is for everybody else, but always use the 3.9.11 sources.
I don't have access to any G4's. Could someone with a G4 test changing the line... else mflags="$mflags -maltivec -mabi=altivec -force_cpusubtype_ALL" in altas.info to... else mflags="$mflags -maltivec -mabi=altivec -mcpu=7400 -mtune=7400 -force_cpusubtype_ALL" which is what MacPorts uses for their 32bit ppc builds. If that works, feel free to check in that change on both trees. > > On 4/11/13 8:09 PM, Sean Lake wrote: >> Hello Jack, >> >> Thought I'd throw in my data: >> >> Sean >> >> Package manager version: 0.34.7 >> Distribution version: selfupdate-rsync Thu Apr 11 12:54:12 2013, 10.5, >> powerpc >> Trees: local/main stable/main stable/crypto unstable/main unstable/crypto >> Xcode.app: 3.1.4 >> Xcode command-line tools: 3.1.4 >> atlas-3.10.1-1 is set to build with only one job. >> >> /sw/src/fink.build/atlas-3.10.1-1/gcc -V 2>&1 >> bin/INSTALL_LOG/ERROR.LOG >> gcc: error: unrecognized command line option '-V' >> gcc: fatal error: no input files >> compilation terminated. >> make[4]: [error_report] Error 1 (ignored) >> /sw/src/fink.build/atlas-3.10.1-1/gcc --version 2>&1 >> >> bin/INSTALL_LOG/ERROR.LOG >> tar cf error_PPCG432.tar Make.inc bin/INSTALL_LOG/* >> bzip2 error_PPCG432.tar >> make[4]: Leaving directory `/sw/src/fink.build/atlas-3.10.1-1/darwin_bld' >> make[3]: Leaving directory `/sw/src/fink.build/atlas-3.10.1-1/darwin_bld' >> make[2]: Leaving directory `/sw/src/fink.build/atlas-3.10.1-1/darwin_bld/bin' >> Error report error_<ARCH>.tgz has been created in your top-level ATLAS >> directory. Be sure to include this file in any help request. >> cat: ../../CONFIG/error.txt: No such file or directory >> cat: ../../CONFIG/error.txt: No such file or directory >> make[1]: *** [build] Error 255 >> make[1]: Leaving directory `/sw/src/fink.build/atlas-3.10.1-1/darwin_bld' >> make: *** [build] Error 2 >> ### execution of /tmp/fink.LLBX5 failed, exit code 2 >> ### execution of /tmp/fink.0tS6C failed, exit code 2 >> Removing runtime build-lock... >> Removing build-lock package... >> /sw/bin/dpkg-lockwait -r fink-buildlock-atlas-3.10.1-1 >> (Reading database ... 283428 files and directories currently installed.) >> Removing fink-buildlock-atlas-3.10.1-1 ... >> Failed: phase compiling: atlas-3.10.1-1 failed >> >> >> On Apr 10, 2013, at 14:59, Dominique Dhumieres <domi...@lps.ens.fr> wrote: >> >>> Jack, >>> You can get the complete log at >>> http://www.lps.ens.fr/~dominiq/fink/atlas.log.bz2 >>> >>> Dominique >>> ------------------------------------------------------------------------------ Precog is a next-generation analytics platform capable of advanced analytics on semi-structured data. The platform includes APIs for building apps and a phenomenal toolset for data science. Developers can use our toolset for easy data analysis & visualization. Get a free account! http://www2.precog.com/precogplatform/slashdotnewsletter _______________________________________________ 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