On Mar 13, 2008, at 10:27 PM, Alexander K. Hansen wrote: > On Thursday 13 March 2008 08:12:14 pm Ben Abbott wrote: >> On Mar 13, 2008, at 9:17 AM, Alexander Hansen wrote: >>> Ben Abbott wrote: >>>> The current octave.info files still points to gcc4.2 library. >>>> Should this not be updated to respect gcc4.3? >>>> >>>> ConfigureParams: FLIBS=%p/lib/gcc4.2/lib/libgfortran.dylib F77=%p/ >>>> bin/gfortran --infodir='${prefix}/share/info' --mandir='${prefix}/ >>>> share/man' --libexecdir='${prefix}/lib' -enable-shared -enable-dl >>>> -- >>>> disable-static --without-mpi --with-hdf5 --with-fftw >>>> >>>> I expect there are similar occurances in other packages as well. >>>> >>>> Ben >>> >>> It was updated without anybody informing me, but yes, that update >>> should be made. >> >> I notice gcc43 includes a conflict with gcc42 >> >> Conflicts: gcc4, gcc42 >> >> I've been using a local octave.info file to build a bleeding-edge >> version of octave. Recently, I've been having problems that may be >> related to gfortran 4.2, I so I thought I'd try 4.3. >> >> Unfortunately, modifying the .info files for packages that octave >> depends on wasn't so simple (atlas for example). >> >> So if I install gcc43, I'll not be able to build anything that >> depends >> upon gcc42, unless I modifiy each respective .info file:-( >> >> What is the usual procedure when upgrading between gccX.Y to gccX.Z? >> >> Ben > > You're kind of on your own for local stuff.
Sure, no problem there. > For most non-compiler things, any > dependency on gccX.Y is build-time, so it's not such a huge deal to > swap back > and forth. The reason I asked, is that I've been warned that gfortran was recenlty modified and that I might encounter problems if I mix compilers. The comment made is below (link also provided). The ZDOTC function is a Fortran function returning double complex. There was a recent gfortran change in calling conventions for functions (not subroutines) that return complex variables. I suspect you are mixing the two methods here, compiling zqrqhv with a compiler that implements one of the methods, and using a version of ATLAS that was compiled to use the other. http://www.nabble.com/failed-build-with-current-mercurial-sources-to15896284.html Unfortunately, I believe I have dependencies compiled with gcc older than 4.2, and unless I'm mistaken XCode's lapack/blas also qualify. Any thoughts on how I might force rebuilds of all packages so that they are all compiled with gcc 4.2? Any idea which version of gcc included this change to gfortran? Ben ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Fink-users mailing list Fink-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fink-users