If you are still looking for a fortran compiler for the Intel Mac, try HPC (High Performance Computing) website. http://hpc.sourceforge.net/
It's author released an Intel binary of gfortran on Sun. Don't know how he mangaged to compile it. I've been using it and gcc 4.0.1 to TRY and build octave for my Macbook. The make seems to compile the fortran src OK, but I'm getting ld errors from library conflicts (but that's not a fortran compiler prob). I don't know if this will help your gcc build attempts. - Kirk Volland NPS, Monterey --- "David R. Morrison" <[EMAIL PROTECTED]> wrote: > > On Mar 15, 2006, at 7:15 AM, William Scott wrote: > > > > > Hi Josh: > > > > I'm afraid I am completely unable to help due to > cluelessness, and > > you've definitely made far more progress than I > would have been > > capable of. I > > naively put together the ccp4-gfortran fink > package with the hope > > that if it compiled on OS X ppc (It does) it might > work on the > > intel processor, which I have no access to (I'm > just a poor assoc. > > professor). > > > > Having said that, it looks like you are within > striking distance of > > getting this to work, and I would accept with > extreme gratitutde > > any changes you might be able to make to get it to > fly. > > > > It might be worth posting this to fink dev. In > fact, I'll cc it now > > to get the ball rolling.... > > > > The incredibly knowledgable and helpful David > Morrison is also at > > Duke, but I doubt he makes house calls. > > Hi Bill. Actually, I'm on sabbatical this year, > just up the road > from you in Berkeley. > > > > > Thanks for your perseverence. > > > > Bill > > > > > > On Wed, 15 Mar 2006 [EMAIL PROTECTED] wrote: > > > >> Hi Bill- > >> > >> Sorry to bother you with this... Thanks in > advance for any help > >> you have > >> time to provide... > >> > >> I just spent a day or so trying to get > ccp4-gfortran to compile on an > >> intel iMac... Didn't work but I can at least > pass along some > >> information > >> which may be useful in moving the right > direction. I apologize if > >> some of > >> these "fixes" are on the heavy-handed side. > >> > >> 1) I can't get the fink gcc/g95/g77 to compile on > the iMac, and so > >> I had > >> to change the requirements for ccp4 to allow the > use of the > >> placeholder > >> gcc4.0 for the Apple version of the compilers. > >> > >> 1.1) The other dependencies for ccp4 (e.g. blt) > will compile > >> under fink > >> on the intel iMac > >> > >> 2) Since I couldn't get fink to compile a usable > fortran compiler, > >> I ended > >> up downloading a gfortran binary from the wiki. > Much to my surprise, > >> this appears to work... At least it doesn't > complain when asked to > >> compile something. And it will compile a > runnable "Hello, world" > >> style > >> program. > > We've had other reports of a similar nature. What > you have, I > suspect, is a ppc version of gfortran, which your > intel Mac is happy > to run under Rosetta. The binary which it is > producing is a ppc > binary which will also run under Rosetta. > > Packaging this setup for fink is an idea I've been > toying with, but > after discussions with some of the opendarwin folks, > I plan to hold > off on this because packaging it will be hard and > there is a chance > that we'll have a functional native gfortran on > intel some time soon. > > >> > >> 3) When I started to install ccp4, I ran into a > similar error with > >> several > >> different programs, all during the configure > phase... Most of the > >> configure scripts appear to check for the > "Fortran name mangling > >> scheme" > >> and are unable to determine what it is on the > iMac. Most of them > >> print a > >> warning here but several give an error and fail. > The ones which fail > >> (all in the x-windows subdirectory) are xdl_view > (xdl_view/src/ > >> configure), > >> Rotgen (Rotgen/configure), and Mosflm > (Mosflm/configure). It > >> appears that > >> these programs are all going to be compiled with > the > >> -fno-second-underscore flag, which hopefully will > prevent there > >> from being > >> a problem. So I added a few lines to the > ccp4-gfortran.info along > >> the > >> lines of: > >> > >> perl -pi.bak -e 's|exit 1|echo hi|g' > x-windows/Mosflm/configure > >> > >> This is obviously not a delicate way of doing > this, but got things > >> moving > >> forward again. > >> > >> 4) At this point, things seemed to move forward > for a long time, but > >> during compilation of the actual ccp4 programs I > got a lot of > >> errors along > >> the lines of: > >> > >> /usr/bin/ld: warning ../lib/src/libccp4c.a > archive's cputype (7, > >> architecture i386) does not match cputype (18) > for specified -arch > >> flag: ppc (can't load from it) > >> /usr/bin/ld: warning ../lib/src/libmmdb.a > archive's cputype (7, > >> architecture i386) does not match cputype (18) > for specified -arch > >> flag: ppc (can't load from it) > >> /usr/bin/ld: warning > /sw/src/fink.build/ccp4-gfortran-6.0-1002/ > >> ccp4-6.0/lib/ccif/libccif.a archive's cputype (7, > architecture > >> i386) does not match cputype (18) for specified > -arch flag: ppc > >> (can't load from it) > >> /usr/bin/ld: warning > /usr/lib/gcc/i686-apple-darwin8/4.0.1/ > >> libgcc.a archive's cputype (7, architecture i386) > does not match > >> cputype (18) for specified -arch flag: ppc (can't > load from it) > >> /usr/bin/ld: Undefined symbols: > >> _ccperr_ > >> _ccpfyp_ > >> _ccprcs_ > >> _ccpupc_ > >> _lkyin_ > >> _lkyset_ > >> _lrassn_ > >> _lropen_ > >> _lrreff_ > >> _lrrefm_ > >> _lrsymm_ > >> _mtzini_ > >> _parser_ > >> collect2: ld returned 1 exit status > >> make[1]: [abs] Error 1 (ignored) > >> > >> So it looks like there is a (hopefully > straightforward) problem > >> with the > >> -arch flags for the libraries that are being > built earlier. > >> > > Since gfortran is building the ppc architecture, > anything else which > is being linked must also be built for the ppc > architecture. (It's > === message truncated === ------------------------------------------------------- This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 _______________________________________________ Fink-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/fink-devel
