[ncl-install] Installing ncarg on OS X 10.9 with ifort

[Dave Lane]

In these continuing troubles of compiling ncarg (without ncl) on a Mac 
with OS X 10.9 and ifort/gcc, the latest hangup is pasted below. I have 
tired with gcc 4.3 and 4.9 and ifort 2011 and 2013 versions. Same issue. 
Any ideas?  As said before I have been able to compile cleanly with 
gfortran/gcc (4.9) on the same machine (the researcher needs the ifort 
version for the debugger).

--- Dave

Continuing in ./ncarg2d/src/libncarg_gks
ranlib libncarg_gks.a
Making ./ncarg2d/src/liboptional
Making ./ncarg2d/src/liboptional/conraq
ifort -m64 -openmp  -O    -c -o conraq.o conraq.f
ifort -m64 -openmp  -O    -c -o contor.o contor.f
gcc-mp-4.3 -bind_at_load -r -o libconraq.o conraq.o contor.o
ld: warning: unexpected dylib (/opt/local/lib/gcc43/libgcc_s.10.5.dylib) 
on link line
ld: warning: unexpected dylib (/usr/lib/libSystem.dylib) on link line
0  0x102491abb  __assert_rtn + 144
1  0x1024ba2c6 ld::tool::SectionRelocationsAtom<x86>::sectSymNum(bool, 
ld::Atom const*) + 0
2  0x1024a8180 
ld::tool::SectionRelocationsAtom<x86_64>::encodeSectionReloc(ld::Internal::FinalSection*, 
ld::tool::SectionRelocationsAtom<x86_64>::Entry const&, 
std::__1::vector<macho_relocation_info<Pointer64<LittleEndian> >, 
std::__1::allocator<macho_relocation_info<Pointer64<LittleEndian> > > 
 >&) + 86
3  0x1024dd7b8 ld::tool::SectionRelocationsAtom<x86_64>::encode() + 62
4  0x1024af4ca 
ld::tool::OutputFile::updateLINKEDITAddresses(ld::Internal&) + 454
5  0x1024aa0de  ld::tool::OutputFile::write(ld::Internal&) + 138
6  0x1024927f0  main + 1012
7  0x7fff929dc5fd  start + 1
A linker snapshot was created at:
         /tmp/libconraq.o-2014-08-23-150659.ld-snapshot
ld: Assertion failed: (_machoSection != 0), function machoSection, file 
src/ld/ld.hpp, line 714.
collect2: ld returned 1 exit status
make[5]: *** [libconraq.o] Error 1


On 16/09/2014 3:45 PM, Dave Lane wrote:
> Ok, I have the M_PI problem resolved by patching the source as 
> suggested. I'll write back tomorrow with my next hangup :-)
>
> --- Dave
>
> On 16/09/2014 2:43 PM, Dave Lane wrote:
>> Hi Rick,
>>
>> What would the appropriate definition of M_PI be?
>>
>> Also note that I'm compiling with V4.9 of gcc and ifort. The same 
>> file must have compiled ok when I built with gcc/gfortran (both 4.9).
>>
>
>
>
> _______________________________________________
> ncl-install mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-install

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-install/attachments/20140923/ec0cc63a/attachment.html