[mpich-discuss] mpich2 compilation with intel only

Wed Jun 27 16:53:04 CDT 2012

Whoops...you're right, that was supposed to be ifort.

Here's the actual successful configure script:

#! bin/sh

CDIR=/opt/intel/
FDIR=/opt/intel/fc/10.1.008/

sh ${CDIR}/bin/iccvars.sh ia32
sh ${CDIR}/bin/compilervars.sh ia32
sh ${FDIR}/bin/ifortvars.sh

./configure \
     --prefix=/Volumes/SecondHD/rieth/mpich2_intel \
     CC=icc \
     CFLAGS="-m32 -I${CDIR}/include" \
     CXX=icpc \
     CXXFLAGS="-m32 -I${CDIR}/include" \
     FC=ifort \
     FCFLAGS="-m32 -I${FDIR}/include" \
     F77=ifort \
     FFLAGS="-m32 -I${FDIR}/include" \
     LDFLAGS="-L${CDIR}/lib -L${FDIR}/lib" \
     2>&1 | tee c.txt

But my previous question still stands; even though I see the line  
"checking whether we are using the GNU C compiler... yes" in my  
configure output, have I still actually built an Intel version of  
mpich2?

Thanks,
Dave

On Jun 27, 2012, at 5:30 PM, Anthony Chan wrote:

>
> I find odd that you configure with cc=icc F77=ifort and FC=gfortran.
> Is there any reason you choose FC to be gfortran instead of ifort ?
>
> A.Chan
>
> ----- Original Message -----
>> Hmmm, actually now that I try it, I am not able to compile a "hello
>> world" program with -no-gcc and #include <stdarg.h> together.
>>
>> If I remove -no-gcc from my configure shell script, the configure is
>> successful, which leads me to my next question:
>>
>> Is this still considered a true Intel build, even if I've used gcc to
>> build it? From here on will the Intel compilers will run my mpi  
>> calls,
>> or will the GNU compiler set continue to be used?
>>
>> Thanks,
>> Dave
>>
>>
>>
>> On Jun 26, 2012, at 4:48 PM, Dave Goodell wrote:
>>
>>> Looks like your intel compiler is broken somehow when compiling
>>> under "-no-gcc". The "icc" compiler is finding the system-
>>> installed, gcc-specific stdarg.h which is blowing up with a
>>> "__MWERKS_" error message. My guess is that it won't complain if
>>> you drop "-no-gcc" because this will cause icc to #define "__GNUC__"
>>> and similar symbols that will make the system header happier.
>>>
>>> Are you *sure* that the "-no-gcc" option is what you want? Will it
>>> compile simple C programs that "#include <stdarg.h>" without any
>>> involvement with MPICH2?
>>>
>>> -Dave
>>>
>>> On Jun 25, 2012, at 4:50 PM MDT, David Riethmiller wrote:
>>>
>>>> Hi -
>>>>
>>>> I'm having trouble compiling mpich2 version 1.4.1p1 on a Mac OSX
>>>> 10.5 system, using only the Intel compilers. I specifically want
>>>> to avoid the GNU compilers, in order to match the software
>>>> available on a supercomputer to which I will port later.
>>>>
>>>> When running my configure script, I hit the following error:
>>>>
>>>> ...
>>>> checking for size of Fortran type integer... configure: WARNING:
>>>> Unable to compile the C routine for finding the size of a integer
>>>>
>>>> checking for size of Fortran type real... configure: WARNING:
>>>> Unable to compile the C routine for finding the size of a real
>>>>
>>>> checking for size of Fortran type double precision... configure:
>>>> WARNING: Unable to compile the C routine for finding the size of a
>>>> double precision
>>>>
>>>> checking whether integer*1 is supported... yes
>>>> checking whether integer*2 is supported... yes
>>>> checking whether integer*4 is supported... yes
>>>> checking whether integer*8 is supported... yes
>>>> checking whether integer*16 is supported... no
>>>> checking whether real*4 is supported... yes
>>>> checking whether real*8 is supported... yes
>>>> checking whether real*16 is supported... yes
>>>> configure: error: Unable to configure with Fortran support because
>>>> configure could not determine the size of a Fortran INTEGER.
>>>> Consider setting CROSS_F77_SIZEOF_INTEGER to the length in bytes of
>>>> a Fortran INTEGER
>>>>
>>>>
>>>>
>>>>
>>>> I've saved my configure command as a shell script:
>>>>
>>>> #! bin/sh
>>>>
>>>> CDIR=/opt/intel/
>>>> FDIR=/opt/intel/fc/10.1.008/
>>>>
>>>> sh ${CDIR}/bin/iccvars.sh ia32
>>>> sh ${CDIR}/bin/compilervars.sh ia32
>>>> sh ${FDIR}/bin/ifortvars.sh
>>>>
>>>> ./configure \
>>>>  --prefix=/Volumes/SecondHD/rieth/mpich2_intel \
>>>>  CC=icc \
>>>>  CFLAGS="-m32 -no-gcc -I${CDIR}/include -L${DYLD_LIBRARY_PATH}" \
>>>>  LDFLAGS="-L${CDIR}/lib -L${FDIR}/lib -L${DYLD_LIBRARY_PATH}" \
>>>>  F77=ifort \
>>>>  FFLAGS="-m32 -no-gfortran -I${FDIR}/include -L$
>>>> {DYLD_LIBRARY_PATH}" \
>>>>  CXX=icpc \
>>>>  CXXFLAGS="-m32 -no-gcc -I${CDIR}/include -L${DYLD_LIBRARY_PATH}"
>>>>  \
>>>>  LDFLAGS="-L${CDIR}/lib -L${FDIR}/lib -L${DYLD_LIBRARY_PATH}" \
>>>>  2>&1 | tee c.txt
>>>>
>>>>
>>>> I'm attaching the output files. Can anyone see what I'm doing
>>>> wrong?
>>>>
>>>> Thanks very much,
>>>> Dave
>>>>
>>>>
>>>>
>>>> <c.txt><config.log><config.system>
>>>>
>>>>
>>>> -------------------------------------------------
>>>> David A. Riethmiller
>>>> Ph.D. Candidate, Astrophysical Institute
>>>> Ohio University
>>>>
>>>> Clippinger Labs 338
>>>> http://www.phy.ohiou.edu/~rieth/
>>>>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> mpich-discuss mailing list mpich-discuss at mcs.anl.gov
>>>> To manage subscription options or unsubscribe:
>>>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>>>
>>> _______________________________________________
>>> mpich-discuss mailing list mpich-discuss at mcs.anl.gov
>>> To manage subscription options or unsubscribe:
>>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>>
>> -------------------------------------------------
>> David A. Riethmiller
>> Ph.D. Candidate, Astrophysical Institute
>> Ohio University
>>
>> Clippinger Labs 338
>> http://www.phy.ohiou.edu/~rieth/
>>
>>
>>
>>
>>
>> _______________________________________________
>> mpich-discuss mailing list mpich-discuss at mcs.anl.gov
>> To manage subscription options or unsubscribe:
>> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
> _______________________________________________
> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
> To manage subscription options or unsubscribe:
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss