[MITgcm-support] Error by ifort 9.1 compiler!

Van Thinh Nguyen vtnguyen at moisie2.math.uwaterloo.ca
Wed Jan 10 16:14:51 EST 2007


Hi Baylor & Patrick,

I already fixed the problem, which is due to the data format input.
Thanks for all your helps.

Vth

-------------------------------------------------------

On Wed, 10 Jan 2007, Patrick Heimbach wrote:

>
> Hi Van Thinh,
>
> you'd need to be more specific re. error messages, i.e.
> what's the content of STDERR.0000,
> and last few lines of STDOUT.0000
> to figure out reasons for crash.
> Error msg. you quoted doesn't say much more than
> "model has crashed".
>
> -Patrick
>
>
>
> On Jan 10, 2007, at 2:51 PM, Van Thinh Nguyen wrote:
>
>> Hi Baylor,
>> 
>> Thanks for helping me. Follwing your suggestion, I matched up the oftfiles 
>> as follows:
>> ----------------
>> CPP='/usr/bin/cpp  -traditional -P'
>> 
>> # the '-convert big_endian' is crucial on this platform
>> # -D_BYTESWAPIO doesn't work
>>    FFLAGS='-mp -132 -r8 -i4 -w95 -W0 -WB -convert big_endian -assume 
>> byterecl'
>> 
>> FOPTIM='-O5 -align'
>> #NOOPTFILES='ini_curvilinear_grid.F'
>> #NOOPTFLAGS='-O3'
>> 
>> #P3  FOPTIM     = ( $FOPTIM '-tpp6 -xWKM' )
>> #P4  FOPTIM     = ( $FOPTIM '-tpp7 -xWKM' )
>> LIBS='-L/usr/lib -lmpi'
>> 
>> FC='/opt/intel/fc/9.1.040/bin/ifort'
>> CC='/opt/intel/cc/9.1.045/bin/icc'
>> DEFINES='-DALLOW_USE_MPI -DALWAYS_USE_MPI -DWORDLENGTH=4'
>> 
>> # These paths will likely change to newer versions of ifort
>> INCLUDES='-I/opt/intel/fc/9.1.040/include'
>> INCLUDEDIRS='/opt/intel/fc/9.1.040/include'
>> MPIINCLUDEDIR='/usr/include'
>> ------------
>> 
>> But I still have the same problem as before. I already tried with & without 
>> " -assume byterecl", reduced O5 to O3.
>> 
>> You may have another ideas,
>> 
>> Thanks a lot!
>> 
>> Van Thinh
>> 
>> 
>> -----------------------------------------------------
>> 
>> On Wed, 10 Jan 2007, Baylor Fox-Kemper wrote:
>> 
>>> Hi Vth,
>>> Take a look at the
>>> linux_ia64_ifort+mpi_altix_gfdl_bfk
>>> 
>>> optfiles.  It works with ifort 9.1, I know.  However, there is some mess 
>>> with which MPI library you use...
>>> 
>>> I think the critical lines are:
>>> 
>>> FFLAGS='-mp -132 -r8 -i4 -w95 -W0 -WB -convert big_endian'
>>> LIBS='-lmpi -lnetcdf'
>>> DEFINES='-DALLOW_USE_MPI -DALWAYS_USE_MPI -DWORDLENGTH=4'
>>> 
>>> and you have to be sure to get all the include paths for the mpi library.
>>> 
>>> -Baylor
>>> 
>>> On Jan 10, 2007, at 12:03 PM, Van Thinh Nguyen wrote:
>>> 
>>>> Hi all,
>>>> Anyone can help me to fix the following error when I run a MPI on
>>>> Altix with Linux SuSE OP (compiler: ifort 9.1)
>>>> ----
>>>> MPI: MPI_COMM_WORLD rank 4 has terminated without calling MPI_Finalize()
>>>> MPI: aborting job
>>>> ----
>>>> Last time when I upgraded the compiler from ifort 7. to ifort 8., I got 
>>>> the same error as above, Patrick advised me to add
>>>> "-assume byterecl" on the FLAGS line, then I got rid of this.
>>>> This time I just upgrade the compiler from ifort 8. to ifort 9.1, I also 
>>>> did add this option to the FLAGS line, but it doesn't help.
>>>> Thanks,
>>>> Vth
>>>> ----------------------------------------------------
>>>> _______________________________________________
>>>> MITgcm-support mailing list
>>>> MITgcm-support at mitgcm.org
>>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>> 
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>
> ---
> Dr Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
> MIT | EAPS, 54-1518 | 77 Massachusetts Ave | Cambridge, MA 02139, USA
> FON: +1-617-253-5259 | FAX: +1-617-253-4464 | SKYPE: patrick.heimbach
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support



More information about the MITgcm-support mailing list