Building CAM4/CAM5 on SGI octane III

[sirajkhan78@gmail_com]

Hi,

I am using CAM4 more than year and have installed it on two clusters for my use. We recently bought new cluster SGI octane III (Linux admin 2.6.32.13-0.4-default #1 SMP 2010-06-15 12:47:25 +0200 x86_64 x86_64 x86_64 GNU/Linux) for our lab and I try to install CAM4 in this small cluster (5 nodes with total of 60 cores). I installed netcdf 4, intel compiler 11 and mpich 2 but when i try to build the model I get some error as below. I also tried on serial mode but same error. Also same error is there if I use CAM5. This mean that there is some machine issue to compile ncdio.F90 file.

Can any body check this for me.

fort -c -I. -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/chemistry/bulk_aero -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/chemistry/utils -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/physics/cam -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/dynamics/fv -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/cpl_mct -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/control -I/data1/siraj-work/models/ccsm4/models/csm_share/shr -I/data1/siraj-work/models/ccsm4/models/csm_share/dshr -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/utils -I/data1/siraj-work/models/ccsm4/models/utils/timing -I/data1/siraj-work/models/ccsm4/models/utils/pio -I/data1/siraj-work/models/ccsm4/models/utils/mct/mpeu -I/data1/siraj-work/models/ccsm4/models/utils/mct/mct -I/data1/siraj-work/models/ccsm4/models/utils/esmf_wrf_timemgr -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/utils/pilgrim -I/data1/siraj-work/models/ccsm4/models/drv/driver -I/data1/siraj-work/models/ccsm4/models/drv/shr -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/utils/cam_dom -I/data1/siraj-work/models/ccsm4/models/atm/cam/src/utils/cam_dom/cpl_mct -I/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/cpl_mct -I/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main -I/data1/siraj-work/models/ccsm4/models/lnd/clm/src/biogeophys -I/data1/siraj-work/models/ccsm4/models/lnd/clm/src/biogeochem -I/data1/siraj-work/models/ccsm4/models/lnd/clm/src/riverroute -I/data1/siraj-work/models/ccsm4/models/ice/cice/src/drivers/cpl_mct -I/data1/siraj-work/models/ccsm4/models/ice/cice/src/drivers/cpl_share -I/data1/siraj-work/models/ccsm4/models/ice/cice/src/mpi -I/data1/siraj-work/models/ccsm4/models/ice/cice/src/source -I/data1/siraj-work/models/ccsm4/models/glc/sglc/cpl_mct -I/usr/local/include -I/data1/siraj-work/models/mpich2-1.2.1p1/include -I/usr/local/include -DNO_SHR_VMATH -DSEQ_MCT -DFORTRANUNDERSCORE -DCO2A -DMAXPATCH_PFT=numpft+1 -DLSMLAT=1 -DLSMLON=1 -DCOUP_DOM -DPLON=144 -DPLAT=96 -DPLEV=26 -DPCNST=3 -DPCOLS=16 -DPTRM=1 -DPTRN=1 -DPTRK=1 -DSTAGGERED -DCCSMCOUPLED -Dcoupled -Dncdf -DNCAT=1 -DNXGLOB=144 -DNYGLOB=96 -DNTR_AERO=0 -DBLCKX=24 -DBLCKY=96 -DMXBLCKS=1 -D_USEBOX -D_NETCDF -DNO_MPI2 -DSPMD -DLINUX -fp-model precise -ftz -convert big_endian -O2 /data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90
/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90(5057): error #6906: The value of the integer is either too great or too small, and overflow/underflow occurred. [O'0777610000000000000000']
cols(:) = nan
---------------------^
/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90(5058): error #6906: The value of the integer is either too great or too small, and overflow/underflow occurred. [O'0777610000000000000000']
data_offset = nan
---------------------^
/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90(5092): error #6906: The value of the integer is either too great or too small, and overflow/underflow occurred. [O'0777610000000000000000']
pfts(:) = nan
---------------------^
/data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90(5093): error #6906: The value of the integer is either too great or too small, and overflow/underflow occurred. [O'0777610000000000000000']
data_offset = nan
---------------------^
compilation aborted for /data1/siraj-work/models/ccsm4/models/lnd/clm/src/main/ncdio.F90 (code 1)
gmake: *** [ncdio.o] Error 1


Siraj

 

[eaton]

I see in later code versions that the variable 'nan' which the compiler is complaining about has been replace by the variable 'bigint' which is also available by use association from the nanmod module.

 

[sirajkhan78@gmail_com]

Hi Eaton,

Thanks for the reply but I didn't get your answer. Do you mean a new version of Intel compiler? or the new version of CAM5 model ? Actually I tried with CAM4 and CAM5 but every time I get the same error. Can you further guide me how to get rid of this error. I am stuck on this machine.

Regrads
Siraj

 

[eaton]

Sorry I wasn't clear. In the file ncdio.F90 you can fix this problem by replacing the variable nan by the variable bigint.

In the cesm1_0_3 code (which was released today) the file ncdio.F90 has been replaced by ncdio_pio.F90, and in that file all occurances of nan have been replaced by bigint.