Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

WRF-Chem Version 4.6 stuck (no error) in compilation with KPP for a day

acgvar

New member
Good day from Tokyo.

Our university supercomputer was recently upgraded, so it's also high time to compile the updated versions of WRF.

Following the procedure for compiling WRF-Chem with the KPP seems to go smoothly (see attached compile.log).

However, for more than a day of leaving it to compile, it remains stuck at the following procedure (i.e., stuck at the bottom of the compile.log),
`./compile em_real >& compile.log` and `./compile -j 8 em_real >& compile.log` are both showing similar tendencies.

Code:
mpif90 -o module_kpp_mozart_mosaic_4bin_Integr.o -c -O3 -fast -Mvect=noaltcode -Mprefetch=distance:8  -w -Mfree -byteswapio    -I../dyn_em -module /source_folder/compilation/WRF-CHEM/4.6/main -I/source_folder/compilation/WRF-CHEM/4.6/external/esmf_time_f90  -I/source_folder/compilation/WRF-CHEM/4.6/main -I/source_folder/compilation/WRF-CHEM/4.6/external/io_netcdf -I/source_folder/compilation/WRF-CHEM/4.6/external/io_int -I/source_folder/compilation/WRF-CHEM/4.6/frame -I/source_folder/compilation/WRF-CHEM/4.6/share -I/source_folder/compilation/WRF-CHEM/4.6/phys -I/source_folder/compilation/WRF-CHEM/4.6/wrftladj -I/source_folder/compilation/WRF-CHEM/4.6/chem -I/source_folder/compilation/WRF-CHEM/4.6/inc -I/source_folder/dependencies_nvhpc/netcdf4/include  -r4 -i4  module_kpp_mozart_mosaic_4bin_Integr.f90

Any advice to overcome or debug this issue?

Supporting detail:

An excerpt of the `configure.wrf` is as follows,
Code:
DESCRIPTION     =       NVHPC ($SFC/$SCC)
DMPARALLEL      =        1
OMPCPP          =       # -D_OPENMP
OMP             =       # -mp -Minfo=mp -Mrecursive
OMPCC           =       # -mp
SFC             =       nvfortran
SCC             =       nvc
CCOMP           =       nvc
DM_FC           =       mpif90
DM_CC           =       mpicc
FC              =       $(DM_FC)
CC              =       $(DM_CC) -DFSEEKO64_OK 
LD              =       $(FC)
RWORDSIZE       =       $(NATIVE_RWORDSIZE)
PROMOTION       =       -r$(RWORDSIZE) -i4
ARCH_LOCAL      =       -DNONSTANDARD_SYSTEM_SUBR  -DWRF_USE_CLM $(NETCDF4_IO_OPTS)
CFLAGS_LOCAL    =       -w -O3
LDFLAGS_LOCAL   =
CPLUSPLUSLIB    =
ESMF_LDFLAG     =       $(CPLUSPLUSLIB)
FCOPTIM         =       -O3 -fast -Mvect=noaltcode -Mprefetch=distance:8 # -Mfprelaxed # -Minfo=all =Mneginfo=all
FCREDUCEDOPT  =       $(FCOPTIM)
FCNOOPT               =       -O0
FCDEBUG         =       # -g $(FCNOOPT)  # -C -Ktrap=fp -traceback
FORMAT_FIXED    =       -Mfixed
FORMAT_FREE     =       -Mfree
 

Attachments

  • compile.log
    1.4 MB · Views: 2
A quick fix for this problematic compilation was to stop the compilation, and run that last command with minimal optimization, or for the above case, enter the folder,

Code:
mpif90 -o module_kpp_mozart_mosaic_4bin_Integr.o -c -O1    -I../dyn_em -module /source_folder/compilation/WRF-CHEM/4.6/main -I/source_folder/compilation/WRF-CHEM/4.6/external/esmf_time_f90  -I/source_folder/compilation/WRF-CHEM/4.6/main -I/source_folder/compilation/WRF-CHEM/4.6/external/io_netcdf -I/source_folder/compilation/WRF-CHEM/4.6/external/io_int -I/source_folder/compilation/WRF-CHEM/4.6/frame -I/source_folder/compilation/WRF-CHEM/4.6/share -I/source_folder/compilation/WRF-CHEM/4.6/phys -I/source_folder/compilation/WRF-CHEM/4.6/wrftladj -I/source_folder/compilation/WRF-CHEM/4.6/chem -I/source_folder/compilation/WRF-CHEM/4.6/inc -I/source_folder/dependencies_nvhpc/netcdf4/include  -r4 -i4  module_kpp_mozart_mosaic_4bin_Integr.f90

and once the above finished, without cleaning the compilation, resumed by typing the same compile command at the main directory.

1 or 2 other KPP-related scripts encounter the same problem. Did the same process.

The above is bad practice but builds the executables.

Will check how this affects performance.[/CODE]
 
Top