Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

cesm1.1.1 build pio question

Y

yunqian_zhu@colorado_edu

New Member
Hello,I'm using cesm1.1.1, what kind of setting should I use? I tried pio mentioned for cesm1.1.2 and cesm1.2.2.1, but they don't work.my env_mach_specific is:module load intel/17.0.1module load ncarenv/1.0module load mklmodule load ncarcompilers/0.3.5module load mpt/2.19if($MPILIB== "mpi-serial" ) then  module load netcdf/4.4.1.1else  module load netcdf-mpi/4.4.1.1   module load pnetcdf/1.8.0 The error message shows when I build the model:/usr/bin/perl ../bin/genf90.pl /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pio/pionfatt_mod.F90.in > pionfatt_mod.F90Can't open perl script "../bin/genf90.pl": No such file or directoryMakefile:125: recipe for target 'pionfatt_mod.F90' failed gmake: *** [pionfatt_mod.F90] Error 2 Thanks!
 
Y

yunqian_zhu@colorado_edu

New Member
Thank you for the quick reply. But it seems the link is not working:svn co http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_9/pio/ svn: E160013: Unable to connect to a repository at URL 'http://parallelio.googlecode.com/svn/trunk_tags/pio1_5_9/pio'svn: E160013: '/svn/trunk_tags/pio1_5_9/pio' path not foundThanks,
 
Y

yunqian_zhu@colorado_edu

New Member
Hello,Thank you. But it is still not working and showing the same pio error message. The actual folder of my cesm1.1.1 is: cesm1_1_1_ccmi34/. It might make a difference since it is ccmi version of cesm1.1.1?
Thanks,
 
J

jedwards

CSEG and Liaisons
Staff member
I think I see the problem, do this and try again:cp /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pio/genf90.pl  /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/
 
Y

yunqian_zhu@colorado_edu

New Member
Hello,They are the same folder. Do you mean something else?cp: ‘/gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pio/genf90.pl’ and ‘/gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pio/genf90.pl’ are the same fileThanks,
 
J

jedwards

CSEG and Liaisons
Staff member
cheyenne6: /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/piols genf90.pl genf90.plcheyenne6: /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pio:) ls ../bin  esmf_wrf_timemgr  mct  pio  pio_old  timingI don't see it there - am I looking at the correct source
 
Y

yunqian_zhu@colorado_edu

New Member
I did the following modification instead and the build is successful now.In Makefile: GENF90 := /gpfs/u/home/yunqian/cesm/cesm1_1_1_ccmi34/models/utils/pioIn pio_msg_getput_callbacks.F90.in, add:#define TYPETEXT 100#define TYPEREAL 101#define TYPEDOUBLE 102 #define TYPEINT 103
I'm able to run the model, however, the model crashes right away with the following error (/gpfs/fs1/scratch/yunqian/Kelut_addhetnuc_16_21km_log0.5um1.5um20um_Gaussian_1.5Tg/run/cesm.log.190320-140834):1: NetCDF: Numeric conversion not representable9: NetCDF: Numeric conversion not representable9: pio_support::pio_die:: myrank=          -1 : ERROR:37: NetCDF: Numeric conversion not representable13: NetCDF: Numeric conversion not representable17: NetCDF: Numeric conversion not representable21: NetCDF: Numeric conversion not representable25: NetCDF: Numeric conversion not representable29: NetCDF: Numeric conversion not representable41: NetCDF: Numeric conversion not representable33: NetCDF: Numeric conversion not representable9: pionfwrite_mod::write_nfdarray_double:         250 :9: NetCDF: Numeric conversion not representable17: pio_support::pio_die:: myrank=          -1 : ERROR:17: pionfwrite_mod::write_nfdarray_double:         250 :17: NetCDF: Numeric conversion not representable13: pio_support::pio_die:: myrank=          -1 : ERROR:13: pionfwrite_mod::write_nfdarray_double:         250 : 13: NetCDF: Numeric conversion not representable237:Image              PC                Routine            Line        Source237:cesm.exe           0000000001B8352D  Unknown               Unknown  Unknown237:cesm.exe           0000000001551CD1  pio_support_mp_pi         118  pio_support.F90237:cesm.exe           000000000154939E  pio_utils_mp_chec          74  pio_utils.F90237:cesm.exe           00000000016604C3  pionfwrite_mod_mp         250  pionfwrite_mod.F90.in237:cesm.exe           000000000162CFFC  piodarray_mp_writ         650  piodarray.F90.in237:cesm.exe           000000000162B09C  piodarray_mp_writ         223  piodarray.F90.in237:cesm.exe           0000000001630213  piodarray_mp_writ         293  piodarray.F90.in237:cesm.exe           00000000004B2B02  cam_history_mp_du        4334  cam_history.F90237:cesm.exe           00000000004944D0  cam_history_mp_ws        4651  cam_history.F90237:cesm.exe           0000000000488D1C  cam_comp_mp_cam_r         379  cam_comp.F90237:cesm.exe           0000000000479AE7  atm_comp_mct_mp_a         541  atm_comp_mct.F90237:cesm.exe           000000000040A756  ccsm_comp_mod_mp_        3485  ccsm_comp_mod.F90237:cesm.exe           00000000004258C8  MAIN__                     91  ccsm_driver.F90237:cesm.exe           0000000000407EDE  Unknown               Unknown  Unknown237:libc-2.19.so       00002AAAAC60EB25  __libc_start_main     Unknown  Unknown237:cesm.exe           0000000000407DE9  Unknown               Unknown  Unknown237:MPT ERROR: Rank 237(g:237) is aborting with error code 1.237:    Process ID: 37890, Host: r5i2n17, Program: /gpfs/fs1/scratch/yunqian/Kelut_addhetnuc_16_21km_log5um_Gaussian_15Tg/bld/cesm.exe237:    MPT Version: HPE MPT 2.19  12/07/18 05:31:15 Thanks,
 
Y

yunqian_zhu@colorado_edu

New Member
Thank you for the new file. The build is successful.However, when I run the model, it still shows the same NetCDF problem as I posted above.It is because the initial file that I use is not compatible with the mpt/2.19? I tried to change the NetCDF and pnetcdf options in env_mech_specific but they don't solve the problem.Could you suggest some debugging method for me? Thanks,
 
J

jedwards

CSEG and Liaisons
Staff member
That error indicates that you are trying to write values which are out of the range of the variable type - for example writing NaN in output will generate this error.  You can add a flag -debug minimal to the FFLAGS in your case Macros file to get a little more debug information or try rebuilding with DEBUG=TRUE As the nature of the problem has changed and the issue that opened this ticket has been resolved please open a new ticket if the problem persists.
 
M

mark@atmos_colostate_edu

Member
Hey Jim.I'm trying to compile spcam2_0-cesm1_1_1.  Per the instructions in this thread I moved pio to pio.old and downloaded pio 1_5_10.  env_mach_specifc looks like this:module load intel/17.0.1module load ncarenv/1.2module load mklmodule load ncarcompilers/0.4.1module load mpt/2.19if( $MPILIB == "mpi-serial" ) then  module load netcdf/4.4.1.1else  module load netcdf-mpi/4.4.1.1  module load pnetcdf/1.11.0 endif
But I get the following pio build errors:
/gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pio/piodarray.F90.in(1167): error #6460: This is not a field name that is defined in the encompassing structure.   [REQUEST_CNT]    File%request_cnt=File%request_cnt+1---------^/gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pio/piodarray.F90.in(1212): error #6385: The highest data type rank permitted is INTEGER(KIND=8).   [REQUEST_CNT]       allocate(array_of_requests(file%request_cnt), status(file%request_cnt))---------------------------------------^/gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pio/piodarray.F90.in(1212): error #6385: The highest data type rank permitted is INTEGER(KIND=8).   [REQUEST_CNT]       allocate(array_of_requests(file%request_cnt), status(file%request_cnt))-----------------------------------------------------------------^compilation aborted for /gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pio/piodarray.F90 (code 1)
Any ideas?
Thanks,Mark
 
J

jedwards

CSEG and Liaisons
Staff member
Hi Mark - I think that you just need to remove and regenerate the fortran files in directory /gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pioRemove any *.F90 files that have cooresponding *.F90.in files and rebuild.    Thanks for using the forum. 
 
M

mark@atmos_colostate_edu

Member
Thanks for replying so quickly, Jim.So I removed these files:rm alloc_mod.F90rm box_rearrange.F90rm iompi_mod.F90rm piodarray.F90rm pio_msg_getput_callbacks.F90rm pionfatt_mod.F90rm pionfget_mod.F90rm pionfput_mod.F90rm pionfread_mod.F90rm pionfwrite_mod.F90rm pio_spmd_utils.F90rm pio_support.F90rm rearrange.F90And did a clean build followed by a build and now I get this:gmake: *** No rule to make target 'pio_support.F90', needed by 'pio_support.d'.  Stop.
 
J

jedwards

CSEG and Liaisons
Staff member
Sorry, at the end of the Makefile add a linepio_support.F90: pio_support.F90.in and try again.
 
M

mark@atmos_colostate_edu

Member
Okay, did that and now I get this error:mpif90 -c -DLINUX -DMCT_INTERFACE -DHAVE_MPI -DFORTRANUNDERSCORE -DNO_R16 -DHAVE_F2008_CONTIGUOUS -DLINUX -DCPRINTEL -DHAVE_SLASHPROC -DSPMD -DHAVE_MPI -DUSEMPIIO -DSYSLINUX -D_NETCDF -I/glade/u/apps/ch/opt/netcdf-mpi/4.4.1.1/mpt/2.19/intel/17.0.1//include -D_PNETCDF -DTIMING -DPIO_GPFS_HINTS  -O2 -fp-model precise -convert big_endian -assume byterecl -ftz -traceback  -free  -I.  -I/glade/u/apps/ch/opt/netcdf-mpi/4.4.1.1/mpt/2.19/intel/17.0.1//include -I/glade/u/apps/ch/opt/pnetcdf/1.11.0/mpt/2.19/intel/17.0.1//include -I/glade/scratch/branson/spcam_nlev53/bld/mct/mct -I/glade/scratch/branson/spcam_nlev53/bld/mct/mpeu -I/glade/scratch/branson/spcam_nlev53/bld/pio -I/gpfs/u/home/branson/spcam2_0-cesm1_1_1/models/utils/pio -I/glade/scratch/branson/spcam_nlev53/bld/lib/include -I/glade/u/apps/ch/opt/netcdf-mpi/4.4.1.1/mpt/2.19/intel/17.0.1//include   -I/glade/u/apps/ch/opt/pnetcdf/1.11.0/mpt/2.19/intel/17.0.1//include -I../timing  pio_support.F90 ifort: error #10236: File not found:  'pio_support.F90'
Looking in the pio directory it seems like the step where the .F90.in files are used to create the .F90 files is not happening for some reason.  ????
 
M

mark@atmos_colostate_edu

Member
I corrected that.  The last three lines of my Makefile look like this:pio_spmd_utils.F90: pio_spmd_utils.F90.inpio_msg_getput_callbacks.F90 : pio_msg_getput_callbacks.F90.inpio_support.F90: pio_support.F90.in
 
You must log in or register to reply here.
Top