CESM runtime error - Model stops after initializing the atmosphere component

[nitkbhat@gmail_com]

I have successfully built the F compset of CESM 1.0.4. When I submit the job, the model does not complete. It gives the following message in the pbs output file.  I think the error is because of some UNSET variables in ccsm. I see the following message in the atm log. Please find attached the output file, the cpl log and the atm log.  [nitin@master case]$ cat /home/nitin/CESM/cesm1_0_4/CaseF/case/case/run/atm.log.150116-210123 CAM atmosphere model initialization Read in cam_inparm namelist from: atm_in REPRO_SUM_SETOPTS: Using fixed-point-based (scalable) reproducible distributed sum algorithm                     and not comparing with floating point algorithms. SPMD SWAP_COMM OPTIONS:   swap_comm_protocol =            4   swap_comm_maxreq   =          128 SPMD FLOW CONTROL GATHER OPTION:   fc_gather_flow_cntl =           64 cld_sediment_readnl: cldsed_ice_stokes_fac =   0.500000000000000 solar_data_readnl: solar_const (W/m2) =    1361.27000000000  ------------------------------------------      *** INPUT VARIABLES (CAM_INPARM) ***  ------------------------------------------          Initial run  ********** CASE = case ********** UNSET                                                                                                                                                                                        Initial dataset is: Thanks,Nitin 

 

[nitkbhat@gmail_com]

Are the following variables requiring a specific value? ./env_conf.xml:./env_conf.xml:./env_conf.xml:./env_conf.xml:./env_conf.xml:./env_conf.xml:./env_run.xml:./env_run.xml:./env_run.xml:./env_run.xml: Thanks,Nitin 

 

[santos]

Can you attach the file beginning with "cesm.log" as well if you have one (it may start with "ccsm.log" instead)?

 

[nitkbhat@gmail_com]

I saw the glibc detected error in the ccsm.log. I have attached that file. Snippet from the ccsm log:  126  node1127  node1 Opened existing file /home/nitin/CESM_NEW/input_data//atm/cam/inic/fv/cami_0000-01-01_1.9x2.5_L26_c0 70408.nc       65536 Opened existing file /home/nitin/CESM_NEW/input_data//atm/cam/topo/USGS-gtopo30_1.9x2.5_remap_c05060 2.nc      131072*** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x00000000104df160 ****** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x000000000d219b80 ****** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x000000000d722590 ****** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x000000000c2692c0 ****** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x000000000cfc1210 ****** glibc detected *** ./ccsm.exe: double free or corruption (out): 0x000000000dafe830 *** I am using intel compiler family (mpiicc, mpiifort). [nitin@master ~]$ mpiifort -VIntel(R) Fortran Intel(R) 64 Compiler XE for applications running on Intel(R) 64, Version 14.0.2.144 Build 20140120Copyright (C) 1985-2014 Intel Corporation.  All rights reserved.   What is the possible cause of this error?   Thanks,Nitin

 

[jedwards]

Try this:https://bb.cgd.ucar.edu/fv-dycore-spmddynf90-and-intel-14x-compiler

 

[nitkbhat@gmail_com]

Hello, I was unable to access the page as mentioned in the link. It said that I am not authorized to access the link (https://svn-ccsm-models.cgd.ucar.edu/cam1/branches/fvspdmdyn_fix_cam5_3_45/models/atm/cam/src/dynamics/fv/spmd_dyn.F90).Also, I couldn't locate the Depends.Intel file as mentioned in the first solution of the link. However, as the solution suggested the use of -O1 instead of -O2, I changed the FFLAGS_OPT  variable to -O1 from -O2. On running the script again, I got a successful completion of the atmosphere component and also the rest of the model. Thanks. 

 

[jedwards]

I've attached the spmd_dyn.F90 file.    You won't need the Depends.intel file with this, but for future reference if you create that file in your case directory it will be read and used in the build.  

 

[alisonfangdada@icloud_com]

Hello, Icame into the same problem, and which place have you changed the  FFLAGS_OPT from -O2 to -O1?