ERROR running cesm1_0_b11

[agustin@unex_es]

Hi every one, I am traying to run cesm1_0_b11 on a itanium machine with linux os and intel fortran 11.0.74. I ger the following errors when runngin the program

a) in ccsm.log:
.
.
.
(seq_frac_check) [atm init] sum min/max = 0.999999999999992673 1.00000000000000666
(seq_frac_check) [atm init] ofrac min/max = 0.00000000000000000 1.00000000000001177
(seq_frac_check) [atm init] sum min/max = 0.999999999999999334 1.00000000000000044
(seq_frac_check) [atm init] lfrin min/max = 0.00000000000000000 1.00000000000000000
(seq_frac_check) [atm init] sum min/max = 0.999999999999980238 1.0000(seq_frac_check) [(seq_frac_check) [atm init] sum ncnt/maxerr = 0 0.00000000000000000
(seq_frac_check) [atm init] sum ncnt/maxerr = 0 0.00000000000000000
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
ccsm.exe 4000000000EDD931 Unknown Unknown Unknown
ccsm.exe 40000000006813F0 Unknown Unknown Unknown
forrtl: severe (174): SIGSEGV, segmentation fault occurred

b) in cpl.log. the last lines read:

(seq_flux_atmocn_mct) computing only ocn albedos
(seq_mct_drv) : Calling map_lnd2atm_mct
(seq_mct_drv) : Calling map_ocn2atm_mct for mapping o2x_ox to o2x_ax
(seq_mct_drv) : Calling map_ocn2atm_mct for mapping xao_ox to xao_ax
(seq_mct_drv) : Calling map_ice2atm_mct for mapping i2x_ix to i2x_ax
(seq_mct_drv) : Calling mrg_x2a_run_mct
(seq_mct_drv) : Calling atm_init_mct


It seems that it stop initializating the atm model.

Any suggestions

Agustin

 

[fischer]

Hi Agustin,

What resolution, compset, and PEs layout are you using? You can find the
PEs layout in env_mach_pes.xml.

Chris

 

[agustin@unex_es]

The compset is B_RCP4.5_WACCM_C, the resolution is f19_g16 and the number of PES is 128 for each part (atm,lnd,ocn,cpl, NTAKS_ATM=NTAKS_LND=NTASK_OCN=NTASK_CPL= 128). Our machine is an HP superdome with 128 cores 1 TB ram.

Agustin

 

[eaton]

We were running into a strange failure trying to run with cam5 on an x86_64 platform using intel-11.0.074. The WACCM compset you're running with uses the cam4 physics package rather than cam5, so this may not be relevent. But in trying to find our problem we moved to another x86_64 platform with intel-11.1.059 and the problem went away. So it's possible that you're dealing with a compiler problem.

 

[agustin@unex_es]

Incredible but true. I compiled again the code with intel fortran 11.1.064 instead of intel fortran 11.0.74, and now it works.

Thank very much for your help

Agustin

 

[eaton]

Unfortunately it's fairly common that we spend alot of time chasing compiler bugs. For that reason we test using as many different compilers as is feasible. If code is running on 3 out of the 4 compilers that we're testing with then there's a good chance that the problem is due to a compiler bug. On the other hand, sometimes just the one compiler is interpreting the standard more strictly than the others, so modifying our code to make the strictest compiler happy improves the quality of our code base.