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

Error running compset FSDWSF (ERT)

T

thibaut_lurton@cnrs-orleans_fr

Member
Hi to all,I've been trying to run the out-of-the-box compset FSDWSF, which is CAM WACCM with sulfur chemistry and CARMA sulfate.The complete compset configuration is GEOS_CAM4%WCSF_CLM40%CN_CICE%PRES_DOCN%DOM_RTM_SGLC_SWAV.I successfully created and built the case, with a f19_f19 resolution and intel14 compiler, on our local machine.Creating and building the ERT production test also succeded, however, when running this ERT test, an error occurs, under the form: forrtl: severe (408): fort: (2): Subscript #1 of the array ATRCR has value 1 which is greater than the upper bound of 0
Image              PC                Routine            Line        Source            
cesm.exe           0000000006527870  Unknown               Unknown  Unknown
cesm.exe           0000000003F9BC0E  ice_itd_mp_aggreg         479  ice_itd.F90
cesm.exe           00000000048DC8E2  ice_init_mp_init_         912  ice_init.F90
libiomp5.so        00002AAAABAA28A3  Unknown               Unknown  Unknown

(This is how my cesm.log.* file ends in the /run/ subdirectory.)I'm really not sure how I should understand this error. Any advice would be greatly appreciated. I can provide further logs if needed.Cheers,Thibaut.
 
J

jedwards

CSEG and Liaisons
Staff member
THis is an issue that came up in intel 14 - at line 479 of ice_itd.F90 you have:      call compute_tracers (nx_block,     ny_block,   &                            icells,   indxi,   indxj, &                            ntrcr,    trcr_depend,    &                            atrcr     , aice(:,:),    &                            vice (:,:),   vsno(:,:),  &                             trcr(:,:,:)) If you remove the explicit array syntax so that you have     call compute_tracers (nx_block,     ny_block,   &                            icells,   indxi,   indxj, &                            ntrcr,    trcr_depend,    &                            atrcr     , aice,    &                            vice ,   vsno,  &                             trcr)That should solve the problem. 
 
S

santos

Member
There are actually several bugs in Intel 14 that affect CESM; the FV dycore in CAM might crash because of one of these. What model version are you using?
 
T

thibaut_lurton@cnrs-orleans_fr

Member
Thanks Jim for the tip; the ERT test completed and passed ok after that being fixed in the code. My only concern is it took quite a while (4h for a 3-day ERT) but that might be due to my turning the debug option on...Sean, we are using version 1.2.2 of CESM, with CAM4 for this precise test (as specified in the out-of-the-box FSDWSF compset). Are there any specific configurations that we should be vigilant of using intel14?Thanks again for your help.Thibaut.
 
S

santos

Member
If you are not encountering a problem now, it is probably OK. If at some point you do have a glibc error at run time, you can try replacing the FV version of spmd_dyn with the attached version (e.g. in SourceMods, or in models/atm/cam/src/dynamics/fv).
 
You must log in or register to reply here.
Top