‘Segmentation fault’ error: [cn121:363237:0:363237] Caught signal 11 (Segmentation fault: address not mapped to object at address 0x7ffe2b75a5d0)

[shravan]

I've been having trouble trying to run wrf.exe. In HPC, I configured WRF-v4.6.0 with dmpar and compiled it, changing a constant value in module_thompson.F.
real.exe runs successfully, but I am facing a segmentation fault while running wrf.exe. I've again recompiled with the original file without any changes, but even though facing the same issue. Can someone please help me?

[cn121:363237:0:363237] Caught signal 11 (Segmentation fault: address not mapped to object at address 0x7ffe2b75a5d0)
==== backtrace (tid: 363237) ====
0 0x00000000016bb953 solve_em_() ???:0
1 0x00000000014fd76c solve_interface_() ???:0
2 0x00000000005b529d module_integrate_mp_integrate_() ???:0
3 0x000000000041a411 module_wrf_top_mp_wrf_run_() ???:0
4 0x000000000041a3c4 MAIN__() ???:0
5 0x000000000041a352 main() ???:0
6 0x0000000000022555 __libc_start_main() ???:0
7 0x000000000041a269 _start() ???:0
=================================


I attach the files

 

[kwerner]

Hi,
Can you package all of your rsl* files into a single *.tar file and attach that?
1) Based on the rsl files you shared, it looks like the model stops right away. Typically when this happens, it means something is wrong with the input. Can you check the input (met_em files) and make sure you aren't missing any data at any level?
2) If the above is not the issue, can you try running this again, just using the default namelist, making changes only to the times, domain settings, timestep, etc. so that it matches your setup. Do not modify the physics or dynamics options. If that runs, then you'll know the issue is related to a setting you've changed. Make sure to do this test using the non-modified version of WRF.
3) Another test you can try is to use more processors. Try using something more like 36 or 64.

If all of that still fails, please package all of your rsl* files into a single *.tar file and attach that, along with the namelist you used to produce those files. Thanks!