problem k-omega sst

[eka]

Hello,

I work with code_saturne 2.1.4.
I use unsteady flow simulation with k-omega,
when running, I have the following error in the listing :


SIGFPE signal (floating point exception) intercepted!

Call stack:
1: 0x7f83834b79e0 <+0x329e0> (libc.so.6)
2: 0x7f83840f6353 <+0x146353> (libsvml.so)
3: 0x581200 <> (cs_solver)
4: 0x4a7c51 <> (cs_solver)
5: 0x426b78 <> (cs_solver)
6: 0x422663 <> (cs_solver)
7: 0x7f83834a3bc6 <__libc_start_main+0xe6> (libc.so.6)
8: 0x420839 <> (cs_solver)
End of stack

Could anyone tell me what should i do?

Best Regards,
Eka

[Yvan Fournier]

Hello,

In your execution directory, could you run "addr2line --exe cs_solver 0x581200" and post the result ?
(0x581200 is the last address that appears in the backtrace in your "listing" file; testing with the values below can also help).

This should tell you the name of the source file in which the problem occurred, so it may help finding the cause.

Best regards,

Yvan

[eka]

Hello,

I have run ""addr2line --exe cs_solver 0x581200"
and the answer is : ??:0

Then i tried to use code_saturne version 1.3.3, it works.

Regards,
Eka

[Yvan Fournier]

Hello,

You might have some other error in the setup for version 2.1.

Did you try addr2line for the other addresses in the backtrace (from top to
bottom) ?

Regards,

Yvan

[eka]

Hello,

I have tried to run all of the data and get the same results.
Thank you.

Best Regards,
Eka

[eliberge]

Hello,

the same problem appears using LES turbulence model.

I am working on the same mesh file than Eka, and the end of the listing file is :
Signal SIGFPE (exception en virgule flottante) intercepté !

Pile d'appels :
1: 0x7fd8bc5da9e0 <+0x329e0> (libc.so.6)
2: 0x7fd8bdae423c <+0x12c23c> (libsvml.so)
3: 0x7fd8bdae41c3 <+0x12c1c3> (libsvml.so)
4: 0x59bb9e <> (cs_solver)
5: 0x4a88fc <> (cs_solver)
6: 0x425738 <> (cs_solver)
7: 0x421223 <> (cs_solver)
8: 0x7fd8bc5c6bc6 <__libc_start_main+0xe6> (libc.so.6)
9: 0x4208f9 <> (cs_solver)
Fin de la pile

and the answer of the command :
addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x4208f9
is :
/usr/src/packages/BUILD/glibc-2.11.1/csu/../sysdeps/x86_64/elf/start.S:116

Regards,

Erwan

[Yvan Fournier]

Hello,

Could you try addr2line on other addresses from the backtrace (especially 0x59bb9e) ?

[eliberge]

Thank you for your help.

Please find the results of the addr2line command :

>addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x4208f9
/usr/src/packages/BUILD/glibc-2.11.1/csu/../sysdeps/x86_64/elf/start.S:116
>addr2line --exe /lib64/libc.so.6 0x7ff0f7189bc6
??:0
>>addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x421223
??:0
>addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x425738
??:0
>addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x4a88fc
??:0
>addr2line --exe /Calcul/Apps2/saturne4.2.4/v217/libexec/code_saturne/cs_solver 0x59bb9e
??:0
>addr2line --exe /Calcul/Apps/intel/lib/intel64/libsvml.so 0x7ff0f86a71c3
??:0
>addr2line --exe /Calcul/Apps/intel/lib/intel64/libsvml.so 0x7ff0f86a723c
??:0
>addr2line --exe /lib/libc.so.6 0x7ff0f719d9e0
??:0

Regards,

Erwan

[Yvan Fournier]

Hello,

The addr2line information is less useful than I hoped. Using a version of the code compiled in debug mode (see installtion manual), it would be more useful.

I would suggest trying version 2.2.1 to see if you have the same problem, but this means upgrading your data setup.

I could also test your case if you post it here, but I probably won't have the time to do this before a few weeks.

Best regards,

Yvan

[eliberge]

Thank you.

I will try the 2.2.1 version.

Best regards,

Erwan