Re: Floating Point Exception Error when using Source.f

From: Alfredo Ferrari <alfredo.ferrari_at_cern.ch>
Date: Tue, 14 May 2013 22:36:15 +0200

... my experience is that gfortran-4.4.x often miscompiles Fluka routines.
We had another source routine, completely different, which was failing
when compiled with gcc-4.4.x while working perfectly ok with newer
versions of gfortran. This is also the reason why the "fff" script
as distributed refuses to compile on gfortran 4.x with x < 5

               Alfredo


+----------------------------------------------------------------------+
| Alfredo Ferrari || Tel.: +41.22.76.76119 |
| CERN-EN/STI || Fax.: +41.22.76.69474 |
| 1211 Geneva 23 || e-mail: Alfredo.Ferrari_at_cern.ch |
| Switzerland || |
+----------------------------------------------------------------------+

On Mon, 13 May 2013, Macfarlane, Lewis wrote:

> Dear Vasilis,
>
> Thank you for your prompt response.
>
> We are currently using GCC (Gfortran) version 4.4.7 on Ubuntu 12.04 64-bit version. We have tried to install FLUKA using GCC 4.5 and later versions, but we couldn't get this to work due to too many Fortran errors. The source.f does seem to compile and build okay using 4.4.7, we just get the error when trying to run the code.
>
> Could it be a 64-bit specific problem?
>
> Thanks,
> Lewis.
>
>
>
> -----Original Message-----
> From: Vasilis Vlachoudis [mailto:Vasilis.Vlachoudis_at_cern.ch]
> Sent: 13 May 2013 12:18
> To: Macfarlane, Lewis; fluka-discuss_at_fluka.org
> Subject: RE: Floating Point Exception Error when using Source.f
>
> Dear Lewis
>
> I've just tested your input + source.f with the latest fluka 2011.2b (g77 version) and it terminated the 5 cycles with no problem.
> Are you using the gfortran version (4.5, 4.6?) or the g77? Could you please try to test it with the latest version on your machine?
>
> Best Regards
> Vasilis
>
> ________________________________________
> From: owner-fluka-discuss_at_mi.infn.it [owner-fluka-discuss_at_mi.infn.it] on behalf of Lewis.Macfarlane_at_nuclear.co.uk [Lewis.Macfarlane_at_nuclear.co.uk]
> Sent: 11 May 2013 10:20
> To: fluka-discuss_at_fluka.org
> Subject: Floating Point Exception Error when using Source.f
>
> Dear FLUKA experts,
>
> I have some issues trying to run a simple (test) proton case with a source.f subroutine.
> The same subroutine has been used on earlier versions of Fluka/Flair without any issues. Current versions are:
>
> Fluka: 2011.2 (Feb. 2013)
> Flair: 1.1-1 (March 2013)
>
> The input file runs fine when source.f file is not attached, i.e. with a mono-energetic source.
> The source.f file allows the use of a simple rectangular spectrum with a conical beam profile defined by a divergence half-angle.
> When used with the source.f file, the code terminates with the error:
>
> *** Vacuum stopping: Ij, Pla, Ekin
>
> Within the log file, is:
>
> Program received signal 8 (SIGFPE): Floating-point exception.
>
> Backtrace for this error:
> + /lib/x86_64-linux-gnu/libc.so.6(+0x364a0) [0x7fbcdb71a4a0]
> + function kaskad_ (0x501D2C)
> at line 4139 of file kaskad.f
> + function feeder_ (0x4F67B6)
> at line 257 of file feeder.f
> + function flukam_ (0x40D055)
> at line 3095 of file flukam.f
> + function fluka (0x40785C)
> at line 289 of file fluka.f
> + /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xed) [0x7fbcdb70576d]
>
> I have attached my input file and source.f for your consideration. Any help
> would be greatly appreciated,
>
> Many thanks,
> Lewis MacFarlane.
>
>
>
Received on Wed May 15 2013 - 09:38:29 CEST

This archive was generated by hypermail 2.3.0 : Wed May 15 2013 - 09:39:00 CEST