[MITgcm-devel] Beaufort experiment on mac os x

[Martin Losch]

Dimtris,
go ahead.
M.

On Mar 20, 2012, at 1:54 AM, Menemenlis, Dimitris (3248) wrote:

> Torge, thank you for checking.
> 
> Martin, can I go ahead and comment out
> NOOPTFILES = 'gad_c4_adv_x.F gad_u3_adv_x.F' 
> from darwin_amd64_gfortran
> 
> Patrick, you are not allowed to read this message ;o)
> 
> Cheers
> 
> Dimitris Menemenlis
> 
> On Mar 19, 2012, at 11:59 AM, Torge Martin wrote:
> 
>> Okay, new beaufort code compiles and runs on both MacBook Air with Lion 
>> and gcc 4.7.0 as well as MacPro with Snow Leopard and gcc 4.6.2 :-)
>> 
>> In both cases I used the default optfile in MITgcm/tools/build_options
>> darwin_amd64_gfortran and darwin_ia32_gfrotran
>> 
>> darwin_amd64_gfortran should be updated to really have NOOPTFILES = ' '
>> Right now it is still  NOOPTFILES = 'gad_c4_adv_x.F gad_u3_adv_x.F' if 
>> -ieee option is not(!) set, which is default
>> If I remove this line, I can still compile and run the code without any 
>> problems.
>> 
>> Torge
>> 
>> On 3/19/12 9:36 AM, Menemenlis, Dimitris (3248) wrote:
>>> Hi Torge, could I trouble to check out clean code and test the Beaufort set-up one last time.
>>> 
>>> Jean-Michel and Martin, have modified problematic lines in code so that -O3 optimization
>>> should now work without requiring any modifications on your end.
>>> 
>>> It'd be great to know if it solves problem for specific flavor of compilers that you use on your
>>> Mac Air and workstation.
>>> 
>>> Cheers
>>> 
>>> Dimitris Menemenlis
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel