[MITgcm-support] MITgcm, gfortran and Archer

[Jody Klymak]

Hi Renske,

Try replacing the last "&" in your namelists with a "/".  No idea why the conventions are different, and I think there are compile flags to choose one or the other.  

Cheers,  Jody


On Apr 4, 2014, at  6:43 AM, Renske Gelderloos <Renske.Gelderloos at earth.ox.ac.uk> wrote:

> Dear fellow MITgcm users (specifically those in the UK working on Archer, but if anyone else knows the solution I would be happy too of course!),
> 
> I am running a simple setup of the MITgcm on Archer. So far, I have been compiling with Intel and that worked fine (apparently this had some other issues though which is why I am trying to switch). I am now trying to compile with gfortran, but this gives me the following runtime error message:
> 
> At line 2595 of file ini_parms.f (unit = 11, file = './gfortrantmpMV1kxT')
> Fortran runtime error: Cannot match namelist object name .
> 
> Line 2595 of file ini_parms.f refers to reading in PARM03 of the 'data' file, which is in my case 
> 
>  &PARM03
>  niter0=0.,
>  endtime=2400.,
>  abEps=0.1,
>  deltaT=240.,
>  cAdjFreq=0.,
>  taveFreq=2592000.,
>  monitorFreq=2592000.,
>  pChkptFreq=155520000.,
>  forcing_In_AB=.FALSE.,
>  pickupStrictlyMatch=.FALSE.,
>  &
> 
> Again, it works fine with intel, so do not think there is anything wrong with my data file, but perhaps there is a compatibility issue with gfortran or some other known bug? I tried gcc/4.8.1 and gcc/4.8.2, which are the two available modules on Archer. 
> 
> Renske
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support