[MITgcm-devel] no-unused-dummy-argument

[Dimitris Menemenlis]

Merci Jean-Michel.
I just updated linux_amd64_gfortran and now testreport works great!
I was running testreport on a machine with gfortran 4.1.2 when it failed, see below:

[dimitri at skylla build]$ gfortran -fconvert=big-endian -fimplicit-none -mcmodel=medium -Wall -Wno-unused-dummy-argument  -O0 -c cd_code_ini_vars.f
f951: error: unrecognized command line option "-Wno-unused-dummy-argument"
[dimitri at skylla build]$ gfortran -fconvert=big-endian -fimplicit-none -mcmodel=medium -Wall  -O0 -c cd_code_ini_vars.f
[dimitri at skylla build]$ gfortran --version
GNU Fortran (GCC) 4.1.2 20080704 (Red Hat 4.1.2-52)

I don't know at which version no-unused-dummy-argument became part of -Wall.
I got this information here: http://www.hep.by/gnu/gcc/gfortran/Error-and-Warning-Options.html
which I assume pertains to a recent version of grfortran but they don't specify which.

Thanks

Dimitris Menemenlis

On Oct 1, 2014, at 4:17 PM, Jean-Michel Campin <jmc at ocean.mit.edu> wrote:

> Hi Dimitris,
> 
> I've just removed it.
> Just curious: which was the version of gfortran that was causing
> this to stop ? And in which newer version this "no-unused-dummy-argument"
> is part of -Wall ?
> On baudelaire (gcc 4.4.5), the option no-unused-dummy-argument is not
> supported but it does not prevent to compile, so I was assuming it
> would be the same for older version as well.
> 
> Cheers,
> Jean-Michel
> 
> On Wed, Oct 01, 2014 at 09:28:20PM +0000, Menemenlis, Dimitris (329D) wrote:
>> Bonjour Jean-Michel, quick question:
>> for backward compatibility with MITgcm users that have slightly out of
>> date gfortran compilers on their workstations, would it be OK if I
>> comment out the no-unused-dummy-argument
>> in tools/build_options/linux_amd64_gfortran ?
>> 
>> Also this option is already implied by "-Wall" for newer compiler
>> versions, so I think that it is redundant, while it causes compilation
>> to stop for older gfortran versions.
>> 
>> D.
>> 
>> 
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel