[MITgcm-devel] [MITgcm-cvs] MITgcm/doc CVS Commit

[Jean-Michel Campin]

Hi Martin,

You are right. What I did fixed the isomip ad test but
not the other flag combination.
By the way, if we could avoid to have some declaration (integer i, j)
after some executed instructions (CALL SHELFICE_COST_SHIFWFLX)
it would look better.

> Do you want me to check this in?
Yes, please do. 

Thanks,
Jean-Michel

On Tue, Mar 13, 2012 at 04:27:26PM +0100, Martin Losch wrote:
> Hi Jean-Michel,
> 
> you are too fast for me. I think that this initialisation should be moved down to just before the comment: "c--   Sum up all contributions." so that it is valid for all flag combinations.
> 
> Sorry for this bug. Do you want me to check this in?
> 
> Martin
> 
> On Mar 13, 2012, at 4:23 PM, Jean-Michel Campin wrote:
> 
> > This should be fixed now.
> > Jean-Michel
> > 
> > On Tue, Mar 13, 2012 at 10:43:02AM -0400, Jean-Michel Campin wrote:
> >> Hi Martin,
> >> 
> >> It looks like there is a problem in the changes you made in 
> >> pkg/shelfice:
> >> the 2 adjoint testreport I run over night using open64 compiler
> >> + check-uninitialise-var debugging option are failing
> >> for isomip experiments:
> >> http://mitgcm.org/testing/results/2012_03/tr_aces-opa_20120313_0/summary.txt
> >> and:
> >> http://mitgcm.org/testing/results/2012_03/tr_acesgrid-tuv_20120313_0/summary.txt
> >> but were running fine previously.
> >> 
> >> It stops with a floating-point-exception in:
> >>> stopped at [subroutine shelfice_cost_final(integer*4):4481 0x08422b92]
> >> (at the end of the forward run, after writing pickup files).
> >> 
> >> If you have an idea regarding which part to check.
> >> 
> >> Cheers,
> >> Jean-Michel
> >> 
> >> On Mon, Mar 12, 2012 at 12:51:37PM -0400, Martin Losch wrote:
> >>> Update of /u/gcmpack/MITgcm/doc
> >>> In directory forge:/tmp/cvs-serv20834
> >>> 
> >>> Modified Files:
> >>> 	tag-index 
> >>> Log Message:
> >>> describe changes in shelfice
> >>> 
> >>> 
> >>> _______________________________________________
> >>> MITgcm-cvs mailing list
> >>> MITgcm-cvs at mitgcm.org
> >>> http://mitgcm.org/mailman/listinfo/mitgcm-cvs
> >> 
> >> _______________________________________________
> >> 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
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel