[MITgcm-devel] pkg/streamice

[Daniel Goldberg]

Sent from my iPhone

> On Sep 11, 2014, at 3:51 PM, Jean-Michel Campin <jmc at ocean.mit.edu> wrote:
> 
> Hi Dan,
> 
> OK, it seems that streamice_initialise_petsc call can move safely to
> streamice_init_fixed.F (@ the top). 
> 
> Regarding streamice_finalize_petsc, you are right, there is no
> obvious place to put this call. Part of the reason is this useSIGREG
> implementation (in forward_step.F) that allows to stop at any time
> (but does not work will all packages, and not sure about MPI).
> 
> I will try to see what we can do about this. May be adding a new S/R
> at the end of forward_step.F, hidden to TAF like do_the_model_io.F,
> and only called if modelEnd=True; with a 2nd call elsewhere
> for the case useSIGREG=F.
> 
> Regarding streamice_readparms.F:
>> do you mean the lines 146-154 of version 1.4?
> Yes, it was this part. But I realised that I forgot to set
> STREAMICEison=F if useStremIce=F (although STREAMICEison does not
> seem to be used in many places). I can add back these changes if you agree.

Hi j-m,

Sounds good, I have no problem with that code added back in. You are right I do not really use streamiceison but I am ok with that initialization as well. Thanks for doing that.

Let me know if you have a good solution for streamice_finalize_petsc.

Dan

> 
> Cheers,
> Jean-Michel
> 
>> On Thu, Sep 11, 2014 at 10:39:41AM +0100, Daniel Goldberg wrote:
>> Hi Jean Michel
>> 
>> On Thu, Sep 11, 2014 at 7:52 AM, Jean-Michel Campin <jmc at ocean.mit.edu>
>> wrote:
>> 
>>> Hi Dan,
>>> 
>>> 2 things about pkg/streamice:
>>> 
>>> 1) I would like to move the call to streamice_initialise_petsc
>>> from the_model_main.F to the top of streamice_init_fixed.F
>>> (it would be a better place for such an initialisation).
>>> The only relevant streamice call in between is streamice_readparms.F
>>> and I don't think petscInitialize has any overlap with
>>> streamice_readparms.
>>> Can you confirm ?
>> 
>> i think this is fine. i don't think there are any petsc calls before
>> packages_init_fixed.
>> 
>> note there is still a call to streamice_finalize_petsc in the_model_main,
>> and i don't know a good place to hide this, since there are no
>> "packages_cleanup" calls or anything like that..
>> 
>> 
>>> 
>>> 2) in streamice_readparms.F, it seem that you removed during revision 1.5
>>> (on Jun 04, 2014) all the changes I made at revision 1.4 (on May 27,
>>> 2014).
>>> Can you explain why you remove this modification ?
>>> 
>>> Since this was about calling STREAMICE_READPARMS even when
>>> useSteamIce=F , I added, at the same time, at the top of
>>> streamice_readparms.F,
>>> something like: IF (.NOT.useSteamIce ) RETURN
>>> now this quick return is gone but STREAMICE_READPARMS is still called
>>> when useSteamIce=F
>> 
>> i apologize. i would not have done this -- the only possibility is i was
>> not making changes to an updated version. do you mean the lines 146-154 of
>> version 1.4? Shall i add these back in? i am not sure what they are meant
>> to do, will it impact anything else?
>> 
>> 
>>> Cheers,
>>> Jean-Michel
>>> 
>>> 
>>> _______________________________________________
>>> MITgcm-devel mailing list
>>> MITgcm-devel at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>> 
>> 
>> 
>> -- 
>> 
>> Daniel Goldberg, PhD
>> Lecturer in Glaciology
>> School of Geosciences, University of Edinburgh
>> Geography Building, Drummond Street, Edinburgh EH8 9XP
>> 
>> 
>> em: D <dgoldber at mit.edu>an.Goldberg at ed.ac.uk
>> web: http://ocean.mit.edu/~dgoldberg
> 
>> _______________________________________________
>> 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