[MITgcm-devel] [MITgcm-cvs] MITgcm/pkg/ctrl CVS Commit
Matthew Mazloff
mmazloff at ucsd.edu
Wed Mar 7 11:28:04 EST 2012
Hi Gael,
Actually, perhaps just keep your modification to ctrl_pack as just
using the mitgcm to pack requires other changes to the code anyway.
The best thing to do will be eventually figure out a low memory
packing algorithm, but for now I can just deal with hacking the code
so it is only used to pack.
And I have no major edits in pkg/ecco/forward_step.F & pkg/ecco/
the_main_loop.F, so no problem for me that they are retired.
Thanks,
Matt
On Mar 7, 2012, at 12:15 AM, Gael Forget wrote:
> Hi Matt,
>
> sorry for the hassle. I will try to get to this by the end of the
> week.
> (A few things I promised to other people I want to do first...)
>
> One thing that would be useful once we fix the issue would be to
> have this
> feature testreport'ed. It should avoid us from breaking it again. I
> may not be
> very responsive while away next weeks, but e.g. Patrick could help
> you with that.
>
> Another thing in passing, late march or so, I will take a crack at
> retiring
> pkg/ecco/forward_step.F & pkg/ecco/the_main_loop.F. You may want
> to backup whatever edits you may have made to those routines.
>
> Cheers,
> Gael
>
> On Mar 6, 2012, at 2:13 PM, Matthew Mazloff wrote:
>
>> Hi Gael,
>>
>> Due to the large amount of memory packing and unpacking takes, I
>> find it most computationally efficient to have separate executables
>> that only do the packing. I compile these executables as forward
>> code. So this check-in is problematic for me -- I will have to
>> comment it out. I do see the value in it, but we also have a flag
>> EXCLUDE_CTRL_PACK. Is there some way we could make the model to
>> pack in forward mode -- could we add another CPP option.....or is
>> there some compromise you can think of?
>>
>> Thanks
>> Matt
>>
>>
>>
>>
>>
>>
>>
>> On Feb 27, 2012, at 4:26 PM, Gael Forget wrote:
>>
>>> Update of /u/gcmpack/MITgcm/pkg/ctrl
>>> In directory baudelaire:/srv/scratch/gforget/MITgcm/pkg/ctrl
>>>
>>> Modified Files:
>>> ctrl_pack.F
>>> Log Message:
>>>
>>> - dont pack if not mitgcmuv_ad.
>>>
>>>
>>>
>>> _______________________________________________
>>> 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20120307/822649e5/attachment.htm>
More information about the MITgcm-devel
mailing list