Changes between Version 8 and Version 9 of 2011WP/2011Stream2/Dynamic Memory


Ignore:
Timestamp:
2011-01-04T12:17:16+01:00 (10 years ago)
Author:
trackstand
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • 2011WP/2011Stream2/Dynamic Memory

    v8 v9  
    486486I like Gurvan's suggestion of a module containing globally-accessible work-space arrays. We could add some error-checking functionality to this by having an 'in_use' flag for each work-space array in the module. Before using a work-space array, a developer should check that the appropriate flag is .FALSE. and if it is, set it to .TRUE. while they are using it. Once they are done using the array the flag should be set back to .FALSE. 
    487487 
     488I also note that there are different views on how to specify the domain decomposition (jpni and jpnj values). I was thinking that we would calculate them, choosing values such that the domains are as square as possible (following Italo's suggestion above) whereas Gurvan mentions reading them from a namelist file. I suggest that we support both options: by default jpni and jpnj can be calculated at run-time (this saves the user from having to change the namelist file when changing the number of processors they're running on) but this behaviour can be overridden by specifying them in the namelist file. 
     489 
     490(Note that this choice of domain decomposition and the shape of the sub-domains forms a part of the optimisation project presented by Stephen Pickles at the meeting in Southampton.) 
     491 
    488492---- 
    489493