[Carpet] component numbers; CCTK_DELTA_SPACE from one component to another

[Erik Schnetter]

On May 23, 2006, at 14:28:26, Jonathan Thornburg wrote:

> Hi, Erik,
>
>>> If I do a BEGIN_LOCAL_COMPONENT_LOOP ... END_LOCAL_COMPONENT_LOOP,
>>> is it guaranteed that Carpet::component will sequence through the
>>> contiguous set of integers from 0 to
>>> Carpet::vhh.at(Carpet::map)->components(Carpet::reflevel)-1
>>> inclusive?  Or is it possible that the local component numbers will
>>> be some subset of this?
>>
>> The global component numbers run from 0 to N-1, where N is often  
>> the number of
>> processors.  The local loop iterates over that subset which is  
>> local to a
>> processor, i.e., most of the time a single integer in the range 0  
>> to N-1.
>
> Is "that subset which is local to a processor" always a single
> continguous interval A <= component_number <= B for some A and B
> or might it be a more general subset of [0,N-1]?

It will in general not be contiguous.

-erik

-- 
Erik Schnetter <schnetter at cct.lsu.edu>

My email is as private as my paper mail.  I therefore support encrypting
and signing email messages.  Get my PGP key from www.keyserver.net.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : /archives/developers/attachments/20060523/f657ee53/attachment.pgp