Re: Co-Array Fortran
- From: Walter Spector <w6ws_xthisoutx@xxxxxxxxxxxxx>
- Date: Tue, 30 Oct 2007 17:48:58 -0700
gary.l.scott@xxxxxxxx wrote:
...
from walter spector
"a(i) = b(i) + c(i)[np]
The A and B references are local. But C is a coarray, and so would
load values from coarray process 'np'. (Which may or may not be local
-
depending on the value of np.) "
the last sentence seems to imply that with a [np] reference present,
it could denote either a local or remote reference depending on the
value. that superficially seem to conflict.
We are saying the same thing.
Upon startup, there are some fixed number of 'images' of the program
executing simultaneously. Each image has a complete (co)array C. C is not
'distributed'. If C is 100 elements long, it is 100 elements long on
EACH of the images. In this sense coarrays are like MPI programming.
Each image has its own version of the array.
So if, say, some image wants data from coarray C on image #10, it
would execute:
a(i) = b(i) + c(i)[10]
If, per chance, image #10 executed the above statement, the communication
would be to its own version of coarray C. Otherwise, the communication
would be a remote version of C (on image #10.)
Does that make sense?
Walt
.
- Follow-Ups:
- Re: Co-Array Fortran
- From: gary . l . scott
- Re: Co-Array Fortran
- References:
- Co-Array Fortran
- From: gary . l . scott
- Re: Co-Array Fortran
- From: Dan Nagle
- Re: Co-Array Fortran
- From: Gary Scott
- Re: Co-Array Fortran
- From: Walter Spector
- Re: Co-Array Fortran
- From: gary . l . scott
- Re: Co-Array Fortran
- From: Dan Nagle
- Re: Co-Array Fortran
- From: gary . l . scott
- Co-Array Fortran
- Prev by Date: Re: Co-Array Fortran
- Next by Date: Re: Co-Array Fortran
- Previous by thread: Re: Co-Array Fortran
- Next by thread: Re: Co-Array Fortran
- Index(es):
Relevant Pages
|