(j3.2006) Question about co-arrays

[Bill Long]

To continue with Aleks' comment below, if sometime in the future we 
extend C interoperability to UPC interoperability, then this issue 
should be examined again.  Such interoperability has already been 
requested, but I firmly feel that f08 is too early to consider this.

Cheers,
Bill


Aleksandar Donev wrote:

>Hi,
>In addition to what John said, to address specifically:
>  
>
>>In light of Note 5.10 [91:8+1-3]
>>    
>>
>The difference between a co-array of a derived types with pointer components 
>and of type C_PTR is that the user can do something useful with pointer 
>components, specifically, they can have data-refs of the form:
>x[i]%p
>and the compiler knows, from the reference, that this is remote data, and will 
>do the magic to get it.
>On the other hand, there is nothing useful that can be done with a C pointer 
>x[i], since C does not understand co-arrays and cannot fetch data from remote 
>images. We specifically disallow such co-indexed pointers in C_LOC and 
>C_F_POINTER because it makes no sense to claim interop for something that is 
>not even part of the C standard.
>Best,
>Aleks
>_______________________________________________
>J3 mailing list
>J3 at j3.scs.gmu.edu
>http://j3.scs.gmu.edu/mailman/listinfo/j3
>  
>

-- 
Bill Long                                   longb at cray.com
Fortran Technical Support    &              voice: 651-605-9024
Bioinformatics Software Development         fax:   651-605-9142
Cray Inc., 1340 Mendota Heights Rd., Mendota Heights, MN, 55120

            

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://j3.scs.gmu.edu/pipermail/j3/attachments/20070207/1dece953/attachment.html